Method and device for real-time detection, location and determination of the speed and direction of movement of an area of relative movement in a scene
Abstract
According to the present invention, a method and a device for detecting, determining location, speed, and direction of the relative movement of a television image section in real time are proposed. The proposed method provides the possibility for the space-time processing of an input video signal (S). The time-domain processing of the video signal consists in determining the variations of the intensity of each image element in the current frame and of the corresponding image element in the preceding frame, producing a binary signal (DP) indicating the presence or absence of the significant intensity change and a digital signal (CO) indicating the intensity of the said change. The space-domain processing of the video signal consists in determining, for each of the mentioned signals, the element of the image sensing array designed for scanning the image frames. The array element coordinates are used for detecting and determining the parameters of the relative movement of a television image section. For the realization of the method, a device is used that contains a time-domain signal processor (15), a memory unit (16), a space-domain signal processor (17), a time delay unit (18), a timing unit (20), and an adjusting unit (19).
Term
No projected expiry on record.
- Priority
- Filed
- Granted
- Today
14 claims: 6 independent, 8 dependent
- 1A method of real-time detecting and determining the location of the plot being carried out relative movement within the observation zone observed by using a surveillance system that has a digital video signal of the type included in the output the sequence of the corresponding frames, each of which consists of a sequence strings, each of which consists of a sequence of pixels, as well as a definition speed and oriented direction of movement, which differs in that which is to implement the output digital video signal of such a sequence operations:1. Спосіб здійснення у реальному часі виявлення и визначення місцеположення ділянки, що здійснює відносне переміщення у межах зони спостереження, за якою спостерігають за допомогою системи спостереження, що має на виході цифровий відеосигнал такого типу, що включає послідовність відповідних кадрів, кожний з яких складається з послідовності рядків, кожний з яких складається з послідовності пікселів, а також визначення швидкості й орієнтованого напряму переміщення, який відрізняється тим, що полягає у здійсненні з вихідним цифровим відеосигналом такої послідовності операцій: Smoothing processing of the mentioned output digital video signal using a digital constant whose numerical value can vary independently for each pixel of said output signal;згладжувальна обробка згаданого вихідного цифрового відеосигналу з використанням цифрової константи, чисельне значення якої може змінюватися незалежно для кожного піксела згаданого вихідного сигналу;introduction into the OP, first, the frame of the said output signal, which has undergone anti-aliasing, and, secondly, the corresponding smoothing constant this frame;введення в ЗП, по-перше, кадру згаданого вихідного сигналу, що зазнав згладжування, і, по-друге, константи згладжування, що відповідає даному кадру;Time processing, which includes for each pixel position, determining, first, the presence and, secondly, the magnitude of a significant change in the level pixel signal when moving to the current frame directly the previous smoothed and inserted in the FP frame, and the generation of two digital signals, the first of which is a binary one-bit signal with two possible values, one of which indicates the presence, and the other indicates the absence significant change in the transition between two consecutive frames, and the value said binary signal changes the value of the introduced constant in the ZΠ by reducing it, if said signal indicates a significant change, or an increase - if the signal is mentioned indicates the absence of such a change, and the reduction or increase is carried out step by step, while the second digital signal, that is, a signal characterizing the value the change is a multi-bit signal with a certain limited number of bits that quantitatively determine the magnitude of this change;and часова обробка, яка включає, для кожної піксельної позиції, визначення, по-перше, наявності і, по-друге, величини значущої зміни рівня піксельного сигналу при переході до поточного кадру від безпосередньо попереднього згладженого й введеного у ЗП кадру, і генерування двох цифрових сигналів, перший з яких є двійковим однобітовим сигналом із двома можливими значеннями, одне з яких вказує на наявність, а інше вказує на відсутність значущої зміни при переході між двома послідовними кадрами, причому значення згаданого двійкового сигналу змінює значення згаданої введеної в ЗΠ константи шляхом її зменшення, якщо згаданий сигнал вказує на наявність значущої зміни, або збільшення - якщо згаданий сигнал вказує на відсутність такої зміни, причому зменшення або збільшення здійснюють покроково, тоді як другий цифровий сигнал, тобто сигнал, що характеризує величину зміни, є багатобітовим сигналом з певною обмеженою кількістю бітів, що кількісно визначають величину даної зміни;і spatial processing, including performing such operations for Each frame of an incoming digital video signal: the distribution of pixel values only for a part of the pixels of the frame that correspond to a specific observation moment (the part scanned through the matrix, over a time interval that corresponds to the frame length), first, the binary signal and, and secondly, the mentioned digital signal characterizing the magnitude of the change, by matrix, in which the number of rows and columns is insignificant compared, respectively, with the number of rows and the number of pixels per line in the video signal, for receiving characteristics of pixel values;definition in this two-dimensional a matrix form of a region in which said binary signal has a definite value a specific value, indicating the presence or absence of a meaningful change, and said digital signal characterizing the magnitude of the change is changing (or not changes) to a significant value by the pixels of the matrix, which are adjacent along a certain direction, starting from the original pixel, within the same range plot of a frame and, accordingly, at the same moment of observation;and generating signals indicating the presence and location of the land plot relative displacement, and also indicate the relative velocity in the transition from the frame to the frame, as well as the direction of the movement, if it occurs, relative to the environment, and these signals are determined by the instantaneous matrix the division of said two digital signals, a binary signal and a signal that characterizes the magnitude of the change. просторова обробка, що включає виконання таких операцій для кожного кадру вхідного цифрового відеосигналу: розподілення піксельних значень лише для частини з пікселів кадру, що відповідають певному моменту спостереження (частини, що сканується через згадану матрицю, впродовж інтервалу часу, що відповідає тривалості кадру), по-перше, згаданого двійкового сигналу і, по-друге, згаданого цифрового сигналу, що характеризує величину зміни, по матриці, в якій кількість рядків і стовпців є незначною у порівнянні, відповідно, з кількістю рядків і кількістю пікселів у рядку у відеосигналі, для отримання характеристик піксельних значень;визначення у цій двомиттєвій матричній формі такої ділянки, в якій згаданий двійковий сигнал має певне конкретне значення, вказуючи на наявність або відсутність значущої зміни, і згаданий цифровий сигнал, що характеризує величину зміни, змінюється (або не змінюється) на значущу величину по пікселах згаданої матриці, які є суміжними вздовж певного напряму, починаючи від початкового піксела, в межах тієї саме ділянки кадру і, відповідно, в той саме момент спостереження;і генерування сигналів, що вказують на наявність і місцеположення ділянки, що здійснює відносне переміщення, і також указують відносну швидкість при переході від кадру до кадру, а також напрям цього переміщення, якщо воно відбувається, відносно оточення, причому ці сигнали визначаються за згаданим миттєвим матричним розподілом згаданих двох цифрових сигналів, двійкового сигналу й сигналу, що характеризує величину зміни.
- 6Method according to any of the preceding paragraphs, which is different the fact that the mentioned constant has the form 2p, where p is a number not exceeding 16, hence - which can be expressed not more than four bits, and this constant is reduced or increased by way of subtracting or adding a unit from / to p. 6. Спосіб за будь-яким із попередніх пунктів, який відрізняється тим, що згадана константа має форму 2p, де p - число, що не перевищує 16, отже - яке може бути виражено не більш ніж чотирма бітами, причому цю константу зменшують або збільшують шляхом віднімання або додавання одиниці від/до р.
- 8Device for real-time detection of and determining the location of a site that carries out relative movement within Observation areas observed by the observation system which outputs a digital video signal of the type containing the sequence the corresponding frames, each of which consists of a sequence of rows, each of which consists of a sequence of pixels, as well as speed and orientation the direction of movement, using the method described above, and this the device receives at its input the output video signal that is different that includes:8. Пристрій для здійснення у реальному часі виявлення й визначення місцеположення ділянки, що здійснює відносне переміщення у межах зони спостереження, за якою спостерігають за допомогою системи спостереження, що має на виході цифровий відеосигнал такого типу, що містить послідовність відповідних кадрів, кожний з яких складається з послідовності рядків, кожний з яких складається з послідовності пікселів, а також визначення швидкості й орієнтованого напряму переміщення, із використанням описаного вище способу, причому цей пристрій приймає на своєму вході згаданий вихідний відеосигнал, який відрізняється тим, що включає в себе: means (15) for smoothing the said output digital video signal using a digital constant (CO), the numerical value of which may vary independently for each pixel of said output signal;засіб (15) для згладжування згаданого вихідного цифрового відеосигналу з використанням цифрової константи (СО), чисельне значення якої може змінюватися незалежно для кожного піксела згаданого вихідного сигналу;a means (16) for entering in the SP, first, a frame of the said Output signal (LI) that has undergone anti-aliasing, and, secondly, the corresponding smoothing constant (SI) this frame засіб (16) для введення в ЗП, по-перше, кадру згаданого вихідного сигналу (LI), що зазнав згладжування, і, по-друге, константи згладжування (СІ), що відповідає даному кадру, time processing unit (15) for analysis, for each pixel position in the frame of the input video signal, change over time pixel values for the same pixel position when switching to the current frame from the front of the smoothed and inserted in the frame of the Z. said digital signal, and this unit includes a PL, executed from the ability to receive, store and publish information about the previous one smoothing frame, comparator (15a), designed to determine whether or not exceeds the absolute value of the difference between the pixel signal of the current pixel and the magnitude that characterizes the pixel signal in this same pixel position in a predetermined frame stored in said RF, a threshold value, by generating a binary or one-bit signal (DP) with two values, one of which indicates exceeding the threshold, and the other indicates the absence of such Excess, as well as computing device (15s), executed with the possibility determine the multiphase digital signal characterizing the change value, with a small number of bits whose value depends on the value of the value change the same pixel when moving to the current frame directly pre-smoothed and digitally inserted digital video signal;and блок часової обробки (15) для аналізування, для кожної піксельної позиції у кадрі вхідного відеосигналу, змінювання із плином часу значень піксельного сигналу для тієї саме піксельної позиції при переході до поточного кадру від безпосередньо переднього згладженого й введеного у ЗП кадру згаданого цифрового сигналу, причому цей блок включає в себе ЗП, виконаний із можливістю приймати, зберігати й видавати інформацію про відповідний попередній згладжений кадр, засіб порівняння (15а), призначений для визначення, чи не перевищує абсолютне значення різниці між піксельним сигналом поточного піксела і величиною, що характеризує піксельний сигнал у цій самій піксельній позиції у попередньому кадрі, що зберігається у згаданому ЗП, певне порогове значення, шляхом генерування двійкового, або однобітового, сигналу (DP) з двома значеннями, одне з яких вказує на перевищення порогового значення, а інше вказує на відсутність такого перевищення, а також обчислювальний засіб (15с), виконаний з можливістю визначати багатобітовий цифровий сигнал, що характеризує величину зміни, з невеликою кількістю бітів, значення якого залежить від величини зміни значення того саме піксела при переході до поточного кадру від безпосередньо попереднього згладженого і введеного в ЗП кадру цифрового відеосигналу;і a spatial processing unit (11), the input of which is fed said serial binary digital signal and a digital signal characterizing the magnitude of the change generated by the time processing unit for the pixels of a certain frame, and said spatial processing unit includes a means that characterizes the value of the pixels, which divides only the two- digit signal and signal characterizing the magnitude of the change corresponding to the same moment of time, that is, one frame, by matrix (21), in which number of rows and the number of columns is smaller than, respectively, the number of rows and the number pixels on the line in the frame of the mentioned digital video signal, and the frame scans through the matrix over the corresponding time interval frame length, localization tool to determine plot-aggregate pixels within the matrix, in which at a particular time point said binary signal has a certain specific value, and a means for determination in the matrix of a plots-pixel aggregate, in which at that instant of time the digital signal characterizing the magnitude of the change is changing for the adjacent ones pixels to a significant value, as well as a tool that is in response to the instructions from the two previous means generates the signals that determine the said plot-set of pixels, and, therefore, indicate the presence and location the area carrying out relative movement within the surveillance zone, as well at relative speed when switching from frame to frame and orientation direction movement of the area when it really moves relative to the environment. блок (11) просторової обробки, на вхід якого подаються згадані послідовні двійковий цифровий сигнал і цифровий сигнал, що характеризує величину зміни, що генеруються блоком часової обробки для пікселів певного кадру, причому згаданий блок просторової обробки включає в себе засіб, що характеризує значення пікселів, який здійснює розподілення лише двійкового сигналу й сигналу, що характеризує величину зміни, що відповідають тому саме моменту часу, тобто одному кадру, по матриці (21), в якій кількість рядків і кількість стовпців є меншими, ніж, відповідно, кількість рядків і кількість пікселів у рядку в кадрі згаданого цифрового відеосигналу, причому кадр сканується через згадану матрицю впродовж інтервалу часу, що відповідає тривалості кадру, локалізувальний засіб для визначення ділянки-сукупності пікселів в межах згаданої матриці, в якій у певний конкретний момент часу згаданий двійковий сигнал має певне конкретне значення, і засіб для визначення у згаданій матриці ділянки-сукупності пікселів, в якій у той саме момент часу цифровий сигнал, що характеризує величину зміни, змінюється для суміжних пікселів на значущу величину, а також засіб, який у відповідь на вказівки від двох попередніх засобів генерує сигнали, що визначають згадану ділянку-сукупність пікселів, і, отже, вказують на наявність і місцеположення ділянки, що здійснює відносне переміщення у межах зони спостереження, а також на відносну швидкість при переході від кадру до кадру і орієнтований напрям переміщення згаданої ділянки, коли вона дійсно переміщається відносно оточення.
- 11Device for pp. 8, 9 or 10, which is different because the said means (15s, 15d) for the anti-aliasing has an input for receiving the said digital video signal S (PI) and defines for each successive pixel the frame of the video signal Smooth signal (LO) which characterized by the fact that changes over time of the input digital video signal are reduced by using the threshold signal received on the other the input, and the constant (CO) correlated with each pixel position of the frame, the meaning of which varies consecutively in such a way that smoothing leaves unchanged, although it makes it weaker, the nature of the changes to the incoming digital a video signal, and said anti-aliasing means operates in conjunction with referred to by ZΠ (16), which serves for receiving, storing and displaying the values of the smoothed signal and the mentioned Updating Constants for each plot-pixel frame combination, as well Also gives for each pixel position at least a sequence of values updated constant and binary signal values, indicating whether or not the threshold value is exceeded by the absolute value of the difference between the value of the pixel and its smoothed value;and the said block spatial processing makes the distribution of a matrix with a reduced number lines and columns of information on the outputs of said means for smoothing, and It is the sequential values of the mentioned constant and the mentioned one-bit signal. 11. Пристрій за пп. 8, 9 або 10, який відрізняється тим, що згаданий засіб (15с, 15d) для згладжування має вхід для приймання згаданого цифрового відеосигналу S(PI) і визначає для кожного послідовного піксела кадру згаданого відеосигналу згладжений сигнал (LO), який характеризується тим, що зміни з плином часу вхідного цифрового відеосигналу зменшуються шляхом використання порогового сигналу, що приймається на іншому вході, і константи (СО), співвіднесеної з кожною піксельною позицією кадру, значення якої змінюється послідовно таким чином, що згладжування залишає незмінним, хоча і робить його слабкішим, характер змін вхідного цифрового відеосигналу, причому згаданий засіб для згладжування функціонує у взаємодії із згаданим ЗΠ (16), який служить для приймання, зберігання і видавання значень згладженого сигналу і згаданої константи, що оновлюються, для кожної ділянки-сукупності пікселів кадру, а також видає для кожної піксельної позиції принаймні послідовність значень оновленої константи і значень двійкового сигналу, які вказують, чи перевищується згадане порогове значення абсолютним значенням різниці між значенням піксела і його згладженим значенням;причому згаданий блок просторової обробки здійснює розподілення по матриці із зменшеною кількістю рядків і стовпців інформації з виходів згаданого засобу для згладжування, а саме послідовних значень згаданої константи і згаданого однобітового сигналу.
