Scaner for decoding optically read label and opticaly read label for such device

Description:

The present invention relates to a scanning device for decoding the label containing encoded information, which includes the optical reading tag, means for creating the optical image of the area and generating the card corresponding to the light reflected from the label and the challenge of each graphic element.

The goods, different components, letters, packaging and the whole range of respective products that are transported, often need identifying information that refers to the origin, the number of the race, the destination, the name, the price, the quantity of products and others. In some cases, decoding the information encoded on the labels attached to these products allows the automation of reading the numbers that refer to the sale and inventory or to the work of electronic houses. In other cases, the use of such encoded labels consists of automatic guidance and sorting of mail, packages, luggage, etc., but also for the placement of labels containing instructions for using the raw material or the fillers during manufacture. Labels for such products are usually marked with barcodes, one of which is the universal product code. Many bar code systems are known.

Commercial barcodes usually have the following drawback of excessive data density, a density that has to be connected to the growing demands of encoding more and more information on small labels. Attempts to reduce the size and distance between bars in different barcode systems in order to increase the data density did not solve this problem: optical reading devices, having a special separation capacity for barcode determination, including bars up to 5 mm or less, are usually not economically feasible, as the small tolerances inherent to the label printing process and equipment The complicated optics required to distinguish binary linear codes at these dimensions would greatly complicate the process. To accommodate the increasing data volume, labels with very large barcodes should be made, which would make such labels not compact and therefore not applicable to small products. Another important factor is the price of the label material, such as paper. The small label has a lower cost price than the large label, and the price is an important factor when operating in large quantities.

Bar code alternatives include: round shapes using radially encoded cuneiform elements [1], or binary black and white encoded concentric circles [2,3,4], nets of lines and columns filled with encoded squares and rectangles [ 5], microscopic spots located in cells that form a regular network [6] and multicolored fields of data from densely packed points or elements [7]. Some of the above coding systems as well as other coding systems known in the art suffer primarily from insufficient data density, such as in the case of concentric circles encoded on square and rectangular boxes. In the case of networks consisting of microscopic points and multicolored elements mentioned above, such systems require special means of transport and guidance, which limits their application only in cases of particularly precise guidance.

Due to the contemporary transport dimensions and speeds (which use conveyor belts with the width of 0.9-1.2 m), having the linear speed of the lane moving of 2.5 m / s and more, loading lanes packaging of different heights and encoded information labels and the need to use small, cheap and compact labels with an area of ​​about 1 inch², in optical and decoding systems intended for searching and reading labels with data encoded on these fast moving packaging, high voltages appear. There are difficulties with the optical scanner that only detects the label image. In addition, after the identification and identification of the image of the label must be decoded quickly, before other operations with the packages on the conveyor, often it must be done in fractions of a second.

These problems led to the need to create a simple, inexpensive and fast means of signaling the presence of the tag with the data encoded in the field of view of the optical device, installed in such a way, that the scanner could move across the conveyor belt. This means is preferably connected to the high-density data network, described in detail below.

Data networks, containing mixed targets, are known in the art, for example they are concentric geometric figures containing rings, squares, triangles, hexagons and multiple variations of them [8, 9]. Also, the use of systems containing concentric circles as detection and position indicators is described and these symbols are captured by the products to be read optically [10, 11]. But in these systems, two different symbols are used to detect the data field and its position, which complicates the logic circuit required to highlight the symbols, and also reduces the data capacity of such a field. In addition, when two symbols are used, the failure of one symbol creates problems in determining the data field and the operator's possibilities to restore the information in the data field. In the latter system, special position markers are used, oriented at the opposite ends of the data runway, which has linear markers of the encoded data with limited capacity.

The systems described above usually use scanning with the optical transducer, which allows the output video signal to be created corresponding to the change in light intensity reflected from the data network and the position and orientation symbols. In such systems, the video signal, after quantification, has a concrete array of bits, which can be related to a given series of bits. The disadvantage of these systems is that they require two different symbols: one for image perception and another for determining its orientation. Because of the need to connect the optical signal's numerical signal to the series given by the symbols presenting the position, the orientation and the probability of a wrong reading, it is higher than when using the mode and system of the given invention, because in the known systems of label perception it is ensured о inflexible detection of the target signal level.

The circular network is shown! data [1] with the identification target located in the center, containing о consecutive consecutive concentric circles. Target identification I assure! the presence of a means of searching the round label with the help of the optical translator and determining its geometric center, therefore also the geometric center of the circular data network. This is being done! with the help of a logic circuit that perceives the impulse diagram I present! о shape configuration! of bull's-eye for target identification. But as for barcodes! the data network has limited data capacity о! and the system needs it! the second circular scanning process. Thanks! using both linear and circular scanning for such a system with limited data capacity! its complication may occur for о insignificant increase in capacity.

To increase the data capacity of the network, codes have been elaborated that use the sets of high-density color points [7]. But for these systems optical analysis devices are needed! of the image with manual steering !, which are absolutely incapable of carrying out the recording and decoding of the image at the rapid transport of the data networks on the packaging on the conveyor belts. Analogously, for the high density coding system in which microscopic data encoding points are used [6], a special means of transport is needed, which would ensure that the data network is moved in the default direction, but not in some direction as it could happen with a luggage carried on the conveyor belt. In this way, the coded label! you must! either read face to face, using a linear reading device coupled with the means of transport of the label, to ensure an accurate decoding of the information entered on the label. This patent also shows that c! the position of the fat card! the translator will have to be carefully checked so that the information can be read.

Technically! multiple colors are used to create bar code systems! in order to have no optical problems when scanning very small bars. A bar code! in which more than two are used! optical features for encoding data in the data network, for example using the altitude of the black, gray and white bands is described in [12]. But systems like this, though I present! upgrades to bar code systems! previously known, however, they cannot reach the compactness and density of the data as in the invention described! lower.

From the point of view of the mentioned disadvantages of the known scanning devices for decoding the machine read tag, the problem solved by the invention is! in the creation of a scanning device for decoding the machine readable label with new and improved labels, compact with optical reading !, having о major informational density !, which can be read through an optical translator, when the label can be fixed! luggage or other product, transported with a high speed conveyor system, regardless of the orientation of the luggage or the variable height of the luggage, on which it is fixed! the label with optical reading !, with the system in question! being possible! certainly decoding the label into a form! bent, bent, skewed, partially erased or partially broken, at the same time! this system includes the possibility of correction, so that the incorrectly read information can be restored! or omitted and do so preferably for encoded information! priority, and the system called additional uses! relatively cheap logic circuits.

The scanning device for decoding the optical read tag! contains a block of illumination of a set area of ​​the label crossing, a block of optical image formation of the area established with the label installed! to each element of the formed optical image, as well as a decoder block of the binary card into electrical signals, related to the output of the optical image forming block, the creation of the binary card and the excitation of each element, the decoding block being executed in the form of a block of the two-dimensional reset synchronization signal after tact on the optical label, of a block of detection of the optical the geometric centers of the coded polygons of the label when identifying their optical properties and of a block of decoding of the polygons when inverting the coding process, consecutively linked.

In the scanning device the block of two-dimensional restoration of the synchronization signal after the tact is executed in the form of blocks, joined consecutively, of nonlinear transformation of the digital signals of the label with the identification of the junctions of the optical properties between the adjacent polygons, of Fourier transformation, the two-dimensional presentation that the direction, size and brightness of the obtained junctions, digital signal filtering, except for the incorrect direction and the distribution of the junctions of the optical properties, and of the inverse Fourier transform, which provides the restored synchronization signal after the tact.

In the scanning device is introduced a block of normalization of the image information on the label until the default level for each optical property, related to the block of optical image formation, binary card creation and excitation of the elements.

In the scanning device, a block of image scale transformation is inserted on the label with equal horizontal and vertical amplification, linked to the block of optical image formation, binary card creation and excitation of the elements.

A threshold processing and histogram building block is introduced in the scanning device, which presents the optical properties of the image in each polygon of the label, linked to the block of optical image formation, binary card creation and element excitation.

In the scanning device, the block of detection of the geometric centers of the label polygons is executed with the possibility of determining the region with maximum brightness in the preset area of ​​the restored signal of synchronization after tact and to ensure the continuous comment cycle of such a complete signal from the largest domain. brightness and at the cyclic bypass of each adjacent region with the next highest brightness, at the same time each of the regions found correspond to the center of the polygon.

In the scanning device, a block of detection of the area of ​​the concentric rings of the optical label is introduced in the correlative processing of the input signals and the predetermined frequency signal.

In the scanning device the block of optical image formation, binary card creation and excitation of the elements is executed with the possibility of filtering analog signals corresponding to the values ​​of light brightness, when determining the presence of concentric rings and optical label in the preset area.

The optical reading tag contains informationally coded polygons, the geometric centers of the adjacent polygons being located at the peaks of the predetermined two-dimensional network, and the polygons possess at least one of the two optical properties, having three, five or more sides adjacent or partially adjacent, and is executed with concentric rings in the area separated by the area with polygons, each concentric ring having, at least, one of two optical properties in alternating sequence.

In the optical reading tag the polygons represent regular hexagons, and the two-dimensional network represents о regular hexagonal network.

In the optical reading tag, the concentric rings are located in the center of the label.

In the optical reading tag the optical properties are characterized by the colors black, white and gray.

The data network contains о the square network with the I surface², having connected hexagons that form rows and columns and о identifier target in the center, which determines the geometric center of the data network. The geometric target can be any geometrical figure with optical characteristics that allows the generation of a distinct video signal for scanning with the help of the optical translator on a linear path passing through the geometric center of the target. The target of distinction consists of a lot of concentric rings with contrasting reflection capabilities that result in the creation of a periodic video signal to a scanner. Due to the use of an analogue filter for the detection and decoding of the data network, the signal is compared with the preset frequency, as a consequence it can be performed о fast and accurate frequency tuning and subsequent determination of the supported data network. The analog output signal from the optical transducer, which represents the label with encoded information, is subsequently quantified and decoded. Due to the use of an analog tape filter the label can be identified without decoding it with the information encoded on it.

Centrally finding the target of identification can be determined the landmark on the data network. If the center of the target of identification is located in the center of the label, the center of the target of identification and of the given network can be determined at the same time. It is preferable for the identification target to be at the center of the label, but this is not mandatory in the given invention.

In the data network read optically, according to the invention, 100 or several hundred and more alpha-numeric symbols protected by errors ре о surface of about 1 fol can be encoded², when coding hexagons that use three reflector features of the type black, white and gray. For the transducer with the given optical separation capability 10, the system of the invention allows to create a denser information sharing than is possible in the boundary system. For example, if an optical transducer with high separation capacity is used in the invention system, hundreds of alpha-numeric symbols can be encoded on the 1-sheet surface.². Hundreds of alpha-numeric symbols on о surface (A of 1 fol²) can be easily identified by means of a translator with relatively small separation capacity.

The optical reading labels of the invention can be made with variable data density, using two or more contrasting optical features. In the future for variable density data and the introduction of the identification fin, a more complicated scanning device and the use of large decoding algorithms are needed compared to the bar code system.

According to the invention, data encoding can be performed by encoding the set of a binary series of bifi into a bunch of adjacent hexagons, each hex having at least one of two optical characteristics, although encoding can be performed from one hex to another. The flow of numeric bifi can be created using the calculator, based on data entered manually or transformed in some way into a flow of bifi, or a flow of bifi can be created previously. The data to be encoded are wiped in a set of bifi with о given sequence and in the geographical regions of the network to increase the number of crossings between hexagons having different optical characteristics.

The messages to be encoded are divided into priority and simple messages, which are marked separately in different geographical regions of the data relay. The priority message can be arbitrarily repeated in the non-priority region to minimize the possibility of losing the priority message due to scanning errors such as stains, breaks, bends and defects of other types of data relay. The priority message is encoded in the central region of the data relay, along with the identifying target that is confined in the preferential variant in order to protect the message from defects, which are more likely in the peripheral regions of the data relay. It is preferable to introduce in the data repository the possibility of error correction using the greater capacity of the information placed in the given invention to guarantee a high degree of data integrity during the message decoding.

When using the invention, о refea with dense image elements will be used for printing the label with hexagons of different characteristics, although other printing methods can be used. The image element is mapped so that when the label is printed, the optical characteristics of each hexagon are decoded, so they can be decoded later to restore the data entered when encoding particular hexagons. Such a printing process is well known in the art and for printing hexagons with optical characteristics required for the given invention. Standard printing devices can be used for printing.

