Audio/video communications processor.

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Industrial application field]

The present invention relates to an audio / video communication processor system, and particularly for an audio / video communication processor for the remote communication and computer fields, a group of workstations having means for communicating audio and video and allocating users. It is suitable for application to useful systems.

【0002】

[Conventional technology]

First, the terms used in the specification are defined.

【0003】

"Weighting" is a method in fixed binary notation in which software identifies a group of binary numbers in a specific place, that is, in the hundreds or thousands of decimals.

【0004】

"Compression" is a method of using software or hardware to generate code that controls the replacement of a long string of binary data with a replacement code that is one digit or two digits smaller. This code may be in modulation format or binary data format.

【0005】

"Pel" is a term used to describe the elements of a video signal that are currently digital. Binary numbers represent the range of luminosity, chromaticity and hue in digital imaging systems. The smallest independent unit of visual acuity was displayed in binary on the display screen. If written without abbreviation, it is Picture ELement (pixel), and PEL is an abbreviation for this.

【0006】

"Pixel" is another way of saying Pell.

【0007】

"Communication processor" is a term used to describe a new class of communication equipment. Using this new class of communication equipment, workstations equipped with resources for users are connected to video conferencing via remote communication networks. This new class of equipment consists of several subprocessors, and these subprocessors are assembled to form a communication processor. The concatenated form is the communication processor.

【0008】

A "packet" is a collection of binary data on the network side of a communication processor. The network side of the communication processor includes network and communication processor control information, video data, and audio data.

【0009】

"DACS" is an abbreviation for Digital Access Cross Connect, which is either DS format or non-DS format between several T1 carriers regardless of the signal code of DS carrier or T1 carrier. It is a machine that exchanges slots for 64,000 [bps] hours. DS refers to the digital telephone channel coding of DS1, DS2 or DS3. If the base speed is 64,000 [bps], it may be a non-telephone channel that can contain any form of binary data. T1 refers to the digital multiplexing basic speed of 1.544 [Mbps], which may or may not accommodate 24 DS digital telephone channels.

【0010】

"TASI" stands for Time Assigned Speech Interpolation and is a term for compressed speech. Audio, or proper digital audio, is processed according to either mu-weighting or a-weighting. One is the North American standard and the other is the European standard. Other countries use either North American, European or CCITT standards. On international lines, it is necessary to convert between the two when they are not compatible. The TASI algorithm reads the data and removes the data that represents silence, so only the sound of the actual language is transmitted. In addition, TASI may insert a code for audio data in a well-known fixed pattern. This audio data is newly compressed. The remaining bandwidth saved in this way is used for other purposes.

【0011】

"Artificial intelligence (AI)" is a software program that responds to an event and generates an event based on a group of rules. This rule does not necessarily have to be strict. Usually these rules are flexible rules such as instructions given to beginners. A crucial property of these rules is that they cover all possible and even unlikely events. These rules are a permutation of all possible combinations of events. The purpose of formulating these rules is not to allow situations with inappropriate or impossible responses to external stimuli.

【0012】

"Independent synchronization" is a characteristic of two or more black circuits, and even if the synchronization process is executed once and canceled, these black circuits remain synchronized for a long period of time. However, the minimum acceptable time period is 24 hours.

【0013】

The conventional technique will be described below.

【0014】

As a prior art in the field of the present invention, some patents will be mentioned, followed by a description relating to a product having some elements. Some of these elements may be directly related to products made using the present invention. U.S. Pat. No. 4,949,169, granted August 14, 1990, describes an interface architecture that interconnects multiple video display devices together via high-speed digital communication links with limited bandwidth. There is. The interface architecture at each display node specifies that each node displaying an enlarged or reduced video window transmits sequential pixels consisting of data composed of separated Y-fields and C-fields from a digital TV source. To do. Audio information is transmitted with the video over a portion of the network bandwidth that is not used in the video. An object of the present invention is the need to newly add control hardware and control software by providing a hardware system in which existing hardware can be used in various video display devices and communication adapters coupled thereto. To minimize.