- 12A device according to any one of the preceding claims. 8-11, which is different the fact that said spatial processing unit (17, 18), which distributes by matrix with a reduced number of rows and columns of information from the outputs of the above a means for smoothing, namely, successive values of said constant (CO) and said binary signal, includes means (17a) for detecting in said matrix distribution of a site-set of pixels, in which at the same time either the value of the binary signal corresponds to the case of exceeding the threshold the value and meaning of the constant for pixels that are adjacent along in a certain direction, differ in significant quantities, as well as for generation Output signals indicating the location of the site, as well the speed and orientation of the movement in this area, or the value The binary signal corresponds to a case where the threshold is not the value of the mentioned constant for adjacent pixels is not exceeded are different. 12. Пристрій за будь-яким із пп. 8-11, який відрізняється тим, що згаданий блок просторової обробки (17, 18), що здійснює розподілення по матриці зі зменшеною кількістю рядків і стовпців інформації з виходів згаданого засобу для згладжування, а саме - послідовних значень згаданої константи (СО) і згаданого двійкового сигналу, включає в себе засіб (17а) для виявлення у згаданому матричному розподілі ділянки-сукупності пікселів, в якій водночас або значення згаданого двійкового сигналу відповідає випадку перевищення порогового значення і значення згаданої константи для пікселів, які є суміжними вздовж певного напряму, відрізняються на значущу величину, а також для генерування вихідних сигналів, що вказують місцезнаходження згаданої ділянки, а також швидкість і орієнтований напрям переміщення у цій ділянці, або ж значення згаданого двійкового сигналу відповідає випадку, коли порогове значення не перевищено і значення згаданої константи для суміжних пікселів не відрізняються.
- 14Method of monitoring the car's driver's vigilance for Detection of possible driver falling asleep, which includes the following steps:14. Спосіб контролювання пильності водія автомобіля для виявлення можливого засинання водія, який включає такі стадії: (a) generating a video signal that is originally responsible sequential image of the driver's face, and then - continuously, in real time, sequential image of driver's eyes only;(a) генерування відеосигналу, що первісно відповідає послідовним зображенням обличчя водія, а потім - безперервно, в реальному часі, послідовним зображенням тільки очей водія;(b) processing said video signal only with respect to said eyes in order to consistently, in real time, be identified in the image Eye vertical movements of the eyelids, corresponding to the blinking, determine the speed these displacements and detect speed values that do not exceed the threshold the blinking speed, which corresponds practically to the driver's state, intermediate between the state vigilance and a state of falling asleep;and (b) обробка згаданого відеосигналу лише щодо згаданих очей для того, щоб послідовно, в реальному часі, виявляти у згаданому зображенні очей вертикальні переміщення повік, що відповідає морганню, визначати швидкість цих переміщень і виявляти значення швидкості, що не перевищують порогової швидкості моргання, що практично відповідає стану водія, проміжному між станом пильнування й станом засинання;і (c) initiating a warning signal to the driver about danger, just mentioned threshold speed of blinking is exceeded. (c) ініціювання сигналу попередження, що повідомляє водія про небезпеку, тільки-но згадана порогова швидкість моргання перевищується.
Independent claims6
480 paragraphs in 26 sections, as filed
UKRAINE
(19) id (11) 59366 (13) C2
(51) 7 006T7 / 20
MINISTRY OF EDUCATION SCIENCE OF UKRAINE
STATE DEPARTMENT OF INTELLECTUAL PROPERTY
DESCRIPTION
TO THE INVENTORY PATENT
(54) METHOD AND DEVICE OPERATING IN A REAL TIME, FOR DEFINITION, DETERMINATION OF LOCATION, SPEED AND DIRECTION OF MOVEMENTS OF RELATIVELY TRANSFER IN SCENE
(21) 99010431
(22) 22 071997
(24) 15 09 2003
(86) PCT / РР97 / 01354, 22 07 1997
(31) 96/09420
(32) 26 07 1996
(33) RR
(46) 15 09 2003, Bull No. 9, 2003
(72) Pirim Pagrick, RR
(73) HOLDING BUREAU, SH
(56) and 5, 5486430, ZO 01 1996
(57) 1 A method for real-time detecting and determining the location of a site that performs relative movement within a surveillance zone observed by a monitoring system having an output of a digital video signal of this type, including the EQUALITY OF THE RELEVANT Frames, each of which consists of a sequence of rows, each of which consists of a sequence of pixels, as well as the value of the velocity and the orientation of the displacement, which differs from the realization of the original digital from osyhnalomtakoyi sequence
smoothing processing of said output digital video signal using a digitalconstant, whose numerical value maychange independently for each pixel of the output signal,
the introduction into the OP, first, of the frame of the output signal, which was subject to smoothing, and, secondly, the smoothing constant corresponding to the given frame,
time processing, which includes for each pixel-noi position, determining, firstly, the presence and, secondly, the magnitude of the significant change in the level of the pixel signal during the transition to the current frame from the immediately preceding smoothed and inserted into the FP frame, and the generation of two digital signals, the first of which is a binary one-bit signal with two possible values, one of which indicates the presence, and the other indicates a lack of significant change in the transition between two sequential frames, and the value of the mentioned binary change signal WWE is the value mentioned in the RFP-denoyi by its constant decrease yakschozhadanyy signal indicates the presence of significant zmi-
us, or increase - if the mentioned signal indicates the absence of such a change, and reduction orincrease step by step, while the seconddigital signal, that is, the signal characterizingthe magnitude of the change, is a multibundle signal with a certain limited number of bits that quantitatively value the value of this change , and
spatial processing, including the performance of such operations for each frame of the input digital video signal of the distribution of pixel values, write for a part of the frame pixels corresponding to a certain observation moment (the part scanned through said matrix during the second time frame corresponding to the frame length) first, the mentioned binary signal and, secondly, the mentioned digital signal, characterizing the magnitude of the change, according to the matrix, in which the number of lines and columns is insignificant in comparison, respectively, with the number of lines and the ring bits of pixels in a line in a video signal to obtain the characteristics of pixel values, the definition in this two-dimensional matrix form of a region in which said binary signal has a certain specific value, indicating the presence or absence of a significant change, and the mentioned digital signal, characterizing the magnitudechanges, changes (or does not change) on a significant magnitude of the pixels of the matrix, which are summers along a certain direction, starting from the initial pixel, within the same area of the frame and, respectively, at the same moment of observation, and the generation of signals , indicating the presence and location of the site, carrying relative movement, and also indicate the relative speed in the transition from frame to frame, as well as the direction of this movement, if it occurs-Xia, relative to the environment, and these signals are determined for the instantaneous matrix distribution of said two digital signals, a binary signal and a signal characterizing the magnitude of the change
2 A method according to claim 1, characterized in that it also includes the formation, firstly, of the histogram values of the signal distributed in the matrices, and, secondly, the histograms of the angles of inclination of the two coordinates of the IZ variables in the plane of the angle of inclination, the detection of the region, in which is observed in the meaningful manipulations that are processed, in each of the formed
CO
59366
59366
4
WITH
their histograms, for each identified region, the value of the presence or absence of a site that performs relative movement, and, if it is present, the definition of the location, speed and direction of movement
3 A method according to claim 1 or claim 2, characterized in that
that said matrix is a square matrix with the same odd number of rows and columns (21 + 1) having embedded matrices containing 3x3.5x5, 7x7,, (21 + 1) x (2I + 1) elements located
in the center of said square matrix, and these matrices are analyzed to determine the least-stated matrix in which said digital signal changes in a certain orientated direction starting from said center, wherein the value of said binary signal indicates an increase of the threshold value for the given direction u
4. The method of claim 1 or claim 2, wherein said matrix is a hexagonal matrix and contains embedded hexagonal matrices, the dimensions of which are gradually enlarged, located in the center of said hexagonal matrix, wherein the matrices are analyzed to determine the least-stated matrix in which said digital signal is varied in a certain orientated direction
5 A method according to claim 1 or claim 2, characterized in that said matrix is a T-shaped matrix with one row and one column, with embedded matrices containing 3x3 pixels, 5x5 pixels, 7x7 pixels,, (2I + 1) x (2I + 1) pixels, are analyzed with the one-line and one column for determining the smallest matrix in which the signal changes in a certain orientated direction, namely, the foam having the greatest angle of inclination and constantquantitative characterization
6 A method according to any one of the preceding claims, characterized in that said constant has a background 2<sup>R</sup>, where p is a number that does not exceed 16, hence -which can be expressed by no more than four-bombs, and this constant is reduced or collapsed by subtracting or adding odnitsa from / to p
7 A method according to claim 6, characterized in that, with the help of the Mall algorithm, analyzes the successive reduced-numbered, full frames, and extracts the largest of these regions, which gives the characteristics of the movement, velocity and direction, which correspond to the value of p
A device for real-time real-time detection and determination of the location of a site that performs relative movement within a surveillance zone observed by a surveillance system having an output of a digital video signal of this type comprising a sequence of respective frames , each of which consists of a sequence of rows, each of which consists of a sequence of pixels, as well as a value of the velocity and the orientation of the displacement, using the method described above, and this device is pr ymaye on svoyemuvhodi said output of the video-nyayetsya the segment that includes
a means (15) for smoothing said output digital video signal using a digital constant (CO), the numerical value of which may
vary independently for each pixel of the given output signal,
a means (16) for introducing into an RF, first, a frame of said output signal (I_I), which has undergone a smoothing, and, secondly, a smoothing constant (SI), which corresponds to a given frame,
time processing unit (15) for analysis, for each pixel position in the frame of the video video signal, the change with the time of the values of the pixel signal for the same pixel position in the transition to the current frame from the front immediately smoothed and inserted into the RF frame of the digital signal, and this The block includes an RF, made with the capability of storing, storing and displaying information about the previous pre-smoothed frame, a comparison tool (15a), designed to determine whether or not exceeding the absolute value The difference between the pixel signal of the current pixel and the magnitude that characterizes the pixel signal in this same pixel position in the preceding frame stored in said PZ is a threshold value by generating a binary or single-bit signal (PIR) with two values ,
a spatial processing unit (11), the input of which refers to said successive binary digital signal and a digital signal characterizing the magnitude of the changes generated by the time processing unit for the pixels of a certain frame, said spatial processing unit comprising a means that characterizes the value of the pixels that carries out the distribution of only the binary signal and the signal characterizing the magnitude of the change that corresponds to the exact moment of time, that is, one frame, visatrix (21), in which the number of lines and the number of sto peaches are smaller, LESS than VI BACKWARDS, RANGE OF ROWS An amount of pixels per row in the frame of said digital video signal, and the frame is scanned through the aforementioned matrix during the time interval that corresponds to the frame length, the localization path, and the definition of the Dipyan-set of pixels in the media of said matrix, in which at a certain time moment, said binary signal has a certain specific value, and a means for determining in the given matrix of a site-set of pixels, in which at a particular moment of time, a digital signal that characterizes the value of CHANGE, CHANGES for adjacent pixels to a significant value, as well as a tool that in response to the instructions from the two previous means generates signals that determine the area-set of pixels, and, therefore, indicate the presence and the location of the site, carrying relative displacement within the zone of observation creeping, and te-nah on the relative speed at transition from cadres to the frame and the oriented direction of displacement eta-
5
of this site, when it really moves relative to the environment
9 device according to claim 8, characterized in that said spatial processing unit includes in the se-be the cascade of the first means of delay (g), and each of these means causes a delay, the duration of which is equal to the time interval between the start of two successive lines, and the cascade of other means delay (b), the lines of each line, each of these means causes a delay, the duration of which is equal to the time interval between two consecutive pixels in a row, while the outputs of all other means of delay (b) and the input of the cascade of other means of delay (b) skin pevnyykonkretnyy th line in time for the said product (17a) to detect simultaneously issued znachennyazhadanoho binary signal and said second signal digitally characterizing the change
10. An apparatus according to claim 8 or 9, further comprising a means (24-29) for forming histograms according to the values of said space spatial processing unit and histograms at the angles of inclination of the two coordinate axes with a change in the plane the angle of inclination, as well as a means for detecting in each histogram of the region, in which there is a significant change in the processed values, so that this region is determined by the output of this medium, and the definition of all the histograms of the output signals that identify the area that performs the relative displacement Within the surveillance zone, if such movement occurs, and determine the location of this site, as well as the speed and orientation of the targeted movement, if the site is moved relative to its identity
The device according to claim 8, 9 or 10, characterized in that said smoothing means (15c, 156) have an input for accepting said digital video signal 3 (RI), and determines for each pixel of each frame the said video signal is a smoothed signal (I_O) , which is characterized by the fact that changes over time of the incoming digital video signal are reduced by using the threshold signal received at another input, and the constant (CO) correlated with each pixel position of the frame, the value of which varies sequentially in a way that smoothes The character remains unchanged, although it does make it weaker, the nature of the changes in the input digital video signal, and said means for smoothing the function-in conjunction with said RF (18), which serves for receiving, storing and displaying the values of the smoothed signal and said constants that are in it,
59366 6
pa and its smoothed value, and the said space spatial processing unit SHALL SIZE DISTRIBUTION FROM A MATRIX WITH LOWER LOADS and columns of information from the outputs of said means for smoothing, namely the sequential values of said constants and the said single-bit signal
12 A device according to any one of claims 8-11, characterized in that said block of spatial objects (17, 18), which performs matrix distribution by a reduced number of rows and columns of information from the outputs of said means for smoothing , and the same successive values of said constant (CO) and said binary signal includes all means (17a) for detecting in said mathematical distribution of a site-set of pixels, whither at the same time, or the value of said binary signal corresponds to the case of exceeding the threshold value and the value this constant of the dyapixels, which are adjacent along a certain direction, differ in a significant value, as well as for generating output signals indicating the location of the site, as well as the velocity and orientation of the direction of movement in this di-valley,
13 A device that helps when driving a land vehicle on the road or in the control of a passenger vehicle in the vicinity of an airport mountaineering runway (LR), characterized in that it comprises a device according to claim 10, 11 or 12, as well as a means for displaying the right and left side of the road or the sides of the runways, VV<sub>d</sub> accordingly, and a means for orienting at least one axis of coordinates at a variable angle of inclination in such a way that it remains practically orthogonal relative to the corresponding curve or side (position Ro)
14 A method of monitoring an auto-driver's vigilance to detect a driver's falling asleep, which includes the following steps
(a) generating a video signal that initially responds sequentially to the face of the driver, and then - continuously, in real time, sequentially representing only the driver's eyes,
(b) processing said video signal only in respect of said eyes in order to consistently, in real time, detect in the said image an eye vertical movements of the eyelid corresponding to the morphology, to determine the velocity of these displacements, and to detect velocity values not exceeding the threshold speed of blinking that practically tells the driver's condition, intermediate between the state of dust and the state of falling asleep, and
(c) initiating a warning signal informing the driver of danger, the above-mentioned threshold speed of blinking is exceeded
The object of the present invention is to provide a method for carrying out a relative movement within the zone of the οπο-ί device for detecting and locating a surveillance site, and for determining the speed and orientation-
7
the direction of this relative displacement, in real time
Relative movement means the displacement of the said site (which may be an "object" in the broader meaning of the word, including the alive or part of a living being, for example, an arm) in a virtually immovable environment, and more or less inland real estate of the said site (or " object ") in the moving object, at least in part
The invention relates to the processing of an output digital video signal from a surveillance system consisting of an optical input system, or an objective designed to capture an imaging zone of the observation, and an optoelectronic transformer system, or an image signal generator (sensor) made from the possible transformation of said image obtained by it into an output digital signal
In general, the surveillance system may have been a video camera or a kinematic that observes the observation area monitored, in which case said output digital signal will be a digital video signal obtained at the output of the camera with a cipher output, or on the output of the analog-to-digital converter, the input of which is connected to the output of the camera, which outputs an analog video signal
The observation system can also consist of an objective of any optical device (a binocular, a visual refractor, a detector), at least a portion of the output beam from which is processed by means of a photomultiplier, for example, a CCD or a KMON type , with the usual connected electronic system, and the transducer takes an image formed by said part of the beam of light, and by means of said electron system it is converted into an output of a cipher-free video signal
The invention mainly relies on the processing of the output digital video signal of the observation system, in particular - cameras with a digital output, to detect signals that indicate the existence and location of the SALES that are moving relative within the aforementioned surveillance area, as well as the velocity and the direction of displacement, if the said area is really transferred within this observation zone in relation to a practically immobile environment, in real time