In accordance with the present invention, a new, improved method of regenerating the coded data in the hexagons repository, bit by bit, has been developed. Preferred are hexagons that make up the data stream. The encoded labels can be passed through an illuminated region and read optically, by means of an electronic circuit of the optical transducer or by means of a manual scanning device passed over the labels. The optical transducer forms an output signal, which is an analog electrical signal that corresponds to the intensity of light reflected from the particular region of the label. With the help of an analog filter, the analog signal of the optical transducer is compared with the default frequency о which corresponds to the frequency that determines the false identifier, if it exists. At a positive agreement, the label is identified and the center of the identifier is determined, which determines the landmark in the data warehouse. The analog signal is also quantified by means of a digital analog converter and stored in the buffer area of ​​the image. The numeric data entered, representing the entire label, are accessible for further processing in the decoding process.

With the help of the logic circuit with a written program, the numerical data is transformed into a card of hexagons-interfaces with different optical characteristics. In the preferential embodiment of the invention the calculation is made in the standard deviations of the intensities of the reflector characteristics, written with the optical translator on each element of the image in the given group of elements of the image that surrounds this first element of it, for this reason the large standard deviations correspond to the passing regions. between the contrasting hexagons.

Further, the numerical data are transformed using the filter programs to determine the orientation, direction and dispersion of the hexagons. The main stages of this process are:

(1) Filter the nonlinear version of the digital image.

(2) Determining the orientation of the label, it is preferable to determine the three axes of the image and to determine the parallel axis on both sides of the label.

(3) Determining the center of each hexagon and determining the level of gray color in each center.

(4) Transforming the gray color level into a bit stream.

(5) Arbitrary correction of errors in this bit stream.

(6) Arbitrary transformation of bit stream into a sequence of symbols.

It should be mentioned that although the invention describes the process for hexagons having two or more optical characteristics, in particular the steps of adjusting the optical image due to label slippage, ruptures, etc., it can be used for labels of other types and other cells. .

The technical result of the invention consists in the fact that it is possible to carry out the operations of reading the information on the label with high speed and accuracy in a large volume of label information, including ensuring the protection of information from losses due to mechanical defects or mechanical defects. other label failures, due to the fact that the information communication and error detection data are encoded in regular hexagons, placed in a two-dimensional cell matrix and having three reflector characteristics of the type black, white and gray.

The invention is explained in figures, which are indicated below.

Fig. I, the top view of the identification target, which according to the invention consists of concentric circles.

Fig. 2, the fragmentary top view of the optically read tag, having adjacent hexagons for encoding the data according to the given invention.

Fig. 3, the top view of the optical read tag that has adjacent hexagons and has three optical characteristics for encoding binary data and identifying target according to the given invention.

Fig. 4, the aspect of an agglomeration of hexagonal cells (3 x 3) located adjacent and which can serve as a basic block for encoding information in the preferred embodiment of the invention.

Fig. 5, the card of an agglomeration graphically represented by a data network containing 33 rows and 30 columns forming a network of 11 rows and IO columns of a hexagonal coding block in the form of a 3 x 3 cell figure.

Fig. 6, the general image of the camera tuning system, according to the invention, intended to adjust the position of the optical transducer according to the height of the luggage.

Fig. 7, a detailed description of the decoding process of the given invention.

Fig. 8, the program of the structure of the coding and decoding process and the data flow.

Fig. 9, the consecutive program of the processing stages of the invention.

Fig. IO, the appearance of a set of regular tangent hexagons, viewed from above, so located that the geometric centers of the neighboring polygons are located at the ends of a hexagonal network.

Fig. Il, the appearance of a set of irregular tangent hexagons, viewed from above, so located that the geometric centers of the neighboring polygons are located at the peaks of a hexagonal network.

Fig. 12, the appearance of a set of partially tangent polygons, viewed from above, executed essentially in the form of hexagons, so located that the geometric centers of the neighboring polygons are located at the ends of a hexagonal network.

Fig. 13, the appearance of a set of tangent polygons, viewed from above, executed essentially in the form of hexagons, so located that the geometric centers of the neighboring polygons are located at the ends of a hexagonal network.

Fig. 14, the upper aspect of the optical read tag, which has tangent polygons, executed essentially in the form of hexagons, so located that the geometric centers of the neighboring polygons are located at the ends of a hexagonal network and which includes identification target according to the invention time.

Fig. 15, the appearance of a set of tangent equilateral rectangles, so located that the geometric centers of the neighboring rectangles are located at the ends of a hexagonal lattice.

Fig. 16, the appearance of a set of rectangles, without looking above, which determines the intermediate spaces between these rectangles such that the geometric centers of the neighboring rectangles are located at the ends of a hexagonal lattice.

Fig. 17, the appearance of a set of nonangangent pentagons, viewed from above, which determines the intermediate spaces between these pentagons so that the geometric centers of the neighboring pentagons are located at the ends of a hexagonal network.

Fig. 18, the appearance of a set of nonangangent squares, viewed from above, the rows and columns being located in chess such that the geometric centers of the neighboring squares are located at the ends of a hexagonal network.

Fig. 19, the appearance of a set of partially tangent octagons, viewed from above, which determines the intermediate spaces such that the geometric centers of the neighboring polygons are located at the points of a rectangular lattice.

The possibility of encoding the information using the contrast colors of the adjacent hexagons or of the cells located in the cellular structure with о preset consecutive and о network, allows the information recorded on the label to be restored using an electrooptic translator.

Polygonal cells other than hexagons that are so located that the geometric centers of neighboring polygons are located at the ends of a hexagonal network, or other established network, can also be used to encode information on the optical read tag.

Such polygonal cells, when located with their corresponding centers in blocks determined on the two-dimensional geometric network and are encoded in a given sequence, assigning different optical characteristics to the set of these polygonal cells, can be read with the electro-optical translator according to the method of the invention described. In continuation.

The polygonal cells of the invention represent encoded information blocks, formed on the basis of a closed dashed line, while these cells are located in a two-dimensional structure established on the optical reading tag. When using the invention, label configurations can be used in which a large variety of polygonal shapes and networks of different geometries, such as pentagons, hexagons, octagons, rectangles or squares, are used. Neighboring polygonal cells can be completely tangent, partially tangent, or non-tangent in the optical read tag of the invention.

Tangent polygons are polygons so located that the geometric centers of neighboring polygons are located at the peaks of a predetermined two-dimensional network, and the boundaries of such polygons are tangent to the boundaries of the immediately adjacent polygons. The partially tangent polygons are the polygons so located that the geometric centers of the neighboring polygons are located at the peaks of a predetermined network and these polygons are separated along their corresponding borders by other adjacent polygons, based on which we have a lot of intermediate spaces. , scattered among the polygons on the optical read tag. Non-tangent polygons are some singular polygons so located that the geometric centers of the neighboring polygons are located at the ends of a predetermined two-dimensional network and the boundaries of single polygons do not contact the boundaries of the surrounding polygons. In addition, the predetermined polygonal cells and two-dimensional networks or networks on which the centers of neighboring polygons are located may be irregular, may have axes located non-evenly, or regularized with axes located equidistantly in the configuration. Such two-dimensional networks have axes independent of the axes of symmetry, if any, of polygonal cells.

The hexagons used in the label of the invention have some advantages for encoding the information on the label. These advantages are as follows:

I. At the given optical separation the hexagons can be grouped more compactly than other polygons. For example, at separation the optical given the corners of a square are hardly investigated, so excessive optical separation should be required for reading the squares. For the given optical separation, the circumferences would be optimal, but the space between the adjacent circumferences would have been spent in vain and the process of processing and printing the image of the label would be exacerbated due to the need to assign characteristic I or optical intervals. Hexagons allow optimal packing of information with circumferences or other polygons that contain: octagons, squares, triangles, etc. Squares and triangles create problems because of their sharp corners. Circumferences and octagons create difficulties due to unused space between neighboring circumferences or octagons.

2. О The network of adjacent hexagons has three axes. Using о square or rectangular label, the main axis! of the hexagon can be located in an established position! Fat! on the label side. This! positioning the main axis of the hexagonal network makes it easy to read the information encoded in hexagons, due to the position of the fat hexagon! axis.

When used! label includes о feet! discreet! with the opposite side! sticky, which fixes! packaging or article, on the outside! of the container or other object on which the information is read! Optically it is printed in accordance with the requirements of the given invention.

In the case of the girl! Optically read data network or data network present! the image of adjacent hexagons or cells that have two! or more optical features in shape! reversible! for coding, data registration taking place due! optical characteristics and spatial position of single hexagons! by the other. The hexagons or polygons printed with the contents of this restored information are still called hexagons or polygons with encoded information! the way it is coded! information on the label.

This drawing of the adjacent hexagons with a maximum number of hexagonhexagon crossing boundaries for optimal reading and maximum density! of storing information is called cell structure.

The contrasting features used to print individual hexagons or cells in the data network can vary greatly within the meaning of the present invention !. In the case of the girl! printing means introducing materials with optical characteristics, and printing means! introducing materials with optical characteristics given on the substrate or changing the optical properties in case of using the thermal pattern. Printing also means the insertion of material with optical characteristics on the support, when the support itself has о optical feature! distinct!. For example, when printing hexagons in black and white, when the media is white, only black cells are printed. In this way, white cells also fall under the definition of the term print or print.

In this case, through the optical feature! it means the absorption of light, the reflection and / or the change of the printed cells in different environments. When the cells are printed in black (ink! Black! High density), gray (shades of black) and white (lack of pattern on the white support), as in the case of the preferred embodiment of the invention, we will say that the invention has three optical characteristics. After! as shown in FIG. I, in this case, the set of concentric circles I means two! or more concentric circles 2, one of which is the inner domain of zone 3, determined by the smallest radius! ra circles I or.

In FIG. 2 is shown о part of the label with electrooptic reading! In accordance with the invention. After! as can be seen from FIG. 2, the label contains о lots of hexagons printed adjacent to it forming! о cellular structure! Each of the hexagons is marked with position 4 and contains six equal parts 5. The interior angles of the hexagon are equal to 120 °. In this variant! for realization the hexagon has о vertical axis! long! yy and a horizontal one! xx. The size of the hexagon on xx is slightly smaller than the size of hexagon 4 on the yy due! geometry of regular hexagon.

In the label 6 (fig. 3) with the dimensions of about 1 inch x 1 inch, there will be approximately 888 hexagons or cells 4 (considering that, preferably, the center of the label is occupied with the identification target 7 which consists of -a lot of concentric circles). These adjacent hexagons 4 form horizontal rows R determined by the imaginary line! 8 and the vertical columns C determined by the imaginary lines 9. In this example is the label with the dimensions determined by the imaginary lines 9. In the given example the label with the dimensions of 1 inch x 1 inch has a total of 33 horizontal rows R and 30 vertical columns C of hexagons 4. Each hexagon has the diameter of about 0.8 mm. In the square perimeter surrounding the cell structure, rows R are more than columns C due! geometric packing of adjacent hexagons.

When using the hexagons shown in fig. 3 it can be observed that c! the hexagons are arranged in chess with the cover of the vertical columns such that the vertical hexagons have yy collinear axes. The yy axes of the distal hexagons 4 are in a linear correlation from the outside! vertical! 5 of the displaced intermediate hexagon. The yy axes of hexagons 4 are parallel to both vertical boundaries 10 and 11 of the label, which is shown in FIG. 3. The horizontal rows R are measured on the xx axes at the center point of the hexagon 4.

ю

As described in detail below, hexagons 4 are formed in the printing process, process in which hexagons 4 are printed with two or more optical gradients, for example, in contrasting colors. These colors may be white 12, black 13 and also any other color, but gray 14 is preferable, as shown in FIG. 3, although other contrasting colors can be used. Only two contrasting colors white 12 and black 13 can be used, as shown in fig. 2. In the preferential version! 3 contrasting colors are used: white 12, black 13 and gray 14, as shown in fig. 3. The concrete gradients of white, black and gray are chosen to create an optimal contrast in the identification process with the help of the electrooptic translator. The gray level is chosen so that the optical gradient! to approximate between the black and white gradations used to create the label. Etiquette! 6, shown in FIG. 3, can be created! using о tag! discreet! which has the preferential version! surface of 1 inch², or if! an acceptable (preferably white) background is used, the label can be printed directly on the surface of the headlight packaging! a о the separate highlight. Due to the necessity and significance of the existence of an optical background for one of the contrasting colors, it is preferable! use a separate label, because the background color of the label is easier to control. For further determination of the main axis of the label is important! leveling the hexagons printed on the label relative to the parts of the label, which will be discussed below. The label is printed so that the yy axes of the forming hexagons! the cells are parallel to the vertical parts of the label 10 and 11, as shown in FIG. 3.

When reading the hexagonal relay in order to decode the information contained in different hexagons, it is important to! either between adjacent hexagons a clear color contrast. According to the following, the more optical features used to decode hexagons, the simpler the scan device and the mathematical assurance required! for decoding hexagons. At the same time, having more optical scuffs decreases the density of information on the label. Looking for a reasonable compromise between the amount of encoded information and the prefab for scanning labels with many optical features, it was inferred that c! it's preferable! printing of hexagons encoded with three optical characteristics, namely: black, white and gray.

If! the support or the label has a good white background, then the white hexagons can become a beacon! it is necessary! just printing gray hexagons. In the preferential version! Of the invention, the gray hexagonal cells are executed based on the printing of the cells with black flame !, but only every fifth element of the image of the relay of elements of the matrix printer is printed in this way, in the illustrative example described below. The process is performed based on the use of the shading algorithm with a process known in the art !. This allows to print with the printer the proportion of image elements that determine! the gray color of the hexagon, while the printing of the black hexagon is necessary! printing each element of the image that determines! this hexagon. Black hex cells can be created based on cemeal printing! black! standard. After! as described below, mathematical insurance! of scan analysis in the decoding process allows о efficient delimitation! between the black, white and gray features so it's not necessary! о accurate appreciation! of the color. On the other! part, yes! colors other than black, white and gray are used, or different gray color graders are used to create four or five colors, the contrast of the hue shades will be more minute control for detecting differences in optical characteristics between colors. It is worth mentioning c! The use of the black cam is the simplest and easiest way to create a cell relay with three optical features and it is preferable! when carrying out the given invention.

Due to the square shape of the label in the preferential version! and the nature of the hexagonal cells, the edges of the cellular structures contain incomplete hexagons, after! as shown in FIG. 3 these incomplete hexagons are not used for information transmission.

In the preferential version! of the invention the label also confines the identification source. The target of identification 7, shown in fig. 3, border о mulfime of concentric circles of contrasting colors (shown as white and black). The corresponding black rings are marked with the positions 15, 16 and 17, and the white rings with the positions 18, 19 and 20. It is preferable to pretend! either in the geometric center of the label, because s! be less exposed! complete or partial deformities or destruction if the peripheral part! the label is bent, dirty or defective. The dimensions of the image buffer area (described below) are required to store the data on the label! before its identification and are minimized when the identification card is placed! in the center of the label.

The number of concentric circles in the ID can be! it varies, but it was determined that! six concentric rings 15, 16, 17, 18, 19, 20 and their changing grenades are also needed! from white to black and again to white, etc.

Structure technology is used to connect the calculated structure to which the concentric rings are assumed to be in conjunction with read structures. When the connection takes place, the identification target is searched as described below.

The identification target can have any diameter, smaller than the data network, for creating a surface that can make up 25%, but it is preferable to have 7% of the surface of the network. It is preferable for the identification target to have as small a size as the surface occupied by it on the label cannot be supplemented with encoded information. In the preferred embodiment the diameters of the printed rings are chosen such that the outer border of the outer ring 20 constitutes approximately 7.45 mm. Thus, in FIG. 3, the region of the target of identification 7 constitutes about 7% of the surface of the label of 1 inch² 6. At the same time, the identification target 7 acceptable can be printed on the label 6 with the surface of 1 inch² without changing the amount of information, which can be encoded in the hexagonal network surrounding the identification target. As in the case of incomplete hexagons, at the outer periphery of the label, the bits of hexagons adjacent to the outer border of the target are not used for encoding information. The width of each ring must be as large as the dimension of the side multiplied by о side (xx axis in fig. 1) of the hexagons, which allows to improve the separation capacity.

Six rings are enough. This quantity is reasonable and allows to improve the detection of the rings on the minimum surface of the label with a minimum possible of misreading of the false markings on the label and of other false markings outside the label, somewhere on the conveyor belt.

The identification target may have a different shape than the concentric rings. For example, squares, spirals or hexagons can be used to create contrasting crossings between the concentric figures, because the linear intersections on the identification target allow the creation of set crossings that can be received by the electro-optical transducer and measured with an appropriate filter. It should be mentioned that although the spiral is not an assembly of concentric circles, depending on the size and radius of the spiral, it can be obtained о approximation of concentric circles. It is preferable to target concentric circles, because the signal produced when scanning through their center has a frequency that is the same as in the case of intersection in any direction through the center of concentric circles. This allows simplification of the central identification, what will be discussed below, and allows the identification of the target of identification in the one-dimensional search of the analog or numerical output signal of the scanning device, although the method of the invention allows the use of two-dimensional numerical search for superior accuracy. , in case the numerical signal is analyzed.

In this case, concentric rings are called complete rings, partial rings in the form of semicircles, sectors of concentric rings from 180 ° to 360 ° and concentric spirals that approximate the concentric rings.

Because each hexagon can be encoded in three different optical characteristics, in the preferred embodiment 1585 bits of information can be encoded in each hexagon (log₂ 3). Obviously, if more or less optical characteristics are used than 3, the volume of bits encoded in each hex will change accordingly. The coding algorithm is created with the purpose of approaching the maximum data density and increasing the number of optical characteristics passing from cell to cell, which will facilitate the two-dimensional regeneration described below.

In FIG. 4 shows a set of 3x3 cells, which includes nine hexagonal cells 21 and which is the basic coding block used in the preferred embodiment of the invention. It is a preferable concept of coding, but nonetheless essential. Other coding blocks within the scope of the invention are also achievable. As shown in detail below, 3x3 sets of hexagons 21 are mapped to encode 13 bits of information, in cases where the set is completed with 9 hexagons or less than 13 bits of information, if the set is incomplete, which is due no use of hexagons. On о label with 1 inch surface² to the data network containing approximately 888 hexagons and the identification target, which occupies approximately 7% of the label surface, 1292 bits of formation can be registered.

When coding each set, the bottom hexagons 22 and 23 of each set 21, as shown in FIG. 4, are limited in their corresponding optical characteristics such that they are always determined to be different from the intermediate and adjacent hexagon 24. Thus, only a single bit in the hexagon can be encoded in hexagons 22 and 23. At the same time, it is possible to encode a 13 bits of information in set 60 of the coding account of 11 bits in the remaining hexagons. Because when mapping 7 hexagons there are more possibilities of combining than they are used (for example, 3⁷= 2187 combinations compared to 2¹¹⁼2048 combinations), some combinations are canceled such as all black, gray or white, that is, essentially all black, gray and white combinations. The need to have contrasting colors in hexagons 22 and 23 compared to hexagon 24 is due to the fact that the necessary passes for the data regeneration and arbitrary normalization described below must be guaranteed and also the need for assistance in determining the horizontal offset of the data network, as well as is described below. In cases where the encoding sets have 7 or 8 hexagons, 7 useful hexagons are encoded with 11 bits, and the eighth hex, if any, is encoded with 1 bit. For all other partial sets 3 bits are encoded on each pair of hexagons and 1 bit on each remaining hexagon, which will be discussed below.

So this is a very efficient, relatively inexpensive and easy-to-print label that can be easily decoded (using the respective scanning equipment and analytical mathematical assurance). As mentioned, in the preferred version is used the packing of hexagons in the format of 33 rows x 30 columns, on о label with the surface of 1 inch² and о identification target occupying about 7% of the total area of ​​the label. From the practical point of view 13 bits of information are obtained from a set of 9 hexagons so that each cell has 1.44 bits of information. This is less than 1,585 theoretical bits per hex, due to other conditions of the coding algorithm, because not all 3⁷ structures are used, and some of the optical cell-to-cell passes are eliminated.

From the reasonings provided below, it will be seen that in the preferred embodiment of the invention it is desirable to introduce a level of error protection when coding the label so that the actual volume of information on the label will be reduced in order to provide greater data integrity in the decoding process. .

As is clear to a person skilled in the art, the description of the exemplary embodiment of the label in which hexagonal cells are used is directly applicable to optical readout labels in which other polygonal cells are used. The methods described for printing the optical characteristics of the hexagon are equally applicable for printing the optical characteristics of other polygonal cells in both black, white, gray (shades) and in other colors. Similar inconveniences and advantages in terms of data density are characteristic of labels printed with polygonal cells, different from hexagons, when the black, white "and not necessarily gray" optical features are used for polygonal cell printing. As with labels containing hexagons, labels printed with other encoded polygonal cells can be read using a not too complicated scanning device, when only two optical features, especially black and white, are used to encode information in polygonal cells, because at these colors the maximum contrast is reached.

The procedures for encoding the information and algorithms described for labels containing hexagons are applicable for labels printed with different polygonal cells. As in the case of labels with hexagons, cells with incomplete polygons, which may appear at the border of the optical reading tag or appear due to the partial deletion of the identification target containing consecutive sequences of concentric rings, are not used to encode information.

The cell structure contains a network of hexagons located tangent 25, geometric centers 26, the axes whose axes are located at the tips 26A of the hexagonal mesh 27, as shown in FIG. 10. Regular hexagons, ie hexagons that have six sides and six equal interior angles, form hexagonal networks, which are also regulated by configuration and have three equidistant axes (Al, A2, A3), which are located opposite each other under angles. 120 °.

If the hexagons 28 of the label are irregular but symmetrical, for example, if the hexagons are located along the parallel sides 29, 30, the geometric centers 31 of the adjacent hexagons will describe the irregular hexagonal network 32 shown in FIG. 11. О such irregular hexagonal network will have three more axes (Al, A2 and A3) which correspond to the three axes of symmetry of the irregular hexagons, but these three axes will not be located at each other at an angle of 120 °.

Although the hexagonal network shown in FIG. 12 is not regulated by origin, however it is a о two-dimensional geometric mesh or о network that has the established axes. Thus, the positions and intervals of the geometric centers of the hexagons that are located at the ends of the axes of the intersecting hexagonal network are also predetermined. Subsequently the geometry of the hexagonal network is used in the decoding process described below. In particular, the filtering step performed on the transformed numerical data, which corresponds to the measured image with an optical transducer, is regulated on the reflection of the geometry of the tag set so that the numerical presentation of the measured tag can be used for the exact restoration of the primary mesh. In the restoration process the omitted points are also highlighted from the hexagonal mesh. The missing points of the mesh appear because the optical characteristics have not passed between the polygons with similar optical characteristics.

In the case of irregular hexagonal nets, described in connection with FIG. 11, it is desirable to perform the step of determining the main axis, step (3) (e) of fig. 7 of the decoding process performed after the Fourier transform step of the process of identifying the main axis of the optical read tag. The main axis of the label will have the geometric centers of the polygons located along this axis with intervals other than the other two axes.

The configurations of the labels of the invention approximating the preferred embodiment, containing polygonal cells, as described above, are possible when using certain polygonal cells. In FIG. 12 shows the label configuration, in which polygonal cells 33 are used, which are extremely similar to hexagons, but are polygons with 20 sides and not hexagons. Similarly, polygons created with a number of sides smaller or greater than 20 can also be printed. Polygons 33 are partially tangent, as opposed to imaginary tangent hexagonal cells 34, in which they are described.

Intermediate intervals 35, in the example of the label in fig. 12 may or may not be printed with optical characteristics different from those of the encoded polygons. Intermediate intervals do not possess coded information, which is why their presence leads to a lower data density. But if the intermediate intervals divided between the polygons have another optical characteristic, as opposed to the neighboring polygons, then several transitions between the optical characteristics of the polygons and the intermediate intervals can be observed and because of this, in the transformation domain high energy will appear. During the decoding process described below, but the overall jamming level of the system will also increase.

Due to the fact that the label polygons shown in FIG. 12 are located on a hexagonal mesh which has three axes located equidistantly, the geometric centers 36 of the polygonal cells 33 are located at the ends of the hexagonal network 37. Because the drawings are symmetrical, the position, location and spatial orientation of the sides of the polygons are determined and can be determined in the field of transformation of the decoding process with minor changes, required in the two-dimensional synchronization restoration program. In particular, some small changes can be made with the filters used to filter the transformed numerical data, which represent the coordinates, the range and the brightness of the optical characteristics that were measured with the optical transducer on the illuminated image of the label. Such changes are obvious to a specialist in the given field.

The label shown in FIG. 12 is used with hexagon shaped polygons. Some polygons separately have more than three axes of symmetry, but because they are very close to the hexagon, the medium-resolution optical transducer can read them as hexagons. The geometric centers 36 of the polygons 33 are located at the ends of three equidistant axes (Al, A2 and A3) of the hexagonal network 37.