The best created system is to detect and localize an object that carries out displacement, and to determine its velocity and direction of movement is the vision of animals or humans, and an example of this is the hunter in the shelter, which locates the movement of the beast (defines its location), and determines the direction and speed of displacement
Previously, technical solutions for the devices for observation were proposed, such as analogues, such as analogues (Jocomo Indiveri and other magazines, "RossesiyiPdv οί MіsgoIііgо" '96, pp. 15-22), and digital (Pierre-François Radei, in the magazine "Rossesiyиpиdd οί Mіggo Meіgо" '96, pp. 23-29), but in the first article the detectors and analogues are considered
59366 8
blocks with complex structures, and in the second description means for the detection of the outline of the Ingot object, the described devices use dual high-speed and capacious ZP to enable robotics in real time, and about the sites or moving objects , only limited information can be obtained
Therefore, it was suggested to remember the first two-dimensional RF frame output signal from a video camera or similar device consisting of consecutive data that characterizes the pixels that correspond to the observation zone observed by the camera with the time of time ίο, and then remember the video signal of the corresponding next frame corresponding to the ZONE of the observation zone at the moment of time Ti, in the second two-dimensional CP. If the object has moved in the cha-ove interval between the times of time ίο and B, then, first, determine the distance 6, to which the object was transferred within the observation zone between the moments of time it ι οο, and, secondly, determine the time interval T = 1i-1o between the start of two consecutive corresponding frames, for the same pixels. When the speed of movement is set when it is necessary to obtain accurate indicators of speed and direction, characterizing the displacement, the system of this type requires a very large total capacity ZP In addition, obtaining indicators of speed and direction is delayed, this information can be is restrained only after the momentum of time t + P, where P is the time required for calculations for the time interval ίο -Ii and the following disadvantages (the need for a large capacity of the RC and the delay in obtaining the required inform ation) restrict the use of a system of this type
Further, in the French patent number 2611063, one of the authors of the invention for which the author of the invention is based on this application (Patrick Pyrim), describes the method and method of processing in real time flow of sequential data, consisting in particular of the output signal from the cemetery, for the implementation of compression of data According to this patent, with the application of a certain classification law for the first sequence, a histogram of the signal levels is formed; in the CR, a representation function of Gauss corresponding to this histogram is introduced, and from it the maximum and minimum are selected The levels, the levels of the next or the second sequence are compared with the mentioned signal, is introduced in the RC of the first consistency with a constant, identical for each peak village, the binary classification signal is generated, which characterizes the following sequence with respect to the classification law, from the aforementioned binary signal, the auxiliary signal representing the duration and position of the range of significant quantities is generated, and, finally, the said auxiliary signal is used to generate a signal that localizes a range of maximum duration, called the dominant range, and these operations are repeated for the regular sequences Ordered Signal This way and the Classification Device provide the possibility of data compression, holding only the characteristics of the sequential data stream processed, which are of interest In particular, this method allows for digital processing
59366
9
a video signal characterizing a video in order to isolate and locate at least one characteristic from at least one di-valley in the image. Thus it becomes possible to classify the brightness of the signal or the color of the signal and characterize and localize the object in the image
U.S. Patent No. 5,884,804,30 discloses the evaluation of movement by a separate value of the horizontal and vertical changes in the image of the plot being observed. For the detection of displacements from right to left or left, or from top to bottom or bottom up, in, opposite, horizontal or vertical on-line , use differential signals, initiallyimplement the KOC operation with horizontale / vertical differential signals and cadmium differential signals, and, secondly, use the ratio of sums
Horizontal / Vertical Signals and Sum of Differential Signal Signals relative to the KXZ Window. For the above-mentioned US Pat. No. 5,884,303, the image densities in two orthogonal directions, horizontal and vertical, are used with an identical difference K, which is repeated in the two orthogonal on- lines, and the mentioned difference K is determined as the function of the speed of movement, determined by the device under the aforementioned US patent specifies the direction of movement for each of the two orthogonal directions through the use of a collection of the computational operations indicated in strokes 12 (at the beginning and at the end) 113 (in the beginning), the difference signals associated with the use of very complex (and therefore complicated in implementation) electronic computing units that execute, in particular , operations of division, multiplication and addition, in addition,
On the contrary, the method of the present invention is carried out with the help of a digital type device that sets the object of the invention and has relatively simple construction and a relatively small storage capacity, thus providing the necessary information quickly with a wide range of results and applications (in accordance with the completed half image for interlacing frames or full images, depending on the particular application)
In the article by Alberto Tomita W. and Rokyya Ishia "Getting the shape of a hand from a sequence of digital tinting images" in the works of the Institute of Electrical Engineering and Electronics Engineers, Vol. 1994, p. 1925-1930, the detection of displacement is realized by the difference between sequential image-
10
we with the subsequent formation of histograms constructed on the form of a person's hands to get it on a digital area of observation. Histogram analysis is based on the brightness scale, which is characteristic of human hands. It does not provide any means for the formation of histograms of planes-coordinates. The sole purpose of the authors of this article is to identify the displacement human hands, for example, when operating with a conventional mouse, and this displacement is detected when data is entered into the computer
Instead, the present invention is not limited to detecting the movement of the hand but provides the ability to detect the relative displacement of any object (in the widest sense of this term) within the observation zone and does not use histograms constructed on the hemispheric levels that characterize the hand, and instead uses the histograms of certain specific digital variables characterizing the displacement (if it occurs), and the histograms of the plane coordinates
The essence of the invention
According to the present invention
- in a known manner, the output of the cryptographic video signal of the surveillance system, which consists of a sequence of frames (corresponding to the half image in the case where the "picture" correspond to two interlaced frames, or the full image - when the "picture" corresponds to one frame), is processed. each of which consists of a certain number of consecutive lines and a certain number of pixels, or elements of the image, in each of the above lines,
- in order to obtain, using a relatively small capacity, the signals that would indicate if there is within the area of observation of the relative moving area, and, if there exists, would determine the location, velocity, and direction the movement of the site, if the end is really moving relative to the environment,
- by generating two digital signals, one of which characterizes a significant change or the absence of a significant change in the pixel signal for pixels located in the same location of successive relevant frames, and the other changes the magnitude of the change, if it has a location, and the metric distribution these two signals for the pixels of a certain section of the frame at one and the same time point
The first object of the invention is to provide a real-time real-time realization of the detection and determination of the location of a site carrying relative displacement within the surveillance zone, which is observed using a system of observation, which has an output of a digital video signal of that type, containing the SEQUENCE OF RELEVANT frames, each of which consists of sequences of rows, each of which consists of a sequence of pixels, as well as determining the speed and orientation of the direction of movement, which differsthat is accomplished starting with this sequence tsyfrovymvideosyhnalom
- smoothing processing of said output digital video signal using a digital constant, whose numerical value may vary independently for each pixel mentioned
11
59366
th output signal
- Introduction to the RF, first, the frame of the said output signal, which has undergone smoothing, and, secondly, the smoothing constant corresponding to the given frame,
- time processing, which includes, for each peak-position, the definition of, first, the presence and, secondly, the magnitude of the significant change in the pixel-level signal when moving to the current frame directly directly from the previous smoothed and introduced into the frame of the FP , and the generation of two digital signals, the first of which is a binary one-bit signal with two possible values, one of which indicates the presence, while the other indicates a lack of significant change in the transition between two sequential frames, and the value of the mentioned binary signal change is the value of the input introduced into the LC constant by decreasing, if the above signal indicates a significant change, or an increase - if the signal is indicated by the absence of such a change, and the reduction or increase is carried out by the coverwhile the second digital signal, that is, the signal characterizing the magnitude of the change, is a multi-bit signal with a certain limited number of bits that quantitatively represent the magnitude of the change, and
- spatial processing, including the execution of such operations for each frame of the incoming digital video signal
- the distribution of pixel values writes for part of the pixels of the frame corresponding to a certain observation moment (the part scanned through the matrix, during the time interval corresponding to the duration of the frame), first, the said binary signal and, secondly, the mentioned digital signal, which characterizes the magnitude ofchanges, in the matrix, WHICH number of rows and columns is insignificant in comparison, correspondingly, with the number of rows and the number of pixels in the line in the video signal, to obtain the characteristics of the pixel values,
- Definition in this two-dimensional matrix background of a region in which said binary signal has a certain specific value, indicating the presence or absence of a significant change, and the mentioned digital signal characterizing the magnitude of the change, changes (or does not change) to a significant magnitude by the pixels of the matrix, which are summers along a certain direction, beginning from the initial pixel, within the same plot of the frame and, accordingly, at the same moment, the observation, and
- the generation of signals indicating the presence and location of the moving objects, and also indicate the relative velocity in the transition from the frame to the frame, as well as the direction of this displacement, if it occurs-Xia, relative to the cluster, and these signals are recognized -making on the instantaneous matrix distribution of these two digital signals, binary signal and signal, characterizing the magnitude of change
In an over-weighting embodiment, the proposed method is different in that it also includes
- formation, firstly, of the histograms of the values of the sin-gnal, distributed on the matrices, and, secondly, of the huto-
12
grams of angles of inclination of two coordinate axes with an angle of a slope, changing in plane,
- detection of an area in which there is a significant change in the processed values in each of theformed histograms,
- for each identified region, determining the presence or absence of a site carrying relative displacement, and, if available, the definition of the location, speed and direction of movement
In particular embodiments
- said matrix with a square matrix with
the same odd number of rows and columns (21 + 1) that has embedded matrices containing 3x3.5x5, 7x7,, (2I + 1) x (2I + 1) elements located
in the center of said square matrix, and these matrices are analyzed to determine the least-stated matrix in which said digital signal changes in a certain orientated direction starting from said center, wherein the value of said binary signal indicates an increase of the threshold value for the given direction u
- said matrix is a hexagonal matrix and contains embedded hexagonal matrices, the dimensions of which are gradually enlarged, located in the center of said hexagonal matrix, and the matrices are analyzed to determine the least-stated matrix in which said digital signal changes in a certain orientated direction,
- the matrix is a G-shaped matrix with one row and one column, with embedded matrices containing 3x3 pixels, 5x5 pixels, 7x7 pixels,,, (2I + 1) x (2I + 1) pixels, analyzed with the same line and one column for definitions the smallest matrix in which the signal changes in a certain orientated direction, namely, the line having the greatest angle of inclination and constantquantitative characterization
In the preferred embodiment, saidconstant has the form 2<sup>R</sup>, where p is a number that does not exceed 16, hence - which can be expressed no more than four bits, and this constant is reduced or increased by subtracting or adding an unit from / to p
In this case, if necessary, by means of the help of the Mall algorithm, analyzes successive decreasing sections of full frames, and chooses the largest of these sections, which gives the characteristics of displacement, velocity and direction corresponding to the value of p
Another object of the invention is to provide a device for real-time realization of detecting and determining the location of a moving area within the surveillance zone observed by a surveillance system having such an output of a digital video signal of this type, which includes the sequence of the corresponding frames, each of which consists of sequence of rows, each of which consists of a sequence of pixels, as well as determination of speed and orientation of movement, using the method described above y prychomutsey device takes its entrance zhadanyyvyhidnyy video signal and wherein vklyu a-ity
59366
14
13
means (15) for smoothing said output digital video signal using a digital constant, the numerical value of which may vary independently for each pixel of said output signal,
- a means for introducing into the RF, firstly, the frame of the above-mentioned output signal, which has undergone smoothing, and, secondly, the smoothing constant, which corresponds to this frame,
- time processing unit for analysis, for each pixel position in the frame of the input video sig-nal, the change over time of the values of the pixel signal for the same pixel position in the transition to the current frame from the front-right front smoothed and entered in the ZP-k-dru the said digital signal, and thisblock includes an RF, made with the capability to receive, store and provide information in the correspondingprevious smoothed frame, a meanscomparison, designed to determine whethernot exceeds the absolute value I am the difference between the pixel signal of the current pixel and the magnitude that characterizes the pixel signal in the same peak-seat position in the previous frame stored in said PL, a certain threshold, by generating a binary or one-bit signal with two values ,
- a spatial processing unit, the input of which is the said successive binary digital signal and a digital signal characterizing the magnitude of the changes generated by the time processing unit for the pixels of a certain frame, said spatial processing unit including a means for characterizing the value pixels, which carries out the allocation of only a binary signal and a signal characterizing the magnitude of the change corresponding to the same moment of time, that is, one frame, according to the matrix (21), in which the number of rows and the number of columns are smaller, , Respectively, the number of rows and the number of pixels per line in the frame of said digital video signal, and the frame is scanned through the matrix during the time interval corresponding to the frame length, the localization means for determining the plot-set of pixels within the limits of said matrix, in which, at a particular point in time, said binary signal has a certain specific value, and a means for determining within the limits of said matrix of a plurality of pixels, in which at that instant a digital signal characterizing the magnitude of the change is changed for adjacent pixels to a significant value, and a kind of device that, in response to instructions from the two previous means, generates signals that define the said site-a set of pixels and, therefore, indicate the presence and location of the site that carries out relative displacement within the zone and
observation, as well as the relative speed in the transition from the frame to the frame and the orientation direction of movement of the area, when it truly moves relative to the environment
In an overwrite embodiment, the spatial processing unit also has an output that appears to be delayed by a digital video signal, which is an incoming digital video signal with a delay equal to the length of the matrix line, less the length of one line, in order to output output signal in synchro-nism with the analysis of the matrix in said block of timeprocessing
In the over-weighting embodiment, the spatial processing unit proposed in accordance with the present invention includes a cascade delay device, each of the casings of the elements of the cascade causes a delay, the duration of which is equal to the time interval between the start of the two successive rows, and the cascade delay means for each line, each of the elements of this cascade causes a delay, the duration of which equals the interval of time between two consecutive pixels in a row, and at the outputs of all registers and in the inputs of the first p histriv each line in the SEW-tion given time to remedy said fordetection simultaneously issued value of said binary signal and said tsyfrovohosyhnalu characterizing the change
In an over-weighting embodiment, the invention provides, in accordance with the present invention, a means for forming histograms at values at the output of said block of spatial processing and histograms per angles of a slope of two coordinate axes with variables in the plane of the inclination angle , as well as a means for detecting each histogram of an area in which there is a significant change in the processed values, so that this area is determined at the output of this medium, anddefinition of all logos of the output signals that identify the area carrying relative non-movement within the surveillance zone, as such movement occurs, and determine the location of this site, as well as the velocity and direction of oriented displacement, if the said site moves relative to the environment and
If it is necessary to detect the movement of an object in a practically immobile environment, then you need to identify the part of the matrix in which the value of a one-bit signal indicates an excess of the limit and the time of the day, the digital signal characterizing the universe of change has significantly different meanings for the neighboring ones pixels of one frame
On the contrary, if it is necessary to detect a floating object in a moving environment, it is necessary to determine the part of the matrix in which the value of a one-bit signal indicates that the limit is not higher than, and at the same time a digital signal characterizing the magnitude of the change, has almost identical values for neighboring pixels of one frame
In an over-weighting embodiment, in said apparatus for detecting a site that carries out relative movement within the zone of surveillance, and to determine the location of this site and the speed and direction of said displacement, implementing the aforementioned method
59366
15
- said anti-aliasing means has an input for receiving said digital video signal 3 (RI) and determines for each successive peak-village the frame of said video signal is a smoothed signal (BO), characterized in that the time difference of the input digital video signal is reduced by the use of a threshold signal -la, accepted at another input, and the constant (CO), correlated with each pixepposition POSITION, the value of which varies consecutively in such a way that smoothing leaves unchanged, although it makes it weaker, the nature the change of the input digital video signal, said means for smoothing operates in conjunction with said RF (16), which serves for receiving, storing, and issuing updated magnitudes of the signal and said constant for each plurality of pixels of the frame,