In FIG. 13 shows the polygonal figure 38 of analogous form (compared to the polygon 33 of Fig. 12), which can be made completely tangent. These polygons 38 can be described using the virtual hexagon 39 shown in FIG. 13, but between the real polygons it will be impossible to find the intermediate intervals (35 from fig. 12). О such tangential construction is preferable for simplifying the decoding process, but it is not mandatory in the practice of using the invention. Polygons 38 are shown in position when their corresponding geometric centers are located at the ends of a hexagonal lattice. Also, as the polygons 33 of FIG. 12, the polygons 38 are executed essentially in the form of hexagons and at о moderate resolution optical capacity they can be taken as hexagons.

In FIG. 14 shows the dilatation of the label that would have been produced, if it had been printed with a о point printer that prints 200 image elements per 1 inch. Polygons 40 of FIG. 14 indicates the shape of the geometric figure that will actually be printed in place of the hexagon with о such point printer, due to the density of the image elements of the printer. Printers with a higher image density should provide a closer approximation to the hexagon than the polygons shown in fig. 13. Thus, the polygons 38, of FIG. 13 and 40 of FIG. 14 are probably ancillary products due to the limited possibilities of some printers in the process of printing labels containing hexagonal cells or are created as a result of persistent attempts to print approximately such polygons in the form of hexagons. The shape of such polygons in the background is hexagonal and allows them to function as equivalents of the coding cells in the form of adjacent hexagons.

As in the case of FIG. 3, the optical reading tag shown in FIG. 14 also contains о consecutive identification target of concentric rings 42-47. Like the hexagons on the label in fig. 3, the polygons 40, which have in the form of hexagons in FIG. 14, are located in rows R and columns C, surrounded by imaginary lines 48, 49, 50 and 51, respectively. Also, as in the case of the hexagons in fig. 3, the polygons of FIG. 14 they have their geometric centers located at the ends of a hexagonal relay determined by the axes located equidistantly Al, A2 and A3. Thus, the labels with the configuration shown in fig. 14 is encoded and decoded in accordance with the method described below.

If the label is used geometrically or alternatively, then the use of rectangular, pentagonal or octagonal mesh or other similar structures will have to be made in the two-dimensional synchronization restoration process described below. For the different geometries of the determined relay, changes are needed, which must be made at the two-dimensional filtering stage! of the process of restoring synchronization. Filters operate! with transformed numerical data, which correspond to the optical characteristics of the polygons read with the translator in the field of image. Such non-essential adjustments to the filtering scheme can be easily carried out by an ordinary specialist. Where the two-dimensional network! determined! it has non-equidistant axes, that is! when it has an irregular configuration, it may be necessary to determine the main axis of the label before performing the transform! ™ Fourier of the numeric data, which represents! the optically measured image. This is because c! the intervals of the geometric centers of the polygons will not be equal in the field of transformation.

Non-tangible polygons can also be used to create an optically readable label according to the present invention. In FIG. 15 is shown о hexagonal network! from the 52 squares, which are located headlamp! to be reached, and their corresponding geometric centers 53 are located at the ends of a hexagonal relay formed by three axes equally situated Al, A2 and A3. Obviously in the case of the girl! configuration is based! on the hexagon of virtual hexagons 54, described around polygons 52, due to which they are formed! intermediate intervals 55.

Analog relays for square 52 shown in FIG. 15 can be constructed using rectangles. In FIG. 16 shows a set of rectangles 56, the geometric centers of the adjacent rectangles being located at the ends of a hexagonal relay formed by the axes, which intersect! Al, A2 and A3. The picture of the hexagonal structure is also complemented by the virtual hexagons 57 of FIG. 16, described around the nonangangular rectangles 56, based on which the intermediate intervals 58 between the rectangles 56 are created. 17 is also shown о constant label! from non-tangent pentagons 59, the geometric centers of the adjacent pentagons 59 being located along three equidistant axes Al, A2 and A3. The geometry of non-tangent pentagons can be reproduced even more easily, describing pentagons 59 with virtual hexagons 60, forming at the same time! intermediate intervals 61 between pentagons 59. О hexagonal network! alternative can be built! in case the axes Al, A2 and A3 of the relay are located evenly, ins! do not correspond to the symmetry axes of the polygonal figures. Instead, the geometric centers of the adjacent polygons are located at the ends of the intersecting axes. This construction is shown in FIG. 18, where it exists! о a set of tangent squares 62, the geometric centers 63 of the adjacent squares being located along the axes Al, A2 and A3.

Larger polygons can be built analogously on о two-dimensional mesh! established. In FIG. 19 is shown о a set of partially tangent octagons 64, which determines the number of intermediate intervals 65 between the octagons 64. The centers 66 of the adjacent octagons 64 are located at the peaks of the intersecting axes! Al and A2, thus forming the network of octagons 64, which can be used! in the practice of the given invention. Intermediate intervals 65 can be printed with feature! optical! different! of that used! for octagons 64. But this is not mandatory in the practice of using the invention, because! namely the position, orientation and brightness of the optical characteristic in the center of the octagons 64, located in the preset position! on the hexagonal refeaua !, formed! from the axes Al and A2, they are the most important in the decoding process.

It should be noted that although it is shown and described! preferential variant! of the label, many variations of the headlight tag are possible! removal from the volume of invention. For example, the label does not have to be! surface of 1 fol². The unit of 1 fol² was chosen! as a rational example of the size of the label, which allows о acceptable density! of the data of 100 alpha-numeric symbols of the information registered with a high degree of error protection, beacon! have excessive label size. It is desirable that the surface of the label s! either of 1 fol² to reduce the preface of paper and other expenses related to the printing, transport and processing of such labels. Ordinary bar code labels! of analogous dimensions have a much lower data density. Using 4-5 or more optical features or colors to outline the hexagons, it can be packaged with much more! information in a surface of hexagons of established dimensions, but at the same time! it will complicate the mathematical assurance and sensitivity of the scanning system needed to restore this information. Therefore, for purely practical reasons, a coding system with three optical characteristics is needed: black, gray and white.

Besides the fasciculation of hexagons in sets of 3x3 cells described above, other set structures can be used or the fasciculation can be completely eliminated, and the coding algorithm can be of special purpose for the individual structure of the hexagons. In the given invention, it may also vary within wide limits the volume of encoded information of a message to the detriment of error correction.

The coding process for the preferred embodiment of the given invention is described below. It should be noted that the preferred embodiment is described below, but nonetheless, many combinations, variations and changes within the limits of the given invention are possible.

The process can be started with о sequence of data to be encoded on the label. In the preferred embodiment the label is о transport label, and the data is divided into two fields determined as priority messages and simple messages. But it is remarkable that the invention is not limited to two different messages or levels of priority. Many messages and priority levels can be created within the quantitative limits of the given size label and cell number.

For example, when the label is transport, the priority message may consist of 9 symbols, which represents the postal code of the recipient of the luggage, package or letter. Although many natural and legal persons have a 5-digit postcode, this is 9-digit, as the 9-digit postcodes are getting more and more widespread. For this reason, when luggage is sent, the most important о has the postal code. It determines the main destination of the luggage and allows the use of different scanning and control systems to send the package to the place of destination by car, plane or conveyor system.

The simple message may contain, for example, the name and address of the loading point, including the recipient's postal code and also account information.

As a basis for creating priority and simple messages, it is necessary to protect priority messages with excessive error correction, which allows placing (encoding) priority messages in the central region of the label, where it is less likely to damage and destroy it, as well as , repeating and distributing the priority message in simple messages, so if a priority message is partially destroyed, there is a possibility that this message can be restored from the simple message.

Due to the distribution of the priority message in the central region it may be necessary only to decode the priority messages for certain purposes, so only part of the label will have to be processed, which will allow the processing process to be speeded up. This can happen, for example, when the luggage is on the conveyor and only the postal code has to be determined, in order to clarify which of the conveyor branches must continue to move the luggage.

Due to the nature of the simple message, it is not present twice on the label. But, as described below, both simple and priority messages can contain different codes, which protect them from errors and correction modes in order to maximize the possibility of accurate restoration of both messages.

The use of error protection symbols as a component part of the encoded information may be in the preferred embodiment of the invention, in combination with a properly written program and a computer, to allow the system to correct errors during the decoding process, as described below. The use of error correction codes is well known in the art and is within the competence of the specialists in the given field.

From the point of view of the practical use of the invention, the operator making the label can manually enter the data into the corresponding terminal of the atom's calculus, which is intended, as shown below, to include the printer that prints the label with the priority message and with the simple one, properly coded in the label's hexagons. It is not essential that priority messages and simple messages really be created, but this is preferable so that the most important data subject to coding is restored. The preferential variant of the label is also printed with о identification target in the center, which contains a set of concentric circles of two contrasting colors, and the colors are actually chosen from two ranges, which are now printed using hexagons, and those more preferable for maximum contrast would be black and white.

The operator who enters the data manually proceeds so that the corresponding programmed computer encodes each symbol of the input message, using the field identifier in order to create in the computer о consecutive bits, representing the message symbols, encoded in the fields for priority message notation and simple and relative position of each. The operation is performed using the program (fig. 8).

The process can be started with the information that must be encoded and is already contained in a sequence of bits, because it was taken from a storage device or created in one way or another. Because of this, the message, which must be encoded, can exist in such a way that it is manually transformed (with the help of electronic equipment), into a consecutive bit or that starts with a consecutive bit.

As soon as the bit sequence is created or protected from errors, because it is created at the next step, the sequence must be mapped according to the default structure for encoding the hexagonal cell structure according to the invention. In FIG. 5 The input card showing the individual hex cells in 3 x 3 cell sets, leveled in the network or cell structure containing 33 rows and 30 columns of hexagons, is shown. Each row and each column are numbered. The number of rows is from 1 to 33, and the number of columns - from 1 to 30. Note that some hexagons highlighted along the top right surface of the card within the boundaries of the geometric center of the network are marked with X. This indicates that the respective hexagons do not contain bit-mapped information. This is due to the fact that the outer X's represent partial hexagons at the edge of the label, which leads to the presence in each of these lines of a small hexagon. The internal hexagons, denoted by X, represent intervals occupied by the identification target or incomplete hexagons along the perimeter of the identification target such that these internal hexagons, denoted by X, are not mapped at the bit level. All hexagons, which are not marked with X, can record the information. According to the preferred embodiment, each of these intervals must be occupied by a black (B), white (W) or gray (G) hexagon. As noted above, although various packing and mapping techniques can be used, the application of this invention allows the use of packets of 9 hexagons in 3 rows and 3 hexagons in each for organizing the specific bits of information and, as described above, it is preferable to encode 13 bits of information in each package like this from 9 hexagons. In the data network containing 33 rows and 30 columns of adjacent hexagons it is created and can be provided (fig. 5) о nets of 11 rows and 10 columns of sets of hexagons, each of which contains о structure of adjacent hexagons in 3 x 3 cell shape. Note that each set of three rows and three columns of cells within the network limits of 11 sets x 10 sets contains a set of 7 or 8 hexagons due to the geometric packing of the hexagons and their number will change from row to row. Therefore, in this arrangement there are 6 sets containing 8 hexagons and 5 sets containing 7 hexagons. The identification target located in the center allows the creation of incomplete sets. In FIG. 5 shows the graphical presentation of the sets in the hexagons for encoding the information bits in a 33-line data network on 30 columns of adjacent hexagons.

As can be seen from FIG. 4, the sets of 9 hexagons are encoded using the following algorithm:

Take 11 bits and map in a sequence of 7 hexagons denoted by a, b, c, d, e, f and h.

The hexagons g and i are each used to present a bit with the guarantee that each of them is different from the hexagon h.

In this way, 13 bits of information are encoded in a complete bundle of 3 x cells from 9 adjacent hexagons.

For partial sets of 7 or 8 hexagons, 11 bits of information are taken and mapped in a sequence of the first 7 hexagons used. The eighth hex, if any, is used to represent a bit.

For all other partial cells:

Mapping of 3 bits of information in as many pairs of hexagons as possible. All other single hexagons remaining are used to represent a bit.

Considering that when mapping 7 hexagons we get more combinations than 11 bits (ie 3⁷= 2187 as compared to 2¹¹⁼2048), some combinations of hexagons will be eliminated. The combinations eliminated are those that allow a minimum number of passes. To perform these are created the tables for mapping the sets according to fig. 5. Creating and using these tables is within the competence of a qualified programmer. According to FIG. 9 the program for creating the tables BINHEX.LUT and HEXBIN.LUT can be noted MK HEX LUT.

Applying this bit distribution scheme allows the encoding of 1292 bits of information in a data network of 33 rows and 30 columns of adjacent hexagons.

Consistency in which the priority and simple information is located throughout the card is determined according to:

(a) the size of the priority message;

(b) the size of the simple message;

(c) optimal distribution of the priority message in protected places.

Using the input card shown in fig. 5 as a template of the MKMAPS.C program, with the mapping registration that operates with numerical data contained in the memory, it determines first of all how to distribute the information: both the priority message and the simple message, about what will continue to speak.

To minimize the possibility of error and to ensure the possibility of error correction, the preferred embodiment of the invention should contain о effective error protection and correction possibilities. For example, in the preferred embodiment containing 1292 bits of information encoded in a hexagonal network on the о label with о surface of 1 inch², with 33 rows x30 columns of hexagons and an identification target that occupies about 7% of the label surface, it is desirable to use 36 bits of information of the priority message for encoding the 9-symbol postcode plus a find-out symbol. additional number, which may represent the load code. In this example, 120 control orders should also be used for the priority message. This is determined by the number of error correction possibilities required. In the analog example the 560 bits of the simple message include 40 bits of the priority message that are included in the simple message. In the example given 576 simple message control bits are completed in order to maintain the protection and facilitate the recovery of the message. This example illustrates the wide application of the control bit in order to protect and recover the priority message over the simple one. It should be noted that the information provided is provided as an example, and the priority message could have been longer or shorter, as well as the simple message, the number of control bits higher or lower depending on the actual application of the invention. The systematic code takes the specific consecutive of the message and adds the consecutive of the error control in the consecutive of the message. The non-systematic code takes the specific consecutive of the message and includes the consecutive of the error control with the consecutive of the message, so that the message is no longer highlighted, but of course it is restored. The use of systematic or non-systematic coding for error protection is within the scope of the invention. Next we will talk about the systematic code.

It was shown that the "overlapping symbols for error highlighting" stage contains systematic and / or non-systematic coding.

Different linear cyclic error protection codes are known in the art, for example the BCH code, the Reed-Solomon code and the Hamming code. In the preferred embodiment, the Reed-Solomon codes are included separately to protect the integrity of high and low priority messages. ReedSolomon codes are very efficient and are especially useful for controlling many-bit symbol errors.

Based on the examples, some information about Reed-Solomon codes is illustrated. The specific characteristics of the Reed-Solomon code can be given with the following parameters:

m = number of bits in each symbol, n = number of symbols in block = 2m-l, к = number of symbols in message (number of bits in message = k), t = ability to correct in number of symbols = (nk )/2.

The postcode with 9 symbols and a unique alpha-numeric symbol for further perception requires 36 bits without error protection in the example below. The Reed-Solomon code with the following parameters was chosen for the priority message.

m = 6 (6-bit rows) n = 2⁶-1 = 63 t = 10

So, к = n-2t = 43

Because to represent a 36-bit message, 6 symbols of 6 bits are required, the remaining 37 symbols (43-6) are filler symbols that are supposed to be between the encoding and decoding devices and which should not be labeled. So the total quantity! of bits needed! on the label! for the priority message const! from (63-37) x6 or 156 bits.

This error coding scheme allows for the correction of a maximum of 60 (10x6) elementary errors, which represents up to 38.5% of the bits used. Due to the large number of fill symbols, the high capacity to detect Reed-Solomon code errors makes it extremely unlikely! о incorrect reading! of the priority message.

Simple message 4 was coded with Reed-Solomon code with error protection having other parameters namely:

m = 8 (8 symbols in 8 bits) n = 2⁸ - 1 = 255 't = 36 k = n - 2t = 183

Because! I exist! 1292 bits for coding on the label! In this example, the total number of 1336 bits (1292 - 156 bits of a priority message and control bits) are available for single message coding and control bits. In this way the remaining 904 bits (255 x 8-1136) are supposed to be fill bits. This allows you to receive 560 bits (183 x 8-904) for the information in the simple message and 576 control bits.

For greater assurance of restoring the priority message it is also confined to the simple message. Reed-Solomon error-protected code is used for simple messages and allows additional coding! has 86 alpha-numeric symbols of 6 bits and has the maximum capacity! error correction of approximately 25.4%.

Using the encoding described! 4 Reed-Solomon with error protection, a total number of 1292 bits of information can be obtained that are confined on the label !, distribute it as follows:

priority information bits,

120 priority control bits,

560 simple informational bits, (including 40 priority message bits included in the simple message)

576 simple control bits.

Consecutive bit data, including appropriate control bits for storing information, is intended! separate hexagons from the input card (fig. 5). It is worth mentioning c! can be used! о great variety of distribution schemes considering c! the most important criteria to be determined are:

(1) the safe location of the priority message near the identification target (if any!);

(2) creating a structure that within reasonable limits allows the assembly during reading.

The specific error coding program, used in this example, is confined to the annex on the microfiche with the name ERRCODE.C. Coding in Reed-Solomon codes requires multiplying the message code vector to the generating matrix. The matrix swelling is done using the arithmetic of the Galua field. Adding any two! elements of the field occur from the account of the exclusion operation or between both elements. The swelling is done by logarithm in the field of Galua. Logarithms and antilogarithms are taken from the tables received from the primary polynomials, especially for the priority message: 1 + x⁶ and for the simple message: l + x²+ x³+ x⁴+ x⁸. After! as shown in FIG. 9, the GF.C auxiliary program creates! the information tables needed for the arithmetic of the Galua field. The information tables are calculated and stored in the GF.LUT file for use during encoding and decoding. The generating polynomial g (x) for the Reed-Solomon code is determined by the equation:

g (x) = (x + a) (x + a²) ........ (x + a^2t), where a is the basic element! of the Galua field.

The generating matrix for Reed-Solomon codes is formed! to the long split! of each row of the generating matrix. The row к of the generating matrix is ​​the remainder of the long division! of his x '^k· 'The g (x).

The calculation of the generating polynomials g (x) and of the generating matrix for priority and simple messages is done according to the auxiliary program MKRSLUT.C The information tables for the generating matrices are created and kept in the RS.LUT file.

The label contains hexagons, which are printed with standard, accessible and inexpensive equipment. Printer with о 300 x 300 dot matrix on 1 inch² gives good results when printing three-color labels (black, white, gray) with 888 hexagons plus identification target in the center. One of the printers with the given separation capacity is the Hewlett Packard Laser Jet Series II with 0.5 megabytes of memory and the graphics separation capacity of 300 dots per inch. The network of 300 x 300 image elements having a density of 90000 elements per 1 inch² can create approximately 90 image elements for a hexagon in the preferred version. Each element of the image is assigned a value of 0 or 1, which represents the white or black element of the image. This printer is used for two-color network printing consisting of white and black hexagons. It can also be used for three-color network printing consisting of black, white and gray hexagons, if the shading algorithm described above is used.

As shown in FIG. 8, with the help of the MKMAPS.C program, was created the table REGIONS.LUT from 34 rows on 30 columns, which is analogical fig. 5, but which is adapted for noting the choice of black or white color for the rings of the target identification. Some single hexagons are coded as black, white, gray or unusable. О separate table of information HEX MAP.LUT is created by the subprogram of the MKMAPS.C program, which indicates that each of the 300 x 300 image elements belongs to the network of the image elements of the concrete regions in REGIONS.LUT, that is approximately 90 image elements per hex. . The image elements belonging to the orientation rings are encoded in white or black. The rings of the identification target are printed due to the generation of the hexagonal structure in each region and then due to the generation of the rings. Regions partially or completely covered with rings are considered by the REGIONS.LUT program as unusable.

The consecutive coding of error protection bits is mapped according to the consecutive given in the network of 11 x 10 hexagons. As shown in FIG. 8, the sequence indicated by the ORDER.LUT table generated by the auxiliary program registered under the name ORDER.C, was used to indicate the values ​​0, 1 and 2 distinct regions when printing on the label, leaving the regions with the value 3 intact. The level of gray color for each of the hexagons in the 3 x 3 cell set is indicated by the program registered with the name CELL CODE.C.

In this auxiliary program, priority was given to memorizing priority messages in the regions near the identification target, where the label is subject to less degradation. That is why the LABEL.C program is used to create a sequence of bits to be inserted into the laser printer.

It should be mentioned that the use of black, gray and white colors allows the creation of a simple label printing procedure, because it only requires black ink when using a standard shading algorithm, well known in the art. If other color combinations are used (which is achievable), the need for printing with other colors leads to о essential complication compared to the three-color variant: white, black, gray or two colors: black and white.

Therefore, when each printed image element is assigned a white or black value, the label can be printed in order to create a coded format as shown in fig. 3, in which some hexagons are white, some gray and others black and where the target region of identification with white and black concentric circles is at the geometric center of the label.

Since the recording and printing of data on the label is described, it is necessary to describe the subsequent interpretation of the label, or the decoding process. It is desirable that the interpretation of the label be done at high speed, during fractions of a second, for the efficient processing of the luggage (or for another processing, or rather pulling the label).

There are two methods that can be used to capture the image when reading the tag. The label can be read at relatively low speeds with a static manual scanning device with the focal point fixed. It is preferable to use an optoelectronic transducer with an automatic focusing mechanism, which allows dynamic scanning of a rapidly moving luggage with variable dimensions and heights for good efficiency. The decoding process and device described below have been demonstrated in connection with the static scanning device with the outbreak fixed. The process that has the possibilities described in comparison with the static scanning device with the fixed focus, is feasible for a dynamic scanning system with some modifications of the optical system. When processing luggage moving at high speed, a fast scan device is required, which would allow reading of the label moving at a linear speed of approximately 2.5 m / s or more and passing below the fixing point of the device for reading. Therefore, the image processing function consists of the following steps. In FIG. 7 are the main stages of the reading process.

1. Illumination of the label

When the luggage, package or letter moves on a fast conveyor, the region to be illuminated is quite large, because the dimensions of the packaging on the conveyor can be quite large and variable. For example, for the luggage processing system, the dimensions of the conveyor with the width of 1 m and the width of the packaging from a few inches up to 1 m (and analog height) are obtained. Therefore the label with the surface of 1 inch² it can be found anywhere in the conveyor. It is possible that the package is located at sharp angles to the axis of movement of the conveyor belt. Shipments, packages, letters and other similar items may have different heights, so the scanned labels may be 1 inch or less above the conveyor, on the side, and 90 cm or more above the maximum height. of the packaging that can be received by the described system, on the other hand.

In order to illuminate the label, according to the invention, especially considering the range of the wide widths of the package, the heights and angles under which the label is viewed, it is desirable to use a powerful light source, which would give good results for two or more optical characteristics, used for label printing. The light can be infrared, ultraviolet or visible and the spectral light used can be variable. It is preferable that the technique used includes the light reflected from the white, black or gray hexagons on the label.

The light source must create enough light reflected on the optical transducer (for example, a device with load links described below), so that the optical transducer can safely differentiate between the black, white or gray shades, the optical characteristics of the hexagons to be measured. . In a dynamic scanning system, a network of luminescent diodes can be used to create the illumination level of the order of 10 mW / cm.² in the region of label illumination or in a linear network with о cylindrical fixing lens. In the present invention a laser light source can be used, the rays of which are passed through a suitable optical system.

The choice of the light source and the properties of the light source is the competence of the specialists in the given field. It should be noted that because the label has dimensions of only 1 inch² and is at a height of 90 cm on a о band with a width of 1 m, moving at a speed of about 2.5 m / s, it is very important that it be well lit for its exact identification.

In the case of an optical transducer with fixed focus used in the given example, the illumination level of 2 mW / cm² it was enough. This level was created using a luminescent light source.

2. Optical measurement of the reflected image of the label

The second stage of the decoding process consists in the optical measurement of the illuminated region using an electronically guided optical transducer. The translator used in the illustrative example for a static fixed-focus scanning system consists of an industrial television camera, having a load-bearing tube, model WV-CD 130, produced by Panasonic Industrial Company, One Panasonic Way, Secaucus, New Jersey 07094, connected to a television lens including a 5 mm tube with a focal length of 50 mm, optical power 1.3, produced by DO Industries Inc. (Japan) with the NAVITRON ™ trademark. The camera is joined by the image capture plate, indicating model number DT-2803-60, produced by Data Translation Inc.