- there is provided a means for detecting in the said matrix distribution a plurality of aggregate pixels, in which at the same time
or the value of said binary signal corresponds to a case where the threshold value is exceeded and the value of said constant for pixels adjacent along a given direction differ in a significant value, in which case this means generates output signals indicating the location of said areas, as well as speed and orientation of movement in this area, or
the value of said binary signal corresponds to the case where the limit is not exceeded and the value of the mentioned constant for adjacent pixels does not differ, and in this case, this means generates output signals indicating the location of the said site
In embodiments that are given overweight
- the aforementioned smoothing means includes, in conjunction with video SLs or operating SLs, for each pixel of the frame consecutive sequences of the mentioned constant and smoothed digital video signal, a means for calculating for each pixel the absolute difference between the value of the digital video output from the echo camera and the value pre-smoothed digital video signal, a means for comparing the above difference with the threshold value and the generation of a binary signal, one of the two values of which indicates the excess memory term limits and insheznachennya - the lack i'i excess zasibdlya constant updates mentioned that oderzhuyebezposeredno previous value constant izzhadanoho W and reduces it if it otrymuyedviykovyy signal with a value that indicates the re-
16
Extends the limit, or increases it, if it receives a binary signal with a value indicating the absence of excess, and does not reduce or increase if it can lead to, respectively, a negative value or to a value that exceeds a certain threshold, means for updating the smoothed value of a digital video signal that carries an algebraic addition to the previous value of this coaxial signal restrained from said RF, the difference between the digital video signal and the video output of the previous and the previous m of the value of the condensed digital video signal from said RF, to the value of the previous constant obtained from the said RF,
- the mentioned constant has the form 2<sup>R</sup>, where p is a whole number that does not exceed 16, which, consequently, can be expressed by no more than 4 bits, and the reduction or increase of the constant is carried out by way of subtracting or adding one unit with / dp,
- the said matrix distribution tool combines with a delay device that results in sequential delays, the duration of which is equal to the duration of one line of the video signal, the digital video output from the video camera, so as to output the consistent outputs of this video, delayed for one delay period, two delay periods, etc., to the number of delay periods, which the unit is less than the number of rows in the said spatial matrix of distribution, a means for implementing a matrix distribution in successive rows of the given matrix that will accept are, firstly, tsyfrovyyvideosyhnal a camcorder without delay and recollection-tion signal successively delayed from vyhodivzasobu delay and, secondly, the importance zhadanoyikonstanty and said binary signal schostrymuyetsya from said means for zhladzhuvan-tion,
- the matrix distribution tool includes a number of conductors of the digital signal, in particular for each row of the matrix, each of the sequentially connected shift registers, the latter of which causes a delay, the duration of which equals the time interval between two consecutive pixels in the line of the digital video signal , in which the position of the pixel, distributed in the matrix, is determined by a point in the matrix, located below the register of delay, and their number for one line per unit is less than the number of columns of the matrix, and the point that the rasta of Old below nizhnaynyzhchyy register
- said means for detecting the site carrying the movement in the matrix, by simultaneously identifying the value of the binary signal indicating the excess of the limit, and changing the value of the constancy, includes a means for determining, for discrete digital directions, the tilt of the kernel changes the value of the constant, close to the pixel in the center of the mentioned matrix, the formation of the initial point of the mentioned directions, and a means for choosing the most-
17
more steep angle of the mentioned change near the mentioned central starting point and for determining the orientation direction, taking into account the choice criterion for the choice of direction in the case of availability of more than one direction with one and the same maximal angle of inclination of the mentioned change, in which the latter of the aforementioned means outputs signals , reflecting the speed and orientation of the movement of the moving area, with a control signal indicating that said speed and direction signals are sub-data control, together with nachennyam constant
In an embodiment of over-weighting, the device for forming histograms, if present, includes itself
- inputs for reception of a signal, which causes a pogram, and a signal for inspection on the basis for the matrix distribution, and
- a means for the formation of two linear single-dimensional histograms for two plane coordinates and to combine these two linear histograms, with a two-dimensional histogram, corresponding to a segment in which there is a significant change in the input signal, and
- an output that gives a signal characterizing the said site. In addition, in the preferred embodiment, the means for the formation of a spasm includes in itself
- a means for calculating the reference change, the input of which receives a string sequence signal, a signal sequence of columns and a pixel synchronization signal, and the output of which represents the reference change,
- two means for the formation of histograms for two axes, which take two reference signals and form histograms of the axes, and
- means of a plot in which the outputs of two means for the formation of pistograms of the two axes, and the output of which gives a signal containing general information about the inclination of the two axes
In certain applications, this constant can be constructed from a limited sequence number of intervals with a gradually increasing magnitude, which is divided by the absolute value of the difference between the value of the current pixel and directly the previous values of the same pixel after the smoothing, for each pixel position
In one of the preferred options, the dipa definition of the crossover signal exceeds or the absence of exceeding the threshold for each pixel position is the absolute value of the difference between the value of the current input pixel and directly to the previous smoothed value of the same pixel that comes from the RP, compared with the threshold value
In the preferred embodiment, the digital signal characterizing the magnitude of the change, gains in the form of an integer number that is characterized by the tendency to approach the value of the current pixel to the immediately preceding magnitude value of the pixel - for each pixel position
One of the distinctive features of this guideline is that, in order to determine both the binary signal mentioned above and the said digital signal, which characterizes the magnitude of the change, as the significance of the immediately preceding pixel, it uses its smoothed value that the stored-
59366 18
In ZP, in order to make smaller values of the difference between the values of this pixel signal at different times that can be observed in the input digital video signal from the video camera or from any other device with digital output
It is known that the result of the smoothing operation is the gradual replacement of the digital signal with large values of the difference between the values of this signal at different times of the signal with the smaller values of such a difference, therefore, a signal that is easier to quantize, requiring for this smaller number of quantization steps, and so thus requiring a smaller number of bits for a smoothed digital signal that characterizes the value of the VARIATION
Below will be described, with reference to figures, an embodiment of an embodiment of the invention proposed by this invention. The method of this invention is described. An illustration of this example will be provided in the description of this embodiment and should not be considered as limiting the scope of the invention, which is determined by the formula of the invention.
Brief description of figures
1 illustrates in a very general way the system with and! inputs outputs, as well as an input signal for this system
FIG. 2 depicts, in the form of functional blocks, the main components of the device proposed by the present invention, which form a node that performstime and spatial processing
FIG. 3 and FIG. 4 illustrate functional circuits, respectively, of a computing node of time processing and a computing node of spatial matrix distribution, which are an integral part of the device of FIG. 2.
FIG. 5 shows schematically a time processing and spatial processing in the system of the present invention
FIG. 6 shows numerical values, according to the Friman code, of discrete directions starting at the reference center in the matrix of FIG. 4.
FIG. 7 illustrates two small embedded matrix matrices of time distribution
FIG. 8 and FIG. 9 describe two other types of matrices, namely, a hexagonal and Γ-shaped matrix, respectively
FIG. 10 shows schematically a node of the device 2 along the Z-Z foam<sup>1</sup> with an additional node according to an embodiment of the invention, which is advantageous
FIG. 11 shows said additional node in the form of a functional diagram, FIG. 2 and FIG. 11 are along the Z-Z line<sup>1</sup>, indicated by the tri-duct lines of FIG. 2 and FIG. 10
FIG. 12 shows the formation of two single-dimensional histograms for the input signal and, on the basis of these histograms, a two-dimensional histogram of the discharging site
FIG. 13 shows in more detail the block of processing and shaping the histogram, as well as the linear combining unit used with it.
FIG. 14 illustrates a one-dimensional histogram
FIG. 15 and FIG. 16 illustrate the use of a change in the tilt angle of the observation zone
FIG. 17, FIG. 18 and FIG. 19 illustrate other possible. FIG
59366
19th
the methods of using the device of this invention, in FIGS. 17 and 18, in the form of functional circuits, and FIG. 17 illustrates the use of video conferencing, FIG. 18 illustrates the use of control of movement in motorways of different classes, and FIG. 19 illustrates the -application, in which the operator manages a machine with the help of hand movements
FIG. 20 and FIG. 21 schematically represent the use of an invention for detecting a driver's situation
FIG. 22 shows a transformed image of the Mall diagram
First of all, we turn to FIG. 1. In an embodiment, which is preferred, the device 11 proposed by the present invention primarily includes an input 12, which receives a digital video signal 3 from a video camera or a camomile 13 with one or more CCD type CCD sensors having a direct digital output or analogue output converted to analog-to-digital converter in digital output Signal 5, as known from the state of the art, consists of consecutive pairs of interlacing frames, such as TR, and TR), TP<sub>2</sub> and TK'z, each of which consists of consistency of the horizontal rows of sweep, and each row (such as (and, b <sub>2</sub>,, its IT in TRI and I21 in TR<sub>2</sub>)
consists of a sequence of elementary signal-pixels, or elements of the image, RI, which respond to points (such as az, ai<sub>2</sub> and ais for her husband's line and, ah<sub>2</sub>and, a<sub>22</sub> for line N <sub>2</sub>, arii and ai<sub>2</sub> for a number of ή π, ai i and ai_<sub>2</sub> for line I<sub>2</sub>i) the observation zone 13a, which is observed with the help of chamber 13, for this reason, the figure shows 5 (RI) -signal 3, which consists of pixels RI
As is known from the state of the art, 3 (RI) contains alphanumeric synchronization (2T) at the beginning of each frame, and the string synchronization signals (ZB) at the beginning of each line
Thus, one can see that the signal 5 (RI) consists of
- a number of sequences (sequential frames) of the limits of a certain time domain, and
- in each sequence (in each frame) a row of sub sequences (rows, pixels) within certain spatial area
In the time domain, the term "sequential frames" will be used to determine successive frames of the same type (for example, non-odd frames, such as TRi, or even frames, such as TRIs) in frame pairs such as TRi-TRi, which generating successive images in digital video signal 3 (RI), and the term "successive pixels in the same position" will mean sequential pixel values (RIs) located at the same location in the sequential frames of the same type, for example , and in the lineI and in the frame TRi and a-i-i in line І<sub>2 "</sub> in the next corresponding TP frame<sub>2</sub>
In addition, the device 11 has outputs 14 for some of the digital signals generated by them, which are used to indicate the presence of a site or an object (in the general sense, as indicated above) that performs relative displacement, and for determination of its location, as well as the speed and direction of displacement, in the case of its knowledge, in an effectively immovable environment, because
20
a drama composed of the signal ZH, schematically connecting signals indicating the existence of the said site or object and determine the location of the said site or object, the velocity V and the direction of the Eeyan movement, as well as, possibly, the input digital video signal CL, which is delayed for his synchronization with the previous signals, taking into account the time of their processing, and this delayed signal CL is used for visualization of the image perceived by the camera 13 on the monitor or television screen 10, at the same time as it enters inform an entity that characterizes a site carrying relative movement (if such a site exists), in particular, the signal ZH (ν, E) I) that can be used in the assembly and verification node 10a
Referring now to FIG. 2, we consider the consortium of the first part of the device 11 shown in FIG. 1, marked by the dashed lines of the block 11a in FIG. 2.
The main components of node 11a are, firstly, a time processing unit 15 with an AP16 connected to it, and, secondly, a spatial processing unit 17 with a latency unit 18 attached thereto and a clock setup unit 19 and synchronous generator 20 that generates synchronous pulses for a block time processing 15 and set-up sequence 19
Time processing unit 15, which, besides the execution of other functions, performs smoothing of the video signal
- uses a digital video sigal 3 from the eocamera 13, which consists of a sequence of values of pixels RI, and HP pulses generated by the generator of 20 synchronous pulses (initialized by the signal 3) with the frequency of passage of pixels in the frame (in particular, 13.5 MHz), to generate multiple signals, as described below with reference to Fig. 3, and exchanges the values of two of these signals from the CL16, namely, the smoothed value of the I-digit video signal and the value C of the smoothing constant, and after the values B and C of the letter O for the values that are not edayutsya block of 15th ZP16 storage, while the values of schoperedayutsya ZP16 in block 15 and referred to the letter and
- generates an output binary signal PC, indicating the presence or absence of exceeding a certain limit, and a digital signal CO, which determines the updated calculated value of the mentioned constant (value of CO) entering in CL1 b
The units of the calculation ι / or the comparison forming the time processing unit 15 are shown in more detail in FIG. 3, block 15 includes four blocks 15a, 15b, 15s15
The first block 15a of the block 15, which uses the input digital video signal 5, which is made up of a sequence of pixel PI signals, and the value of the I_I signal 2 is shifted to the immediately preceding corresponding frame, before being measured by the block 15 as I_O and temporarily entered into the PL16 (as explained below), calculates the absolute value of AV difference between the input values of RI and I_I for the same pixel position (for example, ai in both the line and the PTT and in line I<sub>2</sub>and UTR<sub>2</sub>)
AB = | RI-I |,
with the frequency given by the sync pulses
59366
22
21
HP generator 20 synchronous pulses
The second block 15b is a verification unit. This block accepts the aforementioned digital AB signal from block 15a and a digital signal characterizing a threshold value of EE that could be constant, but usually depends on the pixel value, it changes in accordance with the change of meaning, so that the gamma is performed. -correction (a known means for changing the EE in such a way that the gamma correction is shown by the triple lines as an optional block 15e) This block compares two digital signals that characterize AV ι 5E in order to determine you are a two-way BR signal, in other words, a signal that has one of the two values - 1 and 0, to indicate, respectively, the presence or absence of an exceedance of the above threshold CE. If the AW is greater than 5E, for SI the given value 1u block 15b indicating that the threshold threshold is exceeded,
In fact, when Pp = 1, this means that the difference between the RI and 1-I, or, in other words, between the input digital video signal and the previous smoothed digital video signal, is very large, and it is necessary to reduce it by reducing the error of the condensation and, conversely, if OR = 0, this constant must be increased
The third block 15c changes the value of the indicated constant to the desired side of the mentioned constant depending on the value. If ER-1, the unit 15c reduces the constant to the unit value and the CO (new value Constant (preceding constant value) -and if P-O, the unit 15s increases constant for the same uniqueness and CO = SI + and
Consequently, the unit 15c receives the above-mentioned two-way signal of the excess of a PC from block 15b at one input and the SI signal, which is the previous value of the constants stored in the CL16, on the other input, and reduces or increases the SI constant at the unit value, which is fed to the input, which becomes COI and is entered into the mentioned PO16, replacing the value of SI
In the embodiment of the over-weight, the mentioned constant, on which the convergence of smoothing depends (as a function of the time necessary to achieve the smoothed value of the input value of the digital video signal), represented by the degree of number 2, namely, as 2<sup>R</sup>, and the integer number is a number reduced or enlarged by a unit value (by 1) in block 15c, in this case -I for p in FIG. If PI-1, the unit 15c subtracts the unit value (1) from the indicator p of constant 2<sup>R</sup>, which is equal to 2<sup>P 1</sup> If OP = 0, the unit 15c gives the unit value (1) to the indicator p of constant 2<sup>R</sup>, which is equal to 2 <sup>p + |</sup>
Choosing the form of a constant as 2<sup>R</sup> provides two advantages, firstly, corresponds to a person's range of human vision, and secondly, simplifies the calculation, and, consequently, the block design 15s
Block 15 should also satisfy the two-way, namely, the CO must remain between the two-boundary values of the CO should not be taken by a negative value (CO> 0) and should not exceed-
Weave the boundary N (OO <N) In the above case, in which SI and CO are expressed in form 2<sup>R</sup>, the upper bound N corresponds to the whole number n, which is the maximum value for p
The upper bound N (or n) can be either sta-loy or variable, if it is a variable, the optional block 15ί (depicted by the dashed lines) changes N (η), for example, according to the user's instruction. The consequence of the increase N is an increase sensitivity to the detection of displacement, whilewith increasing the N sensitivity to the detection of high velocities
Finally, in the fourth block 15c, the first input takes the value of the CO of the new constant, generated in the block 15c, to the second input, the input digital video signal 5 in the form of the pixel value PI, and the third input is the smoothed value of the pre-digital digital video signal, namely I_I зЗП16 This unit performs such a calculation
LO = N + (RH) / CO,
and outputs the output
In fact, the term (RI-I_I) / CO is a change,
which is introduced in the smoothed value of the digital video signal, which takes into account the changed value of the Constant Constant, and it is directly proportional to the algebraic difference between the effective value of the current input pixel RI from the camera 13 and its successive smoothing value I_I, and the inverse-proportional CO
If СО = 2<sup>R</sup>, then
Ι_0 = Ι_Ι + (ΡΙ-Ι_Ι) / 2Ρο,
where used p<sub>0</sub>, the value of p is calculated in the block 15c, which replaces the previous value for the yZP16
Thus, the block of time processing 15 with four blocks of calculation 15a, 15b, 15s, 15s!