Optical measurement may include examining the entire label, using a spatial transducer of the kind described above and the image capture plate or as an alternative may be performed by a load device on an integrated circuit, where the second scan measurement is made from the account of the movement of the package (of the label). A good model for this purpose is the Thomson-GSF THX 31510 CDZ, element 4096 of the fast linear load transducer produced by Thomson - GSF Division Tubes Electroniques, France.

In dynamic systems that contain the movement of the labeled luggage in the conveyor system, the long optical path between the label and the light source is preferred. The first purpose of creating a long optical trajectory is to reduce the variation of the virtual dimension or to increase the label measured with an optical transducer. For example, if the optical path is 1.2 m, then the visible size of the label that is 2.5 cm above the conveyor belt will be totally different from the size of the label that is 90 cm above the conveyor belt. If a trajectory of approximately 6 m is used, the virtual dimensions of these labels will be the same. Due to this fact it is possible to complete the entire region of the translator or close to the whole region to be measured, regardless of height, which allows a high degree of image separation. If a space translator is used instead of a linear one, the same principle applies. This can be done with the help of a long optical trajectory, shown in fig. 6.

To make it possible to focus on the labels of different height packages, a height translator is required. An ultrasonic transducer can be used for this purpose or consecutive rays of light can be crossed by luggage. Either of these methods is applicable and can trigger the closed or open contour mechanism that adjusts the focus for adjusting the position of the optical elements of measurement (for example, the transducer lenses), one against the other continuously, as shown in fig. 6.

In FIG. 6 shows schematically the system of focusing and adjusting the camera that works according to the invention, for adjusting the position of the optical transducer of the camera in accordance with the baggage height. In FIG. 6 shows the lens 67, the coil mechanism, the height transducer and the reverse link circuit according to the invention. The height transducer 68 shown in FIG. 6, it may be an ultrasonic transducer or light beam intersected by the packaging moving on the conveyor. The output signal of the height transducer comes from the microprocessor 69, which in turn includes the coil mechanism 70, moving the coil 71 on which the device 72 is installed with the load or any other convenient optical transducer. The wave position transducer 73 measures the coil position and its output at the microprocessor 69 which terminates the reverse link circuit for measuring and adjusting the coil position 71.

The transducer must be able to measure the light reflected from the illuminated label and emit an analog signal that corresponds to the luminosity of the reflector characteristics of the label, recorded by some separate elements of the electro-optical translator.

The convenient light source could be one installed above the conveyor at the given height and covering the region along the entire conveyor belt, with light of о quality and determined brightness. The light reflected from the label can be reflected with о consecutive reflectors and only then measured with the electro-optic transducer.

The purpose of a curved optical path is to create a compact and rigid system.

The analog output signal is filtered. The electrical analogue signal is used due to the analog tape filter for determining the existence of the identification fake on the data repository. Thereafter, the analog signal is converted to a numerical signal, using a common analog-to-digital converter, included in the image capture plate described below or any other known method. Instead of an analog tape filter, a numerical filter circuit can be used to determine the existence of the identification trick, comparing the numerical data ее о determines with the quantized signal of the analog-numerical converter, which will be discussed in the following.

As an example of a space transducer having an integral circuit with load and a set of detectors, used in accordance with the invention, is the television camera described earlier than the Panasonic WV-CD 130 type. The analog output signal of the transducer is transmitted at the image capture board of type DT-2803-60 which confines о monochrome conversion with 6 bifi for quantification and further processing. With the help of a subprogram respectively, the data from the output of the image capture board are stored in memory as an exact image, written with the optical translator.

3. Reflected image processing

One of the most important aspects of the invention is the processing of the optical image with the purpose of creating it repeatedly and accurately orienting the initial configuration of the label and colors (optical characteristics) of each hexagon. This is due to the use of the following steps, after which the known structure, according to which the label was initially encoded and bit by bit mapped, can be used to decode the information that is confined to the label.

(a) Finding the center of the shaft

Before using the television camera having a device with the load described above and the image capture plate, shown in fig. 10, operates the DTINIT.C 74 initialization program to install the image capture board in the given position and to load the tables in memory, and then works the DTLIVE.C 75 program that installs the image capture board in live mode. After this the DTGRAB.C 81 program directs the operation of the image capture board in the process of quantifying the image in 250 rows and 256 columns, and the created models are stored as 6-bit sizes, leveled right in octaves.

Upon first acquaintance with the label image, an analog tape filter may be used to identify two or more characteristics of the concentric rings of the identification fin. Both of these optical features actually have black and white colors, because the highest contrast is created by the highest energy signal. In order to find a fixed structure of switching from black to white and again to black, etc., at the linear scan of the target of identification and the passage through the center it is preferable to emit a homogeneous frequency signal regardless of the orientation of the label. In this way, the target rings are arranged optimally from concentric contrasting rings. After this the output signal of the translator is reversed and transmitted in two directions for detection. In one direction all the energy is detected in the output signal, and in another direction the frequency energy of the rings is measured. When comparing both output signals, the energy in the ring detector is closest to all the energy in the detector when measuring the scanner through the central target identification. The central target of the identification is determined when this first approximation takes place. But in the preferred embodiment, priority was given to an analog band filter at the first stage of filtering, or to a selective band analog filter, although a numerical filter is applicable.

It should be noted that the stage of finding the identification target noted by FIND.C 76 in fig. 9, is shown as optional in FIG. 7, because in the present invention the manual scanning device can be used and in this case the operator can precisely locate the scanning device in order to guarantee a correct leveling of the translator. This, of course, is a slower process than the use of an automatic translator and quick processing is preferable to the latter. If the automatic (and not the manual) translator is used, finding the target is a mandatory step in the process.

As an alternative to the analog filter described above, a numerical band filter can be used, where the Parks-McCeellan algorithm is used, along with the mathematical assurance of the numerical filter constructions for the IBM PC.

For the present invention, a one-dimensional numerical band filter was used to filter a consecutive bit sequence, which will be discussed later with the following filtering programs. The band to be filtered is the presumptive frequency of the rings. The one-dimensional numerical tape filter was intended for a frequency of arrangement of 400 elements of the image per 1 inch and the length of 125 elements of the image (or 0.3125 inches) and was intended to operate on the basis of the size of the printed rings in the identification target, as shown in FIG. 3. The frequency constituted 300/16 linear pairs per line, creating о normalized frequency (where 400 pairs of lines per line = 1) in 300/16 x 400 or 0.046875. The band filter with 5% lower than this frequency and 15% above was chosen, because the label distortions usually lead to the image shrinking and therefore the frequency increase. Stop bands were created from 15% below the frequency up to 0 and from 25% above the frequency up to 0.5 (Neicvist's limit). The filter coefficients were kept in the IMPULSE file. LUT 77 (fig. 9). For subsequent operations, the first 62 coefficients are omitted, because the filter is symmetrical. Subsequent references can be made to the list of codes initiated from the microfiche with the name of the file FIND.C 76.

The filter of 25 elements of the image in length was made from the selection of the band filter on output intervals corresponding to the horizontal amplification. For example, if the horizontal magnification constitutes 60 image elements per inch, each fifth filter selection will be used (400/80 = 5 image elements). For incomplete stages, linear interpolation of adjacent filter selections is used.

Also, the second two-dimensional filter of 25 x 15 image elements was used. The sizes of the selections for this two-dimensional filter were based on the Euclidean distance of each point from the center of the filter and were subsequently brought to the respective scale for horizontal and vertical amplification. Subsequently for linear intervals linear interpolation is used.

The output signal of the mono-numeric filter rose to the square and was leveled with a recursive low-frequency filter of the first order that provides the exponential window of the process. When the signal from the output of the smoothing filter exceeds the given limit, the optional second step of two-dimensional filtering is used to confirm the existence of the target and to determine its precise position, which is discussed below. In the first part of the two-dimensional filtering, a small filter of 10 image elements per 10 elements was used for the economics of calculations. This filter scans the rectangular region in jural to the position determined by the one-dimensional filter. If the maximum two-dimensional correlation exceeds the given limit, then the final stage of two-dimensional filtering with a complete filter of 25 x 25 image elements is used for о small square window around the maximum. If the best result of this filter exceeds the given limit then it is detected centrally. If none of the limits have been exceeded, the program partially downloads the smoothing filter and returns to one-dimensional scanning. If the one-dimensional scanning is completed without the presence of the target being detected, the program ends with the return of the error.

(b) Normalization of the measured image

The brightness of the reflected light captured by the optical transducer used in the process may vary due to illumination, print density, paper reflectability, camera sensitivity, and other factors, including label destruction, for example, bending, blurring, etc. As an optional but preferable stage, the reflected light measured with the transducer and transmitted to memory can be normalized by the usual procedure. Using о technology known in the art, the NORM.C 78 normalization program shown in fig. 9, was used to analyze the luminance levels of the reflected light from the label, written in blocks of elements of the image in the scanning device to determine the minimum and maximum brightness of the reflected light for the data network. the ordered numerical readout of this scanning device in combination with the image capture board is recorded in the computer for further processing using the normalization program.

Using the equation y = mx + b, where the minimum brightness set in place of x allows the acquisition of a size y = 0, and the maximum brightness, instead of x allows the acquisition of the value y = 63. The brightnesses of the reflected light for each element of the image were adjusted in such a way that the blackest and whiter shades in the given image were taken as standard, and other shades of black, gray and white were brought to this standard. In this way the normalization stage allows the process of processing the measured image to be facilitated. The normalization was carried out using the NORM.C.

(c) Recalculation of the image

For subsequent processes the inscribed image of the label is recalculated to create an image with constant horizontal and vertical magnification. And this time it is an optional step, but it allows for fast and accurate restoring of encoded information. The recalculation was done to assign the image о homogeneous horizontal and vertical separation, for example, the image with the size of 150 elements per sheet, which is shown in the static example with the focal point fixed by the invention.

The recalculation occurs due to the calculation of the fractional addresses of the rows and columns on 1/150 inches, based on the known horizontal and vertical amplification. Each point on the recalculated new homogeneous image is further highlighted by the respective point of the repeated image in memory. To approximate the size of the points in the fractional addresses, linear interpolation is used. Due to the recalculation the center of the label is stored in memory in a known position. The recalculated image is stored for later use at the search stage. It is subsequently used at all stages of the process. The recalculated image of the label is centered in a known position on the network, but it should be noted that this does not indicate the orientation of the label which may be asymmetrical to the translator.

(d) Two-dimensional synchronous restoration

All the stages of the process that follow are called two-dimensional synchronous restoration. The steps are performed with the help of the program and the subprograms registered with the name CLOCK.C 79, shown in fig. 9. This operation is performed in two dimensions on the recalculated image to determine the exact position of the hexagon on the primary data network. The purpose of the synchronous restoration is to determine the places of selection and to correct the effects of sliding, bending or tilting the label, because the label cannot be ideally flat. This is an important step in the process and is not limited to labels encoded with hexagons. О such an operation can also be used in other processes for decoding labels including regular two-dimensional networks such as squares, triangles, etc.

One-dimensional synchronous restoration is a general concept that is well known in the field of signal processing. Synchronous two-dimensional restoration is a continuation of this process and will become clear to qualified specialists as a result of a certain intellectual effort. It should be mentioned that the term synchronous restoration generates non-experts because in fact the process does not refer to synchronization.

(i) Border amplification and nonlinear processing

The first step in synchronous restoration can be performed with the help of different nonlinear mapping operations, known in the art when creating the signal components at the predetermined synchronization frequency, which are omitted at the quantized output of the image from the optical transducer and the image capture board.

The purpose of the nonlinear map is to take (preferably) the normalized and recalculated image that exists at the given point during the process and to create a given two-dimensional nonlinear map, which amplifies the passage between two adjacent contrasting hexagons. In the preferred embodiment of the invention this is done through standard mapping. This step can be performed when filtering with an image differentiation nucleus, for which some methods are known in the art, for example, Laplase and Staves nuclei, after which the absolute size is determined or the results are squared.

When mapping the standard deviations the image with the undifferentiated cell-cell edges is stored. Then the standard deviation card is created to determine the locations of the adjacent contrasting hexagons due to determining the standard deviations of the entire 3 x 3 groups of image elements (different from the sets of 3x3 cells) for determining the standard deviations of the brightness of the image elements. Calculations of standard deviations are made to determine the regions of the image elements having fixed coupling (the smallest standard deviations), which represents the interior of the hexagons or the boundary between two hexagons of the same color, as opposed to the groups of the image elements that have larger standard deviations. , which is a change from a hex color of о to an adjacent hex of another color. Because adjacent hexagons often have the same color, the standard deviation card will not always highlight each hex because the mapping of standard deviations is not possible to highlight the boundaries between adjacent hexagons of the same color, there are omitted borders or edges between hexagons. Other aspects of the process of synchronous restoration are oriented towards the regeneration of these omitted borders.