- receives 5 (RI) from the video camera 13, synchromic pulse HP to adjust the speed and threshold signals CE and N (or n),
- on the basis of the input signals I_I and SI from the connection ZP16 determines the updated signals I_O and CO, which are shown in the mentioned ZP to replace, respectively, I_Іi ІІІ and express the new calculated values of the condensed digital video signal and the constant, respectively, and
- Issues the aforementioned CO signal and a binary signal exceeding the threshold value of P, calculated for RI, I_I and EE, to the block of spatial obrobi-17 through the block of delay 18
The purpose of the smoothing operation is to normalize the value of the input digital video signal for each pixel, or each point of the image, namely, the changes of each PI, by reducing the differences between these changes and the replacement, for each pixel, the successive real values of RI, which changes Lying at this point of the image, on the smoothed values of I_O, which change less substantially than the value of RI
Thus, for each input RI, the block of time-processing 15 interacting with FP16 issues as a substitute a smoothed value of I_O, which changes to a smaller value using a binary signal 0P indicating the presence or absence of exceeding the threshold value and the signal of the cons The CT tints that are updated and transmitted in the block-spatial processing are shown in FIG. 4
23
Each pixel within the frame of a plot may be identified by two coordinates (as a rule, orthogonal), in the abscissa and ordinate axes, namely x and y, by assigning a pixel of two indices, namely (line number) for the coordinate of the yoga] (pixel number in a line) for the coordinate x The contiguous pixel with indices and that and has a video value (value of the video signal) RI<sub>and</sub>
If we consider the change with time and values of the RI<sub>and</sub> for the subsequent corresponding frames at moments of Io, I2, and,, are separated by a time interval T, which corresponds to the period of the image (which is usually two-frames per second period), which may be 0.04 s, if the video signal frequency of the image is equal to 25 Hz or 0 , 0333s, if the video signal frequency of the image is equal to 33Hz or 50Hz for progressive scan sensors (1 frame per image), a video pixel with indices, and] has sequential values, denoted as RI<sub>andYou</sub>, Ritsn, Ryzhg, Rysz, - in moments of time, i<sub>1;</sub> C, from
In accordance with the present invention, in the time of processing of the work, 15 the values of P, d are replaced by successive smoothing values of bosch, and in particular, of the group, of the batches, of the bundles, of the squirrels
The faces of each of the successive pixels, or point images, with coordinates and,] in the period of time I, that is, P<sub>H</sub>(, its real value is P<sub>and (</sub> is replaced by a flattened value determined by the form-laoy
and + (RID-I-Ii, (and i) / CO, e)
In the preferred embodiment, the CONSTANT CO has the form 2<sup>1)</sup>
Calculations carried out in block 15, in particular, in block 156 for each time interval T, which is AND1 FROM IO, AND 2 FROM IT FROM 1<sub>2</sub> etc., ensure the convergence of the value to the value of RIS with speed, which depends on the constant, which varies in space (and therefore depends on that and)) and in time (and therefore depends on I), and which can be written as00, , (
In all cases, the convergence of I_O, d is more freely, the copy of SOUI is greater. If 00 ,, (= 1, the conjugation does not occur
Obviously, in the above formulas, you can use the decoding coordinates in and x for the number of indices and those that indicate the ordinal number of the line and the number of the pixel in the line
Block 15a calculates AV- | RI-BI, where the indices and,] are implicit, where the AB characterizes the instantaneous difference between the digital video signal 3 (RI) with the smoothed signal I_I, the number of each three, and, I
One of the features of the device of the invention is the implementation of a clock processing unit 15 of the operation of smoothing the values of the pixels, for the value, for each pixel, the successive values of smoothing, but also the constant and the binary signal, indicating the presence or absence of exceeding the threshold value absolute value of the difference between the value of the pixel and the magnitude of the pixel for the two consecutive corresponding frames, for the distribution of the forward matrix of numerical values, while the same time moment as the mentioned constant , the so-called binary signal, the part of the pixels of the bounded part of the frame, with this part of the frame being scanned, in order to determine, based on
59366 24
the local changes of these two values at the current time, the location of the PARAMETERS that are active relative displacement, and determine the SPEED and the direction of effective displacement, depending on the distribution. This is done by the spatial processing university described below with reference to FIG. 4
The convergence of the smoothing is carried out by the computational blocks 15b and 15c, which determine the change of the new input CO constant (in fact, it is), so as to accelerate convergence. This is achieved by comparing AB (actually ABC) with the generic value of EE, which may be constant, or, The preferred option is variable, and in case it is variable, it may lie behind the pixel value, so as to make a ga-correction if the pixel value increases, this boundary increases, and vice versa, the result of the comparison of AB with the threshold value 5E ( actually, this is EO "(if the above value varies depending on the pixel value, then this change is carried out in the optional computing block 15e) is generated in block 15b as a binary PIC with two possible values 1 and 0 as described above
The binary PC signal enters the unit 15c, the change of the value of the constant. Thus, the computing unit 15c also accepts the value of the inputconstantiality of SI from CL16 and updates it using a new value of the CO constant, which (together with other values) is entered in P 16, where it replaces the former value of SI, in fact, as the value of SI and CO used SI, - and CO "for two consecutive moments of time, for example, Ιο and itsdistributed time interval T between the two successive corresponding frames (either paired or unpaired)
Block 15c, which receives the PC and SI, adds or disposes of a unit value (CI) to / from the value of the consistency of SI or unit to / from the value of p when the SU is expressed in form 2<sup>R</sup>, depending on whether the quad binary PIP signal indicates an excess of the threshold value (Pp = 1) or the absence of this excess (P = 0)
If the threshold value is exceeded, the value of this constant is considered to be too large and reduce it, and vice versa
Block 15c also ensures that a new value of the CO constant, calculated on the basis of SI, by the addition or subtraction of the unit, remained between O (the CO is not negative) and the threshold N (CO is not more than N). If these two conditions do not suffer freesia, block 15s does not change the value of SI (which in practice will lie in the range from 0 to N, with the inclusion of both boundaries), and then CO = SI
The upper bound N (or n) may be a stable or variable, if it is variable, it is unnecessary to exceed the limit value of IC (or Bird), the change (if it is implemented) is provided by the user-managed block 151
Alternatively, N or η may be RI-dependent (with the value of N, η, and RID adding three indices and,, I) to provide a regulated change of 1.0 (which is calculated in block 156), depending on the value of RI, what can be
25
59366
expressed as "N ^ or ps is a function of RI, / ', in addition, the definition of IRD or -η<sub>υ</sub>ι) = ί (ΡΙ<sub>υ</sub>ι) is carried out in the computing unit, which replaces the present block 15ί, in which the value of PI comes from the eocamera 13, in addition to N
In one expedient embodiment of the constation C (in fact, each C<sub>and</sub>i) must satisfy-such a condition - and the numerical value should be crt twin, more specifically - to be a power dvvy regardless of the values and,] and I, in this particular case C, d = 2<sup>P (iip</sup>= - where p is a small integer number that depends on and,] and ia and may be expressed by a small number of bits. Application of such a condition provides the additional advantages mentioned above
- the law of convergence of smoothing analogous to the physiological laws of human sight,
- the electronic structure of blocks 15s ι 15ά3 is simplified, in particular, in block 15gi, intended for determination of changes in the meaning of the mentioned constant, calculations for each pair and, and by the formulaBO = BI + (RI-BI) / CO simplified, if the CO expressed in the form 2<sup>R</sup> (where p is a small integer), and the threshold value η for p is a small integer, expressed as a limited number of bits
In all cases, a new smoothed value of the input digital video signal C is introduced in the SR16, where it replaces BI (for each pair of indexes)
Referring to FIGS. 2 and FIG. 3, it can be seen that the timeframe 15, which includes blocks 15a, 15b, 15c, and 15c, may also be seen, and possibly a block 15e and / or a block 15y and interacts with the CL16, defines and produces outputs, as explained above, such values (for each of the three indices I,], O
- firstly, the updated smoothed value of BO, which is transmitted to FP16 to replace the previous squeezed value BI,
- Secondly, two digital signals, namely
- a binary signal of the PC, indicating the presence or absence of exceeding the given threshold value (S-1 in excess, S-O in the absence of excess) in the change of the absolute value of the input pixel signal from the video camera, compared with the previous smoothed pixel signal for the same points, and
- a digital signal characterizing the magnitude of the change, which consists of the value of updatedconstancy of CO,
and these digital signals BR and CO are received through the block of delay 18 by the unit of spatial processing 17, which will be described below, and the CO signal is also supplied in CL16, where the value of СО is replaced by the previous SI value for the same peak village
Thus, as can be seen, the capacity CL16 to store, firstly, the successive values of the smoothed pixel signal, and secondly, the said constant, if the number of pixels in the frame is set to P (hence, the number of pixels in the full image is 2P), will be at least- nx 2P (e + I) bits, where e ι is the bitness (that is, the number of bits) of the pixel signal and the mentioned constant, respectively. In practice, the capacity of the RC should not be much larger, its capacity should not exceed 2P (e + I) bits per number of bits
26
necessary for normal operation, in particular, for addressing and access to the smoothed pixel-like signals and the signals of the mentioned constants, depending on the indexes and] If each video image consists of a single frame of progressive scatter, it is enough to use P (e + i) bppv instead 2P (e + 1) bits
Output signals OR<sub>M</sub> and CO<sub>0</sub> from the time block unit 15 corresponding to the time t, are analyzed and used at the site of the spatial processing shown in FIG. 4, wherein the association shown in FIG. 3 and FIG. 4 is shown in FIG. 2.
In fact, the time processing unit 15 processes the frame signals, whereas the spatial processing unit 17 shown in FIG. 4 handles frame sequences and pixels
FIG. 5 schematically shows the time processing of the successive matching frames of the TP<sub>Kommersant</sub> TP<sub>2</sub>, TP<sub>2</sub>, arranged one above the other, and spatial processing performed on the first of these frames, i.e.TR In this frame, the Cartesian coordinates of the xmax and pixels of PI with coordinates y, x, other layers -indexes, and, at the time B sequential pixels with the same indices z in three frames TRi, TP<sub>2</sub>, TP<sub>2</sub> the indexed QB, QB, and UB, respectively, and they have the values P, μ, RI, and RIC, respectively. The plane in FIG. 5 corresponds to the spatial mapping of the frame, while the frame of the frame corresponds to the time processing (with the passage of time I)
The spatial processing unit 17 to which the connecting delay unit 18 (also shown in FIG. 4) interacts with the control unit 19, the operation of which is controlled by a clock generator 20 transmitting the HP clockwise for each of the subsequent pixel signals (see FIG. 2, depicted compiled scheme)
Output signals BRZ and CO<sub>H</sub> from the time block unit 15 are distributed by the block 17 in the matrix 21 of the reduced size, the number of rows and columns in the number is much smaller than the number of rows B and the number of pixels M in one line, respectively, for P<sub>υ</sub> and CO, at the given moment of time. In particular, the matrix can contain 21 + 1 lines along the x-axis and 2t + 1 rows along the x-axis (in Cartesian coordinates), where I and t are small integers of the V variant, which is given advantage, 11 tons are chosen as power of two, for example, I-2<sup>and</sup> and t = 2<sup>ı</sup>, where a and b are integers (for example, approximately from 2 to 5) To simplify the figures and the description-su, as an example, we will accept t-i (although they maybe different) and t = I = 2 -8, in this case matrix 21 should have 2x8 + 1-17 lines of 117 columns
FIG. 4 shows several of the 17 rows Υ<sub>ο</sub>, Yi,, Y-ib, Yyβ i are several of the 17 columns Ho, X-i, Xn, Hie, which form the matrix of block 17
The problem is to divide the matrix 21 with the number of lines 1 + 1, in particular with 17 lines, and with the number of columns t + I, in particular, with 17 columns, the input streams IPv and SOS, in other words, binary BR signals exceed the threshold value digital signals characterizing the magnitude of the changes corresponding to the CO constant, which are transmitted from the time processing unit 15 and are distributed in a larger matrix, matrix of the frame with B (in particular, 312.5) lines and M pixels in each row, in particular from 250 to 800, depending on the TV standard used
27
In order to distinguish between these two matrices, i.e., the video signal matrix I_hM and the Icht block of block 17, indicated by position 21, we will use the indices and] for the two coordinates of the first matrix (which will be visible only when the digital video signal is output to a television screen or monitor torus) and indices X ιΥ for the two coordinates of the second matrix (shown in FIG. 4), at each instant of time, the pixel with the instantaneous value of RI, is charac- terized, at the input of the spatial processing unit 17, the two-digit POC, ^ and SUI matrices ІххM of these two signals is forwarded By scanning through a smaller matrix 21 (2I + 1) x (2t + 1), as explained with reference to FIG. 4, the matrix 21 materializes (21 + 1) x (2t + 1) pixels belonging to the same frame
V matrix 21 each pixel is determined by the line number from 0 to 16 inclusive, the lines Υο-Y-YE, and the serial number of the column - from 0 to 16 inclusive, the column Ho-Hii in the case copyI-t-8, hence 2i + 1-2t + 1-17 In this case, matrix 21 will be a plane representation of 17x17 = 289 pixels, while the video signal matrix will include several tens or hundreds of thousands of pixels, or even more
In FIG. 4, long horizontal rectangular polygons Y0-Y16 are used (of which only four are shown, as well as YO, Y Y Y y Y Yi) <sup>1</sup> Ho-Xh vertical foam (of which only four are shown, namely Ho, Xi X |<sub>5</sub> and Hyu) to show this matrix 21 (block 17) with the image's green or pixels that have the indiums defined at the intersection of the ordinate line and the abscissa column, and the number of which is set to 17x17. For example, the piqueal position of the cross-section of the intersection column 8 and line 8, as shown in Fig. 4 with the letter e, this piquel position is the center of the matrix 21
In order to carry out a sequential simple distribution of the sections of the matrix I_hM along the matrix 21 (21 + 1) x (2t + 1), the block 17 is connected to the block of the tray 18, which, firstly, receives signals Pi and SO (from indexes of these) and, secondly, the input pixel signal 3, or RI (also with its indices), as well as the signal from the generator 20 synchronization and synchronization signals 5I_ line synchronization and signals of the AF of the stag-frontal synchronization (FIG. 2, FIG. 4)
As shown in FIG. 1, signal 3 (RI) contains not only signals indicating pixel values such as ai and ai <sub>2</sub>, which form sequences in time (in the sequential frames) and spatial EQUILIBRIUM (peaked in the lines of each frame), but also the signal synchronization ZT, ZI_, on the basis of which the oscillator pulse generator generates a clock signal, for example, with a frequency of 13.5 MHz , that is, one synchromic pulse per pixel of the video frame, as well as extinguishing signals VI_, which translate the unit 19 into a non-working mode at time intervals corresponding to the said synchronization signals
In response to said signals from HP and VI from the synchronization clock generator (FIG. 2), the control unit 19 outputs a 5I_ line synchronization signal with a frequency that is equal to the ratio of the 13.5 MHz frequency to the number of columns in the frame (for example, 400) in the block delay 18 , together with the signal of the AF frame synchronization, the frequency of which is equal to the ratio of the above ratio of 13.5 / 400 MHz to several-
59366 28
The lines in the video, for example, 312.5, and the signal of HP
Block 18 uses the signals ZI_ and ZS and the sinchronous pulses of HP for controlling the split spatial distribution of the matrix 21
Consequently, the successive strands of the YO-Wei restrict the sig nali of Pp and CO
- without delay (line Υο),
- with a delay of one interval of time TP, which is equal to the length of the frame line (line Υι),
- delayed by 2TR (line Υ<sub>2</sub>),
- with a delay of 16TR (line Υι<sub>6</sub>)
Serial delays with a duration of a series of frames, namely TP, are introduced in a cascade with 16 delay circuits n, r<sub>2</sub>,, Pe, serving rows,
respectively, Ye, Y<sub>2</sub> Υιβ, and the line Υο is served directly by the signals of Py and CO, without any delay after receipt from the block15
All circuits Γ |, Γ<sub>2</sub> , G16 can be represented by a line delay of 16 outputs, and the delay introduced by any of its elements between two successive outputs is constant and is equal to TP
The control unit 19 controls the alternating, for successive sequences of sequential frames, scanning the frame matrix of size I_hM through a matrix size (21 + 1) (2t + 1), using for this a signal of the offset synchronization, using the following procedure
Circular displacement of pixels belonging to a certain line of matrix of the frame, in the matrix 17x17, for example, the pixels Xp-Hie line Υο, is carried out by a cascade of 16 shift registers si on each of the 17 strings Υ0-Υ16 (in general 16x17 = 272 offset registers) located in each the line between the two neighboring pixel positions, for example, the reciver άοι is located between the positions of Rio and Rioi, the re-god is between the positions of RIOi and RIOiI, etc. Each registry delays the TOR, which is equal to the inter-shaft of time between two adjacent pixels of a single row, using the signals of the Ar possible synchronization
Note that due to the fact that in the frame TR (Fig. 1) for 3 (RI) and for PySi, the block 18 is shifted to the TP (full line length) one after the other, in the frame TP (FIG. 1) as well as due to the fact that the unit 18 distributes them with a gradual accumulation of TP delays on the lines Υο, Υι, Υπ, and the curves at each particular moment of time reflect the signals of PF and CO for its rows, and<sub>2</sub>,, Il, that one
it is the section of the frame
Similarly, for a particular line such as it, the signals ai and ai, and 2, of successive pixels come with a shift in the TK, and the offset registers are equal to the delay, which is also TK, the result of this is that the signals P and CO for pixels Certain of the rows Υ0-Y16 matrix 21, available in this series, are synchronized, or, in other terms, they correspond to the same section of the frame
Thus, for both rows and pixels, these lines belonging to a certain section of the frame require a purely spatial processing, since the matrix 21 contains in its 17x17 pickle positions the value of Pp and CO for 17 pixels of the same of 17 lines the matrix of the same digital video signal 5 (RI), although the pixels mentioned, such as ai-i,
29
ZO
proceed sequentially, row by row and pixel by bits of sequential lines (FIG. 1), in block 18, along with the corresponding signals of PF and CO
The signals corresponding to the synchronized SY-gnals of the PC and CO of the matrix 21 are available at the appropriate moment of time on the 16x17 = 272 outputs of the offset registers, as well as in front of these registers in front of 17 lines, namely the registers of 1, 2, 3, and, -doc what comes out 16x17 + 17 = 17x17 outputs for 17x17 pixel positions Roo, Ro, and,,,, Rice 16
Inside of the matrix 21 around i) the center e, which is the coordinates x = 8, y = 8 (that is why the number of rows and the number of columns in the matrix 21 in the preferred variant, are preferred by odd, 21 + 1 and 2t + 1 , respectively), we can consider, especially, the small matrix, which has 3 rows and 3 columns, in which the central one of its 9 elements is a pixel e with coordinates x = 8, y = 8. Let this small matrix M3 have the form
and b c
b e t (M3), dn
where the central element is underlined
This matrix with the number of elements 3x3, which contains 8 positions a, b, o, b, t, d, ή, and around the central element or position e corresponds to eight lines, each of which proceeds from the central position e and goes to one of the remaining eight POSITIONS
For this, the eight directions can be identified by using the Freeman code shown in FIG. 6, and these directions are encoded by the digits 07, starting from the x-axis, with an angle of 45 °. In the Freeman code, eight possible directions, numbered from 0 to 7, can be expressed by a 3-digit number, since 2<sup>3</sup>= 8, for providing 8 possibilities
In considering the above-mentioned small matrix, M3 can, according to Freeman's code, observe eight directions that emerge from the central position
321
4e0
567
as shown in FIG. 6
Returning to matrix 21 in fig. 4 with 17x17 elements of the image or pixels, the following will describe how it is possible to detect a plot that removes the movement relative to a practically stationary object within the surveillance zone, which is followed by a video camera 13 and therefore reflected by a video signal 3, consisting of pixels P ^, ^ , and how it is possible to determine the speed and direct effective movement relative to the practical but still environment
Between two consecutive frames, such as TRi and TC<sub>2</sub> (FIG. 5), pixels of PI "of the signal 3 will be characterized by their changes between the instant B (first frame) and the time point 1<sub>2</sub> (second frame) using two PR signals<sub>H</sub> and SO, (rotated by scanning in matrix 21)
If the value of ER in a certain element of the matrix is equal to unity, this means that there is a significant change in the pixel value in it. Therefore, the area carrying out effective displacement, is detected by that part of the matrix, in each
59366
the point of which is P = 1
In fact, the computing unit 17a is used for the simultaneous analysis of various laid square matrices with centers at a point with dimensions 15x15, 13x13, 11x11, 9x9, 7x7, 5x5, and 3x3 within the matrix 21 with 17x17 positions (elements), a matrix with a size of 3x3 is the matrix M3 discussed above The device determines which zat matrix is the smallest at P = 1 and is oriented along the straight line, which determines the direction of the displacement of the site and the change of СО by +1 to -1. With such a change, the IP should be equal to units in any meaning, if should be satisfied the requirement is carried out in the search for the smallest matrix of the analyzed ones, with the establishment of the mainpin
Further, within this relaying area, in one of the enclosed matrices, for example, in the small matrix M3 with the number of elements 3x3, the device determines the existence of a change in CO for each of the sides from the central position in this direction, from +1 in the corresponding direction to -1 in the opposite direction. For example, if in direction 1 (orientated) there is -1, 0, +1 (in the case of the matrix M3 - the position of d, e, c, respectively), then the movement in this matrix occurs from right to left in (orientated) direction 1 by the code of Freeman (Fig. 6) It is obvious that at the same time in this part pit in the small matrix 0P = 1 The transfer speed is still, copying one of the attached matrixes from 3x3 to 15x15 when the CO is changed from +1 or -1 between the two adjacent positions in a certain direction is more. For example, if we have -1, 0, +1 in Orientation direction 1 (in the case of a matrix measuring 9x9, M9,
Since CO is a power of two and in the preferred embodiment, it is expressed as an indicator of the power of the two, a wide range of speed values can be detected when using just a few bits for a power factor of two, at the same time as detecting a relatively low velocity (which can be selected by increasing the change or adjustment reserve for verification, for example, -2, 0, +2 at positions d, e, with the matrix M3, that is, 3x3, would indicate a speed two times smaller than the speed corresponding to -1, 0, + 1 for the same position of the matrix M3)
To remove uncertainties it is also necessary to provide for such two checks
- when performing the first check, choose the largest change, that is, the largest constant, if observe the change of CO for several on-lines in one of the embedded matrices, for example, in a small matrix M3 containing 3x3 elements,
- when performing a second test, one of two (or more) lines is randomly selected, along which the CO varies equally, for example, by choosing the smallest value of the Freeman code, such a case is usually observed when the actual the displacement direction is approximately in the middle between the two sub-species represented by the Freeman codes for directions, for example, between directions 1 and 2, corresponding to a direction that can be designated as 1.5
59366
32
31
(FIG. 6), forming an angle of about 67.5 ° relative to the direction of the X axis (direct 0 according to the Friman code)
The definition of the direction and velocity of the above-described definition of the effective movement is described above by the computing unit 17a (FIG. 4), which is connected to block 17 and receives 17x17 outputs from the above matrix 21, both for CO, and for OP ( the scheme shows two outputs of 21, namely, Zo i and z oo) Block 17a handles the values of CO and OP fromsequent embedded matrices and uses these values to determine the direction (by the Freeman code) and the velocity of movement (depending on which of the embedded matrices is elected), possibly with the use of the above mentioned ne evirok
Scanning a full frame of a digital video signal through matrix 21 occurs
- first in the group of the first 17 lines of the frame, that is, from the 1st to the 17th, from left to right, taking into account relative movement, as shown for the frame TR<sub>2</sub> on fig. 5, i.e. from the TMi site in the extreme left position, then the TM<sub>2</sub> with a displacement of 1 column on TMi, to TMM to (where M stands for the number of pixels in the frame line) in the extreme right position,
- then, in a similar way, from left to right, a column by column, in rows of a frame from the 2nd to the 18th, etc., every time descending down to one-third, from the 3rd row to the 19th, etc., to the last group at the bottom of the frame, namely the strings from (E-16) to the B-th (where B stands for the number of rows in the frame)
Referring to FIG. 2 and FIG. 4, one can see such signals transmitted by blocks 17, 18 and 19, that is, a spatial processing unit
- signal V, characterizing the transfer rate, determined on the basis of the magnitude of the maximum change in the CO in the identified area, the value of this signal can be, for example, expressed in accordance with a scale of 8 integers, from 0 to 7, if the speed is expressed in the form of the indicator ste-pen of two, respectively, the signal will be 3-bit,
- the signal JI, which characterizes the direction of movement, is determined in the direction of the said maximal change, and the value of this signal can also be expressed by one of eight values from 0 to 7 in the Freeman code and therefore will be 3-bit,
- UB signal, which indicates the probability of the result in terms of speed and (orientation) direction, which allows us to distinguish between all the output signals ν = 0 and □! = □ and the case of lack of outgoing signals due to some problem circumstances, this signal may be magic or 1 (probable out-of-day signal), or 0 (absence of output signal), therefore, for UB only one BGT is required,
- in the preferred embodiment, the signalconstants of CO, for example, 3-bit,
(the mentioned 3 or 4 signals V, II, UB and possibly CO are issued to the output by the unit of spatial processing 17 and associated electronic media),
- the delayed video signal CL, consisting of the input video signal 3, was subjected to delay in the delay block 18 at a time interval, the duration of which is equal to sixteen successiveTr lines, and, therefore, at a time interval, the duration
which corresponds to the duration of the signal distribution 5 in the matrix 21 of 17x17, so that the received digital video signal is synchronized with
- display in matrix 21, and the content of this signal can be directly visualized on the TV or monitor screen,
- all three signals issued for the output of the block19, namely the synchronous pulses of the HP, the signal of the ZB of the line synchronization and the signal of the APS of the column synchronization
Instead of enclosed rectangular matrices (FIG. 4i and FIG. 7), embedded hexagonal matrices (FIG. 8) or G-shaped (FIG. 9) may be used.
In the case shown in FIG. 8, all embedded matrices (of which only the central matrices MR1 and МР2 are shown) have a center at the point of the MRO, which corresponds to the central point (in which the binary signal is equal to 0) of the matrices M3, M9 in Fig. 7 Preferred The system of hexagonal matrices consists in the fact that this system makes it possible, firstly, to use the so-called Ha, Wa, and Spotted Coordinates, and the ΠΟ-SECOND, of the triangles with equal sides for calculating the isotropic SPEED
The matrix of FIG. 9 consists of one row (B) and one column (C), which begin from the central position of the MP, in which the signal P = 1 and the SNG signal increases or decreases by a single value if the displacement occurs
Thus, it is determined whether (relative) movement occurs
- у напрямі координати х сигнал СО є однако-вим у всіх позиціях (комірках) стовпця Си, а двій-ковий сигнал ϋΡ=1 у всіх позиціях рядка Би, відпочаткового положення МРи, із значенням СОи доположення, в якому СО дорівнює СОи +1 або -1включно,
- in the direction of coordinate in the CO signal is nevertheless in all positions (cells) of the line Ii, and the binary signal SY = 1 in all positions of the column Ci, from the starting position of the MR with the value of СОi to the state in which СО is equal to СОi +1 or -1 incl.