The decoding process of the present invention can be used for any of the examples described above. The coding blocks of the different geometries can be easily adapted and these encoded polygonal cells can be located with the geometric centers of the adjacent polygonal cells at the ends of a predetermined two-dimensional data network.

When the optical reading labels of the present invention are read with optical translators as described herein, the concrete geometry or the shape of separate coding blocks or polygonal cells is not determined using the optical translator. Instead, the translator simply tests the optical reading tag with a concrete number of samples at each end and records the brightness of the reflected light that corresponds to the optical characteristic of the concrete region being tested, then these sizes are stored in the storage medium to be processed. In other words, the electro-optical transducer allows recording of the average brightness of the region-by-region light, taken as a sample on the entire surface of the label, regardless of whether something has been printed on the label or not. This is and is assumed when recording the image in memory, with the edges unchanged from cell to cell. For this purpose the decoding process is adapted for reading optical read labels with configurations that vary in a wide range, because the centers of the polygonal coding blocks are located in the intervals and directions established on the two-dimensional network.

The preferential variant of the label in which adjacent hexagonal encoding cells are used forms because of this о hexagonal or hexagonal cellular network and this leads to the fact that at the predetermined test point a higher energy signal is inscribed, because it is provided о vignette density of hexagonal coding blocks. In partially tangent networks of polygonal encoding cells or in networks with nonangangular cells, the level of interference caused by the optical characteristics of the intermediate intervals can lead to the decrease of the signal / interference ratio, registered with the electro-optical translator, from which the efficiency of the entire reading system decreases. of labels. Of course, a lot of other circumstances, including dirt, breaks, stains, and additional label slipping at intermediate intervals between coding cells, if the polygonal coding cells are not tangent, can also lead to increased jamming and thus decreased signal restored. .

It has been practically found that the variant of the coding system with hexagonal cells, as in the case of the label examples, which use polygons which are essentially hexagons shown in fig. 14, can only lead to modest distortions and, consequently, to a small decrease in the informational volume of the system. When using polygonal shapes in the case of modest packing characteristics or non-tangible polygons instead of tangent packing, a weaker signal may be obtained, but nonetheless a profitable one for many cases. But at a certain point the signal / jamming ratio of the system, due to a strong dependence on the shape of the polygonal coding cells, the inefficient completion of cells and the predetermined two-dimensional networks that lead to large intermediate intervals between polygons, will fall to a volume not admissible for small storage and restore information.

The acceptability of the system depends on the quality of the signal restored with the help of the electro-optical translator from the account of the modification of the measuring system, for example, by increasing the number of samples per о surface unit in the label area, it can be improved the restoration of the signal registered with the translator and also the memorization. information and characteristics of restoring partially tangent and non-tangent configurations of labels. Such adjustments that allow the use of less preferable label configurations can be performed by a specialist in the field of signal processing.

Because of this, the process allows the wide extension of the borders of the signal / jamming ratio. Thus, polygonal cells, whether regular or irregular in shape, can be used as coding units in the optical reading labels of the given invention. In addition, because the intervals and directions of the centers of the polygons are known relative to the adjacent polygonal cells, the polygonal coding cells can be located on the established network! and not on the hexagonal lattice !, and the polygons can be located tangentially, partially tangentially or even non-tangentially on the optical reading tag!

After! as will be explained in more detail below, the nonlinear mapping technology, in particular, the standard deviation mapping technology, described here in connection with the preferential variant!, facilitates the reconstruction of omitted crossings or borders between polygonal cells with similar optical characteristics. Moreover, one and the same feature can allow! to overcome the loss of crossings between polygons with similar optical characteristics. This is the situation, when the configuration of labels containing partially tangent or non-tangent polygons is used in practice! This! particularity is carried out! with the help of fast Fourier transform, fast Fourier reverse inversion and transformation.

Thanks! using the optional technology in the preferential version! of the invention, the amount of calculations necessary to create the standard deviation card can be reduced. Normally for calculating only nine! image elements in a block of 3 x 3 image elements should be used 8 operated! of assembly. They can be reduced by two! times, replacing each element of the image with its sum and the elements of the image on the left and right. Two assemblies will be required for each element of the image. Then the same operation is performed! on the new image, omitting the calculated amount for the upstream and downstream image elements. For this, they are needed! two meetings, the total number of meetings reaching four. It can be shown that after these stages each element of the image is replaced by the sum of itself and 8 neighboring neighboring elements.

Standard deviation mapping is a mandatory technology for creating the hexagon map corresponding to the initial data network, but only with the omissions of the crossings between the initial hexagons of the same color.

(ii) The frame

The next program entitled Framing is not mandatory. The framing was used in the practice of the invention to reduce the brightness of the borders that are not related to the outlines of the hexagons. These boundaries appear in two! points: the target rings and the uncontrolled image surrounding the label. The suspension function is used to reduce the brightness of these regions. The details of using the quadrature as a precursor for rapid transformation! Fourier is the competence of the specialists.

(iii) Two-dimensional transformation fast! Fourier

Two-dimensional transformation fast! Fourier of the numerical sizes that correspond (optional) to the standard deviation square card is executed! under the control of commercially accessible programs. During work the computer performs! fast transformation! Fourier of the image created at the previous stage for elaborating the two-dimensional arrangement of the orientation and brightness of the border crossing between the contrasting hexagons identified at the standard deviation mapping stage. In other words, fast transformation! Fourier is a measure of the distribution of orientations and brightness of images between hexagons, when known. In this way the regular distribution and orientation of the boundaries of the hexagons allow to assign certain points in the transformation region to a high energy level. The brightest point will be point 0, 0 in the transformation plane !!, which in the given invention! corresponds to the DC component. Six points surrounding the center point represent the distribution, orientation and brightness of the edges between the hexagons.

To a specialist in the given field it will be clear that in terms of hexagons, the two-dimensional presentation! of the range, orientation and brightness of the separation boundaries between the contrasting polygons, discovered at the previous stage! The standard deviation mapping can be calculated based on the fast Fourier transform of the numeric data corresponding to the measured image of the label. Thus, the range and orientation of the boundaries of the polygon will lead to the fact that some points in the transformation domain will have high energy. The number of high energy points surrounding the center point in the 0, 0 coordinates of the transformation plane will depend on the geometry of the concrete polygonal coding cell used to create the optical read tag. As for hexagons, such points surrounding the center point will exhibit dispersion, orientation and brightness of the edges between polygons or the edges between polygons and intermediate intervals, if the label configuration is partially tangent or non-tangent.

Because the image is a real (and not a complex) image, the field of transformation is a point symmetrical to the coordinate center. In this way, only half of the transformation field needs to be calculated, which saves the computer's working time in half. Eliminating these calculations allows to reduce the number of attempts necessary for the subsequent filtering of the image and the inverse steps of the fast Fourier transform. The fast Fourier transform program used in the illustrative example of the static system with fixed focus was the accessible program R2DFFT in the 87 "FFT-2" software package of Microway Inc.

(iv) Image filtering

Next, the filtering process is necessary to reconstruct the complete scheme of the hexagons in the image field, using the transformed numerical data. This can be done by removing any points in the transformation domain that do not correspond to the given distribution and orientation of the hexagons boundaries identified at the standard deviation mapping step. Six special points in the transformation domain arise because of the label's hexagonal cellular construction. In the field of transformation only three points are actually identified, because the image is symmetrical on points with respect to the coordinate center and the following three points can be assumed from the first three. In the preferred embodiment, the filtration is done in three stages to eliminate the passages from the mapping step of the standard deviations that are too far, too close and / or in irregular directions.

First of all, high frequency filtering is performed on account of the cancellation of all points within the circumference given around the coordinate center of the transformation domain, but at a certain external distance from the coordinate center, insufficient for the six special points in the form of hexagon in the graphical region of transformations. These points correspond to intervals greater than the intervals between the hexagons and because of this they carry information about the omitted passes in the label image. In order to create the omitted passes in the label image, it is necessary to delete the information about the omitted passes in the Fourier transform field.

After that, all the outer points of a given radius, with the exception of the six special points in the transformation field, are canceled. This corresponds to false passages that are too close. This operation is combined with the first to form a ring from the remaining points. Creating this ring is equivalent to spatial bandwidth filtering. The inner and outer radii of the ring are determined by the dispersion of the outlines of the hexagons. Because the diameter of the hexagon must hypothetically contain 5 image elements in the given example, and for the length of transformation into 256 image elements, the peaks of the hexagon in the transformation domain will be 265/5 = 51.2 image elements from the center. Also suitable was the ring with the intimate radius of 45 elements of the image and the extreme radius of 80 elements of the image, which corresponds to the diameters of the hexagons from 3.5 to 5.69 image elements. A filter to facilitate the passage of frequencies of higher order was also used for the deformations of the label of the type of sliding and tilting, which causes the image to decrease.

After performing the special band filtering described above, there is a ring with 6 special points, so that each point has an angle equal to the center (point 0, 0) of the transformation domain. To finish removing unnecessary information, the directed filtering step is used. Any point in the transformation field is canceled. This causes the image field to eliminate any edges that are not in one of the three directions dictated by the hexagonal cell structure.

In order to perform the directed filtering, it is necessary to find the most important points that remain after the spatial band filtering. Preferably this point is one of the six special points of the transformation domain, similar to the peaks of the hexagons. Other special points in the same radius from the center and with the angular position of the 60 ° factors are also evident in the transformation field. Therefore any points with an angular distance greater than 10 ° from each of these points are canceled. Six rows of the ring remain. Due to directed filtering any incorrect information about the location or directions of the image region is removed.

Removing this incorrectly located information allows the complete drawing of each hexagon in the transformation domain to be restored.

The previous examination of the filtering scheme, used in the preferential variant of the label containing tangent hexagons, required modifications in the case of using the two-dimensional networks established for the optical reading label. However, it should be noted that only a small change in the filtering scheme is necessary for a person skilled in the art to adapt to different configurations of the label mentioned above and which are shown in the attached figures.

As the individual coding cells are only beginning to be determined, it is determined that their respective boundaries will have certain angular locations and a set number of sides with the given length. Next, it is necessary to determine the proportion of adjacent polygons, for example, whether they are tangent, partially tangent or non-tangent. It is also necessary to restore the geometric network on which the geometric centers of the polygons will be located. Because the geometry of the label is determined, a specialist in the given field will be able to restore the respective filtering scheme for filtering the energy points in the transformation field so that only the brightest points that correspond to the required distribution and the direction of the polygon boundaries are processed at the fast Fourier transform with a subprogram.

Regarding the filters actually created, it is necessary to understand that the creation of a properly measured band space filter is required, based on the established diameters of the encoding polygonal cells. Subsequently it is preferable to create a filter directed to filter the energy points, different from the special points that correspond to the ends of the polygonal coding cells, from this account any information about the incorrect distribution or the incorrect direction of the polygonal coding cells in the field of image and intermediate intervals is removed. , if any. Due to the removal of such incorrect information, a complete picture of the polygonal encoding cells in the image field is created. Further, the numerical data is ready for the fast inverse Fourier transform, according to the steps described below.

(v) Fast inverse Fourier transform

For the return to the image region, restoring with it the picture of the adjacent hexagons of the data network would be welcome the fast two-dimensional inverse Fourier transform (2D-IFFT) on the filtered data of the transformation domain. The inverse transformation is performed using the standard two-dimensional inverse Fourier transform subprogram (R2DIFT) found in Microway Inc.'s 87 FFT-2 ​​package. At the end of the inverse transformation step, each hexagon is restored to the image field. In the new image the centers of the hexagons have high values. The actual value of the stains in the centers of the hexagons depends on the number of edges in the vicinity. О greater quantity of curbs leads to the generation of a higher energy on the allowed frequencies and, therefore, to high value points. Fewer edges lead to lower points. The value of the points represents the measure of the mandatory levels at the synchronous restoration at any point.

(e) Determination of the main axis

The image of the hexagon was thus restored, but its orientation still needs to be determined.

The cellular structure of the invention has three axes located at an angle of 60 ° to each other. The direction of these axes is determined by the brightest point in the transformation field after the spatial band filtering. Now we can determine which of these three axes is main. This step is not required. If this step is not performed, the label must be decoded three times using each of three axes, but only the о axis allows the admission of an important message. The main axis is arbitrarily chosen as о axis which is parallel to two parts of the label, as described above and as shown in fig. 2.