- or, finally, at an angle to x and b of the binary sine-pin P = 1, and CO is equal to SOi. In the positions (collar) B and in the positions (cells) of the Ci, the angular slope is determined perpendicular to the line passing through two positions in which the COI signal changes to a single value, and the signal PY is always equal to 1
FIG. 9 illustrates a case where Pp = 1, and the value of RTI varies by a single value in two specific positions of Biz and C ^, thus defining the corresponding angle of inclination P<sub>p</sub>
In all cases, the speed of movement depends on the position in which the value of CO varies in unit value
If the value of CO is changed to a single magnitude only in B or only in C, it corresponds to the value from this position, in which the value of CO changed
If the value of CO is changed to a single magnitude in a certain position in B<sub>and</sub> and in a definite position in YES, then the velocity will be proportional to the distance from the IM to the (the point of intersecting the foam, perpendicular to the CB, passing through the MRI)
System described above with links to
59366
33
1 to 9, in the embodiment of the present invention, is complemented by an additional system which will be described below with impurities in FIGS. 11 to 16 by generating a general system 22 shown in FIG. 10. FIG. 10 shows, first of all, a knot in FIG. 2, which produces signals V, VI, VI, C, and CL, and the storage signal P (signals HP, SI, AC), which are directed by the node 11 (receiving the input digital video signal 8) into node 22a (which produces a composite output signal ZH )
In FIG. 2, FIG. 10 and FIG. 11, the Z-Zi connection line is shown between the nodes Na and 22. The outputs of the node 11 are connected to the inputs of the node 22a for transmitting the above signals along the Z-Z foam<sub>"</sub>
The output signal of block 22a (hence the general device 22) is a composite ZH signal, which provides the necessary information about the area that performs relative movement within the surveillance zone 13a, which is observed with the help of the camcorder 13
An additional unit 22a connected to the output block 11a is depicted in FIG. 11 in the form of functional blocks. As mentioned above, this figure is connected along the Z-Zi line (shown in the upper part of FIG. 11) of FIG. 2 along the Z- Zi (shown in the lower part of FIG. 2)
The block shown in FIG. 11 is formed, in principle, from a device for the formation and use of histograms, and includes
- bus 23, which transmits several digital signals, described in detail below,
- six blocks 24, 25, 26, 27, 28, 29 forming and processing histograms of such signals delayed digital video signal CL, speed signal V, direction signal (by Freeman code) E1, signal constants, signals of the first axes x (t) and others axis (t), respectively
- six blocks of linear combination ЗО, 31, 32, 33,34, 35, which combine their input signals, which are obtained from the bus 23, each of which is designed to form the control signal ν<sub>"</sub>, ν<sub>2</sub>, ν<sub>3</sub>, ν +, ν<sub>5</sub>, ν<sub>6</sub> for the six blocks 24, 25, 26, 27, 28, 29, respectively, with the block BO connected to the block 24, the block 31 is connected to the block 25, the block 32 is connected with the unit 26, the block 33 is connected to the block 27, block 34 is coupled to block 28, and block 35 is connected to block 29,
- Block 36 of the moving area, which coordinates the outputs from blocks 28 and 29 for the axes (t) and y (t), and
- the change of the line of the beginning of the reference block 37, which takes the orientation signals x (t) o and y (t) for the axes x (t) and y (t), as well as signals HP Pixel synchronization, SN signals of line synchronization and signals From the strictly synchronized (these three signals are grouped together and indicated in Fig. 2, Fig. 4, FIG. 11 and Fig. 11 by the letter P) from block 19 (see Fig. 2 and Fig. 4), and generates signals x (t) and y (t) and transferred to blocks 28 and 29, respectively
Each of the blocks 24, 25, 26 and 27 transmits its output signals CZR, 37, CE), and 3Y to the bus 23, at the same time as each of the blocks 28 and 29 outputs a signal (t)<sub>2</sub>, in (pl) 2 to the two inputs of the plot section 36, which performs the movement, which combines these two signals from blocks 28 and 29, and outputs a composite signal (t) to the bus 23
34
Blocks 24-29 forming and processing histograms that take on their inputs, first, the control signals Ui, ν<sub>2</sub>, ν<sub>3</sub>, ν<sub>4</sub>, ν<sub>3</sub> or ν<sub>3</sub> from the connected to them blocks of linear combination of 30-35, and secondly, signals CL, V, II, CO, x (t) or y (t) and forprocessing, work the same, so the description of work only one of the mentioned blocks, namely, the block 25 of the formation and processing of fast histograms V, taken as an example, is the only difference of each of the remaining similar blocks 24, 26, 27, 28 and 29, the variable that is being processed, while it is important that all input signals for six blocks24-29 are digital signals, which makes it possible to unify the design and operating modes of these six blocks.
FIG. 12 schematically represents the by-pass 38 and 39 histograms, respectively, in x and y (the axes of the decal coordinates of the matrix 21 with the number of elements 17x17 shown in FIG. 4), for velocity values in the moving sections (FIG. 14 shows the column-elements Circuit 38, such as CI<sub>2</sub>), x<sub>m</sub> and y, "represent the x and y coordinates of the double-summation maximums 38 and 39, respectively, while I<sub>and</sub> and i<sub>ı</sub> for all x and y<sub>with</sub> and Ic for the axis in represent the boundaries of the diagonal of significant or those that constitute the integrity of the velocities, with the Ia and Is being the lower boundaries of us, and b and i are the upper boundaries of the significant sections of the histograms
Vertical lines I_<sub>and</sub> and C abscis I<sub>and</sub> and I and horizontal lines I_<sub>with</sub> and I ordinate I.<sub>with</sub> and form a straight-angled, which limits the marked cross-sectional area of 40 significant speeds (for all directions X and y), and near the main area 40, there are several micro-regions of small velocities that can be neglected
Thus, in order to determine the region of the largest changes of a particular parameter represented by a histogram, in this particular case-velocity V, it is only necessary to determine the coordinates of the four boundaries of I<sub>and</sub>, I, i<sub>with</sub> and It and the two maximums Hm and y<sub>m</sub>- Block 25 continuously provides thisinformation regarding the speed V on the micro-bus 23
Similar blocks 24, 26 and 27 provide continuous information in the same way regarding the areas of maximum values for, respectively, CL, SI and CO, on the microscope.
Finally, similar blocks 28 and 29 continuously produce information relating, respectively, to the region of maximum values of x (t) and y (t), and to block 36, which combines this information with the abscissa and the coordinate designated, respectively, x (t)<sub>2</sub> and in (t)<sub>2</sub>, V is a compound signal xy (t), which is transmitted through output of block 36 to bus 23
As a result, the tire 23 transmits information related to the region, which is characterized by the maximum values of CL, V, II, CO of v (t), that is, x (t) and and x (t)<sub>2</sub>, to determine whether there is a moving area within the area of observation observed using the camcorder 13 to determine the location of this site and to determine the SHIELD-COSTI and (oriented) displacement direction
In FIG. 11, the complete output signal from the outline 23 is denoted by ZH. The above components of the signal ZH, in particular V ij, ie, velocity and orientation,
59366
35
The inward direction of movement of the moving area can be analyzed in digital or analog form, can cause a waveform signal ι / or an audio buzzer signal, in particular if the velocity V exceeds the out-of-bound limit value, and can also transmit- fiber optic or relay line (in the case of applications involving the presence of remote objects) to the control unit, such as the unit 10a (FIG. 1), located near the device 11 according to the present invention, or remote from said device
The blocks in FIG. 12, which are shown above the bus 23, are determined by treating the entire frame by processing the entire set of pixels of the frame, outside the global value indicating whether the moving area is present in the monitoring zone, and , in the case of availability, determine its location, and in the case of the actual displacement, determine the speed and orientation of the direction of movement. This section, which performs relative displacement, reveals the area of observation determined by the directions of the XYU, which are not necessarily orthogonal (e.g. Hell, for the case of application shown in FIGS. 15 and 16), the process being implemented using the blocks shown in FIG. 12 under bush 23
Below, with reference to FIG. 12, FIG. 13 and FIG. 14, a detailed description of the construction and operation of the block of forming and processing histograms, such as 25, and a corresponding linear combination unit such as 31
Block 25 (FIG. 13) has a section 25 a forming the g-stop and a section 255 forming the classifier for the histogram, and the two sections work under the control of the software materialized as part of the integrated circuit 25c, which defines the boundaries I<sub>and</sub>, B, i, and histograms (FIG. 11)
Classifier 25b, as well as the classifiers of other blocks 24, 28, 27, 28, 29 forming and processing histograms (for the latter two - with the help of block 36, combining x (t) and y (t)), transmits information from their exits to the bus 23 and, through this bus, on the bilinear combination 31 which, respectively, simultaneously receives information from all classifiers blocks 24, 25, 26, 27, 28, 29 and, depending on this information, issues or does not issue a control signal ν<sub>2</sub> in block 25
By using software 25c, classifier 265 defines different classes (each of which contains the same number of speed values for the case under consideration) that will determine an asset such as 38 or 39 (FIG. 12)
FIG. 14 shows, for the velocity V, successive classes Cu<sub>2</sub> , WITH<sub>p</sub> and C.<sub>p</sub> along the x-axis, as well as their void 38, which are defined in the classifier 255
FIG. 15 and FIG. 16 show the portion occupied by the g-stograms for x (t) and y (t) formed by the blocks 28 and 29 and combined in block 36 to determine the tilt slope
Consider as an example the case of observation
I'm expensive with a digital camcorder.
the outlet, located on the transport
the same means as the device according to the present invention
15 shows two roadside edges, Vd to the left
36
and All to the right of the road K, as well as the angles of inclination of the projection P, relative to x (t), determined by block 28, and projections P<sub>in</sub> relative to y (t), which is determined by the block 29, and the said angles of inclination are measured, for example, from 0 to 7 (according to the rules that differ from the Freeman code)
In order to ensure maximum accuracy with respect to the right-hand side, that is, the maximum sensitivity of the determination, for example, the velocity due to this road, the projection of P must have an angle-point, the value of which will be as close as possible to the optimal angle of inclination of Ro, which is the perpendicular to you , in the case shown in FIG. 15, the same angle of inclination is the angle of inclination. Thus, the maximum value of the speed histogram will be obtained for the slope 5 defined by the block 28 (FIG. 11)
Similar considerations will be applicable to the frontal side V<sub>d</sub> and the slope of the projection P<sub>in</sub>, and, consequently, the part analyzed by block 29
Block 36, which combines two optimal values of the angle of inclination, produces an optimized information for two edges V and B<sub>d</sub>
16 illustrates the application of the invention to determine the optimum pitch of the P-pro-oscillation slope for the successful driving of the ground vehicle Hg, in continental Europe at right-hand traffic (case a) or in the Great Britain at left-hand traffic (case b), and , finally, for the successful management of the aircraft apa-rat ν<sub>3</sub> to ensure proper landing in the runway (runway) (vipa-dock c)
Consequently, for assistance in driving a land vehicle (passenger or freight car) on the road or in the control of a flying machine (an airplane, a space shuttle) when under the flight to the runway of the aerodrome device in this facility additionally equipped with a means for tracking the right and left (W And B<sub>d</sub>, respectively) on the side or right and left sides of the runway and a means for orienting at least one of the axes of coordinates with a variable angle of inclination in such a way that it remains practically orthogonal relative to the corresponding side or side (position Ro)
To this moment, the description of the invention has focused on the means of defining a site that effectively translates into a practically immovable environment by identifying a section for which BR = 1 Dpa so as to determine the area that does not shrink in the moving environment ( for example, a broken vehicle or miss-it's avars on the highway), have to deal with an opposite situation, the copy must be determined by the location of the sites in which BR = 0, in contrast with the environment in which BR = 1 Obviously, in such a di- the velocity velocity is zero, but the concept directly loses value So computations in Block 17a are carried out in a different way
The device of the invention intended solely for the determination of a fixed site can be simplified by releasing blocks for processing speeds and directions, in particular, blocks 25, 26, 31, 32, and the number of outputs of the Block on the inputs of block 17 can be reduced
In the operating unit 10a of the device, in this guideline, a visual-
37
the histograms ι / or the values of the BR signals or the monitors on the monitor screen
The 5P signal, that is, the delayed digital e-signpost, is usually sent to the TV-monitor screen 10 or the mungator to visualize the signal on the screen located at the same or at a remote location, along with the message of relative non-movement, so that you can it was necessary to monitor the nature of the relative movement of the Soul, it is necessary to look at the television screen or the monitor 10 only when it reports the-Xia about the relative movement, for example, with the help of a visual ι / or audio message
It will be effective to facilitate the detection of DILIAN-KIs with relative movement on the screen, using different colors when rendering the delayed digital video signal CL so that each color or hue of color is characterized by the velocity ι / or the direction of motion
Each of the blocks described above with reference to FIGS. 2, FIG. 3, FIG. 4, FIG. 11 and FIG. 13 is formed by an electronic circuit of a known type, in particular, microprocessors that perform calculations and / orcomparison, or use scan signals , ZP, delay blocks, shift register, blocks of formation of linear histograms and assimilation of these histograms on a plane, microspheres
Combining these electronic circuits into separate blocks Na or 22a, or a node 22, consisting of blocks Na and 22a, can be accomplished by using two very small integral circuits or one such scheme in the size of approximately 10x10mm in the case of the use of technologies of 0.7μm, and the input of such a set of two integrals circuits or a single integrated circuit connects-with the digital video output of the camcorder or other surveillance device, and outputs - with one or several devices located in the charge or in a remote location. In one embodiment, the number and the unit 11a is used only for simplified devices, this device, in the preferred version, in the form of a separate integrated circuit, is located between the mentioned digital output and one or more devices located near us or in a remote location
Consider the examples (which are not to be considered as limiting the scope of the invention) to other pre-application applications of the device of the present invention, in particular the device of FIGS. 1 and FIG. 10, in other words, the nodes shown in FIG. 2 and FIG. 11, which interconnecting along the Z-Z substrate<sub>"</sub>
The first additional application, shown in Fig. 17, involves the automatic cropping of a moving human being, for example, during a videoconference. Automatic cropping eliminates the movement of a person moving indoors, thereby enhancing the clarity of the image of the person on which the observer with the help of a videocapher with digital output, and, in the case of a digital digital epoxy seal, simplifies such a seal
FIG. 17 depicts the camcorder 13 of the aforementioned
higher type, which observes the person P, which is
same move in the room Digital video sig-
nal 5 from camcorder is transmitted by cable, op-
59366 38
fiber or radio relay foam on a television screen or a MONEY screen, and also a device 11 according to this invention, the output of this device is connected to a volt 42, which, in response to the signals of the positioning and displacement of human P, which is received from device 11, controls the motor 43 of the said camera 13 to direct the optical axis of the camera on the person and, in particular, on the face of the person, depending on the position and velocity and direction and displacement, possibly by changing the magnitude of the parameters, focal length, ι / or foc eliminating the mammals when moving a person R forward and back. You can also control the movement of one or several searchlights directed at people-well (artist, singer) on the stage,
Another example of using the device of the invention according to the present invention is shown in FIG. 18, which shows a camera 13 or other device surveillance, the output video signal 3 of which enters the device 11 offered in accordance with the present invention. In this case, the camera 13 carries out video control of the highway section the purpose of detecting an unexpected stop of a vehicle, in particular, on the roadside, or after a collision of a car
Thus, in this case, the task is to establish the presence or absence of a real object (vehicle) in a moving environment (other means of transport), ie, inoculation of a site in which BR = 0 in the matrix 21 with the number of 17x17 elements under normal conditions of the chamber 13 watches the traffic flow, which gives the value BR = 1, with the corresponding values of velocity and direction of movement. However, at the pinion of the vehicle in the observation zone, an object with BR-0 occurs
Block 44 which receives the signals ZHi 5P, detects a region in which BR = 0 in ZH, and produces an emergency signal IC, which, firstly, provides a sound ι / or visual signal of the signal to device 45, and secondly , directs the switch 46, which feeds the video 5 (or rather, the 2P video is delayed) to the screen
10 TVs or monitors that allows the detected warning signal to keep track of the highway in the event of a stop or collapse of the vehicle for the use of necessary measures, for example, depending on the reaction of the vehicle to the stopped vehicle.