If the boundaries of the square label are determined by the main axis, then most of the energy in the restored hexagonal structure will flow within the boundaries of this square.

To determine the main axis, each of the three axes is assumed to be the main one. The subsequent design of the square label is determined for each axis, and the total energy of the restored structure, which is interior on this square, is determined from the numerical energy data of the inverse transformation subprogram. Correct testing is characterized by maximum energy. After this the angle of the main axis is stored for the initiation stage and other search operations. But it is still unclear whether the stored angle is in the correct direction or below 180 ° from the correct directions. It should be noted that it is not necessary to determine in general all three areas of the label, because the energy in the common areas of the three squares must not be determined.

(f) The search

The program named SEARCH.C 80, noted in fig. 9, combines the regenerated and transformed information about the center with the levels of brightness of the initial image in such a way that one can determine the level of the gray color of each hexagon. The search is done in such a way as to minimize the possibilities of losses. The end result is a matrix of gray color levels for each hexagon in the data repository. During the execution of the first part of the SEARCH.C program, 4 important informational relays are created. The CVAL chisel (fixed size) retains the measure of the reset synchronization signal for each hexagon while the GVAL chisel allows storing the sizes (0-63) of the gray levels in the center of each hexagon. The remaining IVAL and JVAL relays allow the storage of the distribution of the center row and column of each hexagon.

(i) Initiation stages

Based on the angle of the main axis determined in step (e) and the known distribution of hexagons (5 image elements) for example, the likely horizontal and vertical displacements from the center of a hexagon to the centers of six adjacent hexagons can be calculated using the computer.

After these calculations, the SEARCH.C program acts on the restore signal received from the memory and the recalculated image of the label, also received from the memory. The main purpose of the initialization subprogram is the merging and condensation of the information from these two sources and the generation of the data matrix that ensures the existence of the gray color scale value for each hexagon.

The search initiation step is bounded by a square in the center of the label with the size of the side of l / 3 fol. This initial point within the limits of this area is the point with the maximum size that is on the set bench! of the signal. After this, the position of the initial point in front of the center of the label is determined. This is the point where the signal is strong and distinct and is located near the center of the label. A strong signal guarantees that the search operation starts from the efficient center of the hexagon and it is desirable that the point be close to the center of the label, so that its absolute position can be determined! lighthouse! seriously influence inclination or inclination. The value of the point quality in the synchronization structure restored! is the value of the point plus the value of eight neighboring points. The initial coordinates of the initial point are transformed into form! polar !, and polar coordinates are set faf! by an angle of the main axis determined previously and this result is again transformed into form! rectangular. These coordinates are brought to scale! in accordance with the distribution of the rows (4.5 elements of the image) and the distribution of the columns (5 elements of the image) from the input matrix of the hexagon. The quality of the timing, the level of the gray color and the distribution corresponding to the initial hexagon are subsequently introduced in the respective CVAL, GVAL, IVAL and JVAL relays. (ii) The main search cycle

The main search cycle determines! position of the centers of the other hexagons. The cycle is over! when the proposed number of hexagons is found. High security! of the faf coding process! effective damage! comes from concrete technology! search described! lower.

Each iteration of the search cycle begins with the selection of the position of the synchronization point restored by the largest size, whose neighbors were not found because of the potential sizes. From this point the search will be continued! to a hexagon away in each of the 6 directions. The effect is constant! in bringing the search strategy along the trajectory to the best quality! to a worse one of the restored synchronization. That way, if! I exist! a weak domain of restored synchronization, for example, in the center of the tag or in a deleted domain, the search algorithm omits it and does not go through it. Graphing the omission of these weak surfaces and their subsequent preservation, the possibility of losses in the refea is substantially reduced. Because! the losses are as unwanted as the wrong reading of the gray color level, this one! particularity of the search algorithm is very powerful!

The subprogram is responsible for finding the best neighbors! quality of synchronization in the main cycle. The subprogram is executed 6 times for each of the neighboring hexagons of the given hexagon. The beginning is calculated! neighbor's position. If! this neighbor is off the labels in labels, the cyclical iteration stops. If! no, the neighbor is in control! if! not found from another! direcfie. Cyclic Iterafia! it will end if! the neighbor will be looking for you! the algorithm assumes earlier searches that are safer than subsequent ones. If! the neighbor is out of this test, it's calculated! positive would presuppose! of the neighbor center in the synchronization structure restored! At this moment it is being done! the gradient search for the maximum synchronization signal. There are 8 elements of the image surrounding the restored position! to settle down! if! a larger sync size was found. This! Gradient search allows for a level of adaptation that is strictly necessary in case of bent or inclined labels. After this the subprogram goes to the next neighbor and ends when all the neighbors are controlled.

As shown above in step (d), as a result of the data transformation processes, the restored image of the label now bears the information referring to the geometric centers of the polygonal coding cells. The polygons where several edges have been detected, ie crossings, will have more energy in the centers. The centers will be located on the default two-dimensional network, which has a certain number of axes located equidistantly or non-equidistantly, depending on the circumstances. The information referring to the spatial ratio of the axes of the predetermined two-dimensional network can be used at the orientation of the main axis.

It should be noted that the algorithm can be modified accordingly so that the decoding process determines the real geometry of the two-dimensional network and from this results the definition in the direction of determining the filtering scheme of the so-called main axis of the label (that is, the axis of the two-dimensional network, which is parallel on both sides of the square tag with optical strips, described here) and provides the necessary coordinates for the search subprogram.

Whether the geometry of the label with a step, optional, such as the one described above, was determined, or was simply introduced in the decoding process through respective modifications of the two-dimensional process of synchronization of the restoration, the set of configurations of the The label, described and shown here, can be easily adapted by a specialist in the given field. It should be noted that the number of axes on which the centers of single coding polygonal cells are located and their respective orientation can be introduced at the stage of determining the main axis for all three axes of the hexagonal network of the preferential variant. For this reason the main axis of the given two-dimensional network can be determined without carrying out the control and analysis of the errors described above in step (s).

Regarding the hexagonal network of the preferential variant, the information from the main axis determination stage and the known distribution of the polygons can be used to calculate the horizontal and vertical displacements evaluated from the center of a polygon to the centers of the surrounding polygons. After these calculations and after making the necessary adjustments of the sorting subprograms, the sorting (including the initiation stage and the main sorting cycle stage) can be performed for the specific configuration of the label, used in the current case. It is worth mentioning that such non-essential adjustments of the SEARCH.C 80 sorting program in the annexed list of initial codes is within the competence of a middle-class specialist in the given field.

When the work completion is completed, the subprogram marks the current position of the control panel so that we do not search for it again. The effect is that this position is omitted as a candidate, whose neighbors have been examined. For each iteration of the cycle, from 0 to 6 new candidates are added and one is excluded. With good means can be used the structure of the data in which the candidates are kept in the order of sizes during the entry and removal operations. One such layer is called priority rendering. It is known that for the linear selection algorithm approximately n is required² operations, while for a well-organized priority rendering using a balanced tree or о unordered structure, approximately n log (n) operations are required. Alternatively, the n-order selection algorithm based on group sorting can be used if the restored synchronization sizes are scaled and scaled down to a small domain of integers.

(g) Creation of histogram and level indication

At the end of the main selection cycle, the position of the centers of all hexagons is determined and the gray colors of all the inscribed hexagons are filled. The next step consists of the numerical limitation of the numerical sizes of the gray color level in the range 0-63 of discrete levels, for example, with black, gray and white (for the black, white and gray label). This is done by constructing the histogram of the brightness sizes of the label image from the central hexagon. Energy levels can be determined by looking for the collapses on the histogram.

(h) Approximate network correction and definitive orientation

E> up limiting discrete levels, there can be two types of distortion. First of all, the network may be outside the plant. This can happen if at the initial selection stage the position of the highest quality synchronization signal relative to the label center is not correctly determined. The second possibility was that the entire label was read efficiently from top to bottom, because the angle of the main axis has a 180 ° ambiguity.

The subprogram on the microfiche contains the function of determining the displacement of the label towards the center. If the label is located correctly, the coordinates of the center row must pass through the label center. In order to determine if a vertical arrangement error was made, the rows above the hypothetical central row are checked, to determine which of them forms a line that runs closer to the center of the label. If the row above or below is closer than the hypothetical central row, then it is done! the corresponding movement up or down. If! the left-hand straightening of the short lines has been made! wrong, this is fine! from the account of moving the short rows with о position to the right.

Mistakes of horizontal arrangement! and the stale reading is controlled! with the use of the information entered in the label! and known as approximate information of the relay. The information is arranged! in sets of 3x3 hexagonal cells as described above. Because! the label can be, for example, with a network of 33 rows on 30 columns, these sets form! о 11 x IO network. The bottom central hexagon of each complete 3 x 3 cell set has a unique property, created during encoding. I exist! о guarantee of transmitting it from any direction, about what was mentioned earlier in relation to fig. 4. For example, if! the bottom center hex is black, then the bottom left and right hexagons must be black! either gray or white. The sub-program on the microfiche uses this priority of transmission capacity for the final elimination of two! possible distortion. For the beginning it is created! о network where every element of the indie relay! if! there was a switch between two! adjacent horizontal hexagons. After! this network is controlled! for each of the 9 hypothetical slips of the approximate relay, located in the 3 x 3 structure around the supposed sliding 0. One of these slides will look the best! tuning between the actual and the supposed passes and this slip position is maintained. After! this is controlled! the same hypothesis in the assumption c! label was reversed. This will happen if! the angle of the main axis is oriented from right to left side! by the printed label and not from left to right.

If! the label was simply overwritten, meaning the top rows changed to the bottom ones, and the top columns to the bottom columns then reversed! also the results of the slides. But for о proper restoration! the label is required! о significant transformation. While reading the short lines (length 29) it is checked! the ones on the left, so when the label is broken they should be checked correctly. It's done! adjustment and this is the procedure! that allows s! the results of the slip hypothesis are created not only as a simple restraint. In fact, the best adjusted slip test labels will be better than the previous tests, if! the label has indeed been read reluctantly.

Determining if! the label was really read and yacked! is о slip in the absolute position, the label matrix can be encoded. With the right determination! the image and the slide are over! image processing function and data decoding processes begin.

4. decoding

The RD LABEL.C subprogram shown in fig. 8 allows reading the file created by the selection program and creates a file with о consecutive bits which in the preferred version const! of 1292 bits. At the same time! use the subprogram CELL DEC.C (fig. 8) for masking the unused hexagons and for using the decoding program which is the reverse of the coding program.

The first step in the decoding process is to create a succession of bits from the hexagons information using the hexagon-bit mapping process, which is in line with the bit-hexagon mapping process used in the encoding operation. Subsequently the informational succession! it is deducted! of the program in a sequence of priority bits and consecutive sequence of simple bits! or as many bits in a row as they are used when coding the tag.

After! this is necessary! the use of error correction for each successive bit with the use of error coding technology, which was used in the process of label coding. For example, if the Reed-Solomon encoding is used to correct the succession of bits created by the selection program, it is generated! an output signal that exists! in the same format as described above for encoding the input file. The error correction can be performed! in the following sequence:

1. It is calculated! syndrome.

2. It is calculated! The error search polynomial, using the Berlekamp-Massey algorithm.

3. It is calculated! position of the error using the sorting algorithm Chien.

4. It is calculated! error size, using the Forney algorithm.

The last step is being taken! only if! errors were detected in steps 2 and 3. It is also calculated! the amount of errors detected; if! an incorruptible number of errors is detected or if! the error is located! in the important field, the fact is noted. Specific procedure! error coding used! in this example it is noted! through ERRDEC.C in fig. 8.

5. Conclusion

When transporting the luggage (when identifying its place on the conveyor) the priority message indicating the postcode of the place of destination can be used to include the appropriate steering cranes or conveyors for sending the luggage with the respective car, aircraft or luggage car, which will transfer port of destination.

Although the invention can be used in conveyor systems with rejection, obviously, it can be used in a wide range of operations in collecting information, processing luggage and in technologies where it is preferable to decode the label on luggage, letters, parts of cars. or from analog devices and forces the system to perform baggage processing or о production operation, for example, on the labeled object. The invention allows these operations to be performed quickly and accurately, which is possible due to a large amount of information on the label and even to protect о a significant amount of information from the disappearance caused by the rupture of the label and other damage.

As shown in FIG. 8, TEXTOUT.C software can be used to periodically play the decoded message on the computer terminal.