Using this invention, the controller at its workplace can easily carry out surveillance over a large number of sections of the highway or tracks, each of which has cameras 13 and device 11 provided by the present invention, in which the output signals ZH and ZR from each device
11 are transmitted over a cable, optical fiber, or di-relay line to one control unit, one block 44 is absent, and the observer needs to look at screen 10 only if the emergency signal signals an incident or an accident, one-time incident or multiple accidents
59366
39
It is unlikely that the highway sections will be
In particular, at the input of block 44, a rotating switch (not shown) may be pre-seen, which sequentially and cyclic directs the signals ZH (and CL) from different nodes 13-11 located along the highway, to the end of block 44, which outputs the signal IL
With the help of this system, it is possible to detect not only the stops of vehicles or collisions, which lead to traffic delays, but also reduce the speed of the flow of transport vehicles (in the case of too slow motion), when U = I, and vice versa, a sharp increase the velocity of the observed zone of observation, which indicates the excess of the permissible velocity of the arrow
Finally, FIG. 19 shows another application of the invention - the interaction of a person-machine with the help of the motions of the hand M or, in a particular case, fingers U3 in the zone of CR, divided into rectangular system coordinates, and C<sub>in</sub>
The digital video camera 13 with digital output connected to the device 11 according to the present invention, as shown in FIG. 1, may be used to detect the movement of the hand M and fingertips to control the computer (for example, like a mouse), or For example, some functions of the machine For example, node 13-11 can be used by a gluttonous person who uses a standard deaf-language, built on hand movements, for entering alphanumeric data, and therefore - and text, in the computer, without the use of the usual keyboard, At the same time, this operation, of course, maybe Also, a person who can speak and speak the language of the deaf, enabling the insertion of text into a computer without the use of a custom keyboard, can also be used. Such a system is not sensitive to change over time and does not require accurate reporting of the beginning and end of the gesture.
FIG. 20 and FIG. 21 schematically depict the use of an invention for controlling an auto-vehicle driver, which involves generating a message when a driver is asleep. FIG.
In this case, the camcorder 13 is fixed on the body inside the vehicle, for example, above the rear view mirror, from where it can observe the driver.
The first operation involves cropping the driver in the same way as it was done in the case of applying FIG. 17. FIG. 20 schematically depicts a driver image 1C on a video screen. First, unnecessary rights and the left area (indicated by a horizontal transverse dash of these parts of the image) are shifted initially, which makes the processing area confined to the central part of the zone -display between the two sites mentioned
Further, within this central area, it is sufficient to monitor the unencrypted DiA-bay AA of FIG. 21, which frames its head
Relative movements detected by the device proposed in accordance with this invention are blinking of the driver's eyes (which are reflected by vertical movements in the area of the AA), the blinking speed in the process of falling asleep. If the speed of blinking of the eye of the driver is lowered below a certain threshold, then a mental signal and the driver wakes up
40
22 shows a means for processing a limited number of (p) bits representing a CO constant when it is necessary to take into account a larger range of values of the speed of movement
For this purpose, the Malacha chart will be used (see the article "And the Ingushetia", 3 MIIAi, IEEE Tgapvasyupa op Rai-Yegp Apaevuyev, "Masnipi Yapiyidepse, July 1989, p. 674-893), provides for the sequential division of a complete video image on the successivehalf marked with 1, 2, C, 4, 5, 6, 7 This happens sealing, which allows you to process the image area Thus, at p-4, ie 2<sup>R</sup>= 16, you can determine the speed within the broader range
If, first, within the full image, the device proposed in accordance with the present invention will determine that the speed of the object (in a broad sense of the given term), carrying out the movement, exceeds the maximum velocity corresponding to constant 2<sup>R</sup>-16, then it will only be necessary to examine the plot 1, 2, 3, 4, the image being observed, until the moving object reaches the maximum speed for the corresponding partial image, after the compaction -saving
To use the Mafla composition using wavelets, it is only necessary to add the block 13A (shown in FIG. 22) to the circuit of FIG. 1 to implement said compression video. For example, this block may be "Economical Multi-format Video Codec 07 601", which is manufactured by the American company "ΑΝΑΙ_Ο" (3YE-UISSE), described in the document "Ανν 601 Rheiitschuga Oaia ZNeEI", published by this company in January 1996. FIG. 2 shows an optional sealing block of this type 13a
Finally, FIG. 2 and FIG. 3 show the number of bits (1,3 (if p-3), 8, 15), transmitted by one or other microscales, which allows us to judge the eco-nomy of the sizes for different functional blocks as they process limited number of bits
It is understood that this invention allows to detect relative movement within a surveillance zone observed by an optoelectronic device such as a video camera that transmits the scene under observation to form a digital video signal consisting of sequences of frames which, in their own the queue is composed of a sequence of strings formed by sequences of pixels, and this digital signal is analyzed to identify the area that operates relative displacement, with the definition of SPEED AND (oriented) direction of the shift If the site is performing an efficient movement of a RELATIVELY practically immobile environment
Given that the device according to the invention recognizes the orientated direction and the speed of displacement of the object (in the broadest sense of this term), we can use a means for using the above two parameters to determine the expected position of the object at a predetermined time point and ahead of the orientation of the camcorder 13 In the direction of this expected position
Note that in order to obtain results thatstrengthen with the help of the device, the proposed
59366
42
41
According to the present invention, there is no need to rigidly fix the camera, that is, it and this device, which are used in pairs with the camera, can be placed on a ground, air or marine vehicle (for example, to implement the method illustrated in FIG. 16)
After a very short initial period of time, the duration of N (about 10 correspondent poster frames), the device proposed in accordance with the present invention, determines the parameters of relative displacement immediately after the end of each frame, with which they carry out time and spatial processing, looking at the recursiveness of the calculations provided by the present invention.
One of the embodiments of the present invention disclosed above and several of its applications have been described above. It is evident that this embodiment and the said applications are described as examples which should not be considered as such that they are of a restrictive nature, the specialists will be Obviously numerous variants of the implementation of the application are not beyond the scope of the subject invention as defined in the claims of the invention.
For example, it is possible to implement methods used by signals other than signals at the outputs of block 11 of FIG. 2 shown in FIG. 11, without however being subject to the scope of the subject matter of the invention, as defined in the claims
With regard to the use of the device on-the-go, then it is to no extent limited to thosevariants of application, disclosed above asexamples. Yes, a scheme similar to the one shown in Fig. 17, can be directly connected to a camcorder or video camera to reduce its sensitivity to displacements caused by random user motions
You can also use one or that
a preferred embodiment of several embodiments of the present invention, connected by one or the other, in the preferred embodiment, with a plurality of cemetery or video cameras with a digital video input located inside the building, with the formation of a "smart room" in which such a system can detect and localize the entity and the movement of one person or several people in it, analyzing this migration in order to provide security or identification ι / or assisting in the implementation of certain actions, and may also use ystovuvatysya dlyasposterezhennya child in the next room or supermarket zakliyentamy
It was assumed that the video signal used includes a pair of sequential shunt frames, in particular, when considering the capacities of the LP16, since in the device of the present invention, the pairing of frames is performed, however, it is possible to use one frame of two (for example, unpaired), which allows you to reduce the bridge of the RC, but at the same time causes a decrease in the speed of obtaining the required information at about twice. You can also use an echo camera or other surveillance device, which in the digital output is used alone dr on the image
In some cases, the application, in conjunction with the system of the present invention, may employ special sensors, for example, one or several acceleration sensors, for processing additional displacement parameters
Obviously, the invention is not limited to the specific embodiments and application methods described, but encompasses all variations of the modifications included as part of the general definition of the invention.
59366
44
43
1¾
GV
Ιχι
.zh.
And! 3!
V
138
IP 1
RIS 2
2
45
46
59366p. with
IO 4
47
59366
48
RIS 6
IO, 7
49
59366
50
51
59366
52
5 (i)
PS 11
53
59366
54
■ ч_ 25b
JV
PS 14
S-C
35
55
59366
56
ISL5
43
57
59366
58
IP 18
Su
..! And 1..1.1.1.1.1..1.1..1.1 1 1.1..1.} AND II AND I AND I AND I AND I 1 AND B
See
CID 19th
I-IC
KIS2
Computer layout М Мацело Signed for print 06 10 2003 ТиражЗЭприм
Ministry of Education and Science of Ukraine
State Department of Intellectual Property, Lvivska square, 8, m Kyiv, SME, 04655, Ukraine
LLC "International Scientific Committee", Artema str, 77, Kyiv, 04050, Ukraine
Contents26
Every citation, both ways
| Document | Relation | Office | Cited during |
|---|---|---|---|
| WO2017180079A1 | Cited by | World Intellectual Property Organization (WIPO) | Applicant |
113 members in 20 offices
Priority claims7
| Document | Office | Kind | Date |
|---|---|---|---|
| 9609420 | France | – | |
| 9609420 | France | A | |
| 9609420 | France | A | |
| 9701354 | France | W | |
| 9701354 | France | W | |
| FR19960009420 | – | – | – |
| WO1997FR01354 | – | – | – |
Members113
| Document | Office | Kind | |
|---|---|---|---|
| FR2751760A1 | France | A1 | |
| FR2751772A1 | France | A1 | |
| CA2261854A1 | Canada | A1 | |
| WO9805002A1 | World Intellectual Property Organization (WIPO) | A1 | |
| AU3775397A | Australia | A | |
| FR2751760B1 | France | B1 | |
| FR2751772B1 | France | B1 | |
| EP0909971A1 | European Patent Office (EPO) | A1 | |
| EP0912964A1 | European Patent Office (EPO) | A1 | |
| TR1999000143T2 | Türkiye | T2 | |
| TR199900143T2 | Türkiye | T2 | |
| FR2773521A1 | France | A1 | |
| CA2320815A1 | Canada | A1 | |
| CA2320974A1 | Canada | A1 | |
| WO9936893A1 | World Intellectual Property Organization (WIPO) | A1 | |
| WO9936894A1 | World Intellectual Property Organization (WIPO) | A1 | |
| AU2058099A | Australia | A | |
| AU2619299A | Australia | A | |
| CN1226329A | China | A | |
| WO9936893A9 | World Intellectual Property Organization (WIPO) | A9 | |
| IL127799A0 | Israel | A0 | |
| IL127799D0 | Israel | D0 | |
| BR9710578A | Brazil | A | |
| CA2341823A1 | Canada | A1 | |
| WO0011609A1 | World Intellectual Property Organization (WIPO) | A1 | |
| WO0011610A1 | World Intellectual Property Organization (WIPO) | A1 | |
| WO0011610A1 | World Intellectual Property Organization (WIPO) | A1 | |
| AU2424299A | Australia | A | |
| AU9346698A | Australia | A | |
| FR2773521B1 | France | B1 | |
| KR20000029599A | Republic of Korea | A | |
| TW393859B | Taiwan Province of China | B | |
| AU722228B2 | Australia | B2 | |
| BR9906974A | Brazil | A | |
| BR9906979A | Brazil | A | |
| EP1050032A1 | European Patent Office (EPO) | A1 | |
| EP1050033A1 | European Patent Office (EPO) | A1 | |
| EP1050033A1 | European Patent Office (EPO) | A1 | |
| JP2000516004A | Japan | A | |
| CN1291320A | China | A | |
| TW436436B | Taiwan Province of China | B | |
| CN1299498A | China | A | |
| EP1105840A1 | European Patent Office (EPO) | A1 | |
| EP1105840A1 | European Patent Office (EPO) | A1 | |
| EP1105842A1 | European Patent Office (EPO) | A1 | |
| IL137173A0 | Israel | A0 | |
| IL137173D0 | Israel | D0 | |
| IL137174A0 | Israel | A0 | |
| IL137174D0 | Israel | D0 | |
| US6304187B1 | United States of America | B1 | |
| EP0912964B1 | European Patent Office (EPO) | B1 | |
| AT208070T | Austria | T | |
| ATE208070T1 | Austria | T1 | |
| DE69707886D1 | Germany | D1 | |
| DK0912964T3 | Denmark | T3 | |
| ES2165622T3 | Spain | T3 | |
| JP2002509320A | Japan | A | |
| JP2002509321A | Japan | A | |
| PT912964E | Portugal | E | |
| US2002071595A1 | United States of America | A1 | |
| EP1050033B1 | European Patent Office (EPO) | B1 | |
| EP1050033B1 | European Patent Office (EPO) | B1 | |
| DE69707886T2 | Germany | T2 | |
| AT219593T | Austria | T | |
| ATE219593T1 | Austria | T1 | |
| EP1050032B1 | European Patent Office (EPO) | B1 | |
| DE69901878D1 | Germany | D1 | |
| AT221235T | Austria | T | |
| ATE221235T1 | Austria | T1 | |
| DE69902225D1 | Germany | D1 | |
| EP1105842B1 | European Patent Office (EPO) | B1 | |
| AT225543T | Austria | T | |
| ATE225543T1 | Austria | T1 | |
| DE69808522D1 | Germany | D1 | |
| EP1105840B1 | European Patent Office (EPO) | B1 | |
| EP1105840B1 | European Patent Office (EPO) | B1 | |
| US6486909B1 | United States of America | B1 | |
| AT228260T | Austria | T | |
| ATE228260T1 | Austria | T1 | |
| DE69904056D1 | Germany | D1 | |
| ES2179612T3 | Spain | T3 | |
| ES2179620T3 | Spain | T3 | |
| DE69901878T2 | Germany | T2 | |
| DE69902225T2 | Germany | T2 | |
| US2003067978A1 | United States of America | A1 | |
| ES2188130T3 | Spain | T3 | |
| JP2003521752A | Japan | A | |
| DE69808522T2 | Germany | T2 | |
| UA59366C2This record | Ukraine | C2 | |
| DE69904056T2 | Germany | T2 | |
| RU2216780C2 | Russian Federation | C2 | |
| CN1132131C | China | C | |
| IL127799A | Israel | A | |
| US6717518B1 | United States of America | B1 | |
| CN1157694C | China | C | |
| IL137174A | Israel | A | |
| CN1192327C | China | C | |
| KR100490972B1 | Republic of Korea | B1 | |
| IL137173A | Israel | A | |
| US7181047B2 | United States of America | B2 |
Numbers
- Publication
- 59366
- Publication, DOCDB
- 59366
- Publication, EPODOC
- UA59366
- Application
- 19990010431
- Application, DOCDB
- 99010431
- Application, EPODOC
- UA19990010431
Titles3
- Ukrainian
- СПОСІБ ТА ПРИСТРІЙ, ЩО ДІЮТЬ У РЕАЛЬНОМУ ЧАСІ, ДЛЯ ВИЯВЛЕННЯ, ВИЗНАЧЕННЯ МІСЦЯ ЗНАХОДЖЕННЯ, ШВИДКОСТІ ТА НАПРЯМУ РУХУ ДІЛЯНКИ ВІДНОСНОГО ПЕРЕМІЩЕННЯ В СЦЕНІ
- English
- METHOD AND DEVICE FOR DETECTING, DETERMINING LOCATION, SPEED, AND DIRECTION OF THE RELATIVE MOVEMENT OF A TELEVISION IMAGE SECTION IN REAL TIME
- Russian
- СПОСОБ И УСТРОЙСТВО ДЛЯ ОБНАРУЖЕНИЯ, ОПРЕДЕЛЕНИЯ МЕСТОПОЛОЖЕНИЯ, СКОРОСТИ И НАПРАВЛЕНИЯ ОТНОСИТЕЛЬНОГО ПЕРЕМЕЩЕНИЯ УЧАСТКА ТЕЛЕВИЗИОННОГО ИЗОБРАЖЕНИЯ В РЕАЛЬНОМ ВРЕМЕНИ
Classification
- CPC, 3
- G06T7/254
- G06V40/161
- G06T7/20
- IPC, 2
- G06T7 20
- G06K9 00