【0015】

U.S. Pat. No. 4,780,761, granted October 25, 1988, states that human vision systems are less sensitive to diagonally oriented spatial frequencies than to horizontal or vertical frequencies. It relates to a recognition device. This transceiver has a method of quantizing the conversion factor according to the model of the human vision system. This system is not designed for video conferencing workstations with telecommunications networks. This system does not time-divide the video subprocessor as part of the video / audio communication processor, and does not function with the network control method, so it cannot be converted to network.

【0016】

U.S. Pat. No. 4,494,144, granted January 15, 1985, describes good video transmission with reduced bandwidth. Bandwidth is reduced by dividing the camera's image pictogram into segments by determining the activity level within each segment and transmitting the signal for each segment at the resolution level associated with the activity level within the segment. To. The most active segments are transmitted at the highest resolution, while the other segments are transmitted at lower resolutions. This system is not designed for video conferencing workstations with telecommunications networks. This system does not time-divide the video subprocessor as part of the audio / video communication processor, so it cannot be networked.

【0017】

U.S. Pat. No. 4,862,264, granted August 29, 1989, subdivides one frame of pictorial information into one set of constituent blocks, measures the total number of active pictorials in each block, and acts in that block. Limited bandwidth by sampling the patent information for each block at a rate related to the total number of typical patents, and adding an auxiliary signal to the coded block to indicate the rate used for that block. Explains how to code a video signal for transmission in. A decision is made for each block whether to transmit completely accurately or to reconstruct the frame from the previous frame. In fact, each block is sampled twice at the same time. The first sampling is sampled at the subrate and the second sampling is sampled at the Nike straight. A circuit that activates the block and a circuit that activates the movement are used to determine whether the transmission is high-precision transmission or low-precision transmission. These samples may be transmitted in analog or digital format.

【0018】

U.S. Pat. No. 4,654,484, granted March 31, 1987, describes an improved device for fast compression, decompression, and display of wideband information transmitted by narrowband communications channels. Video footage is periodically collected in low and high resolution phases from digitalized data that represents the intensity of the gray level of individual pixels. The digitalized data is already grouped into pixels. During the initial period of the low resolution phase, a representative sample of cell luminance values is transmitted from the transmit station to the receive station according to a video compression routine. Next, the receiving station uses a video expansion routine to calculate the luminance value of the pixel for which the luminance value has not been transmitted and display the first image.

【0019】

U.S. Pat. No. 4,682,225, granted July 21, 1987, describes optimal bandwidth compression methods and equipment for telemetry engineering. The optimal sampler from the video signal produces a series of sampled fields. Each field and its range rate information are sequentially transmitted to and stored in the multiple optimum field storage means. This patented device may be used in a spacecraft docking system. In this case, a huge amount of video information and video data must be transmitted in a limited finite bandwidth. The present invention is suitable for space communication systems from spacecraft that have both video and data signals. In particular, the manual signal can control parameters such as range rate, sampling rate, etc., the number of low resolution frames of the video displayed at the same time or data and a part of the downlink communication bandwidth allocated between the videos. ..

【0020】

Although this patent has little to do with the preferred application of the present invention, it should be noted that this high performance system may be significantly improved in the present invention. During the video conference, the user may request reinitialization at any time. The communication processor of the present invention also recognizes a decrease in data transmission and automatic reinitialization. The communication processor of the present invention controls the reinitialization so as not to interrupt the allocation of the existing channel bandwidth between users.

【0021】

U.S. Pat. No. 4,739,413, granted April 19, 1988, describes a method for video-optimized modulators and demodulators with adjacent modulation amplitudes that match the gray values of adjacent pixels. Each modulation symbol corresponds to a specific pixel brightness value. U.S. Patent No. 1 granted on March 5, 1974 Nos. 3,795,763 describe a method for digital television transmission systems that transmit at significantly reduced bit rates and bandwidth. Frequency interleaving technology reduces the sampling rate, and digital differential PCM with edge recording technology reduces the number of bits per sample. The bit rate is further reduced by removing about half of the color data and all synchronous pulses from the transferred data. The synchronization information can be reconstructed by periodically transmitting the synchronization word. The transmitted bit is multiplexed according to a specific format in which the luminance line and the chromaticity line are properly matched on the receiving side. Y and C are separated and sampled below the Nike straight. This sample is quantized and converted into a different sample with reduced bits. Audio is sampled at a horizontal scan rate, and digital displays of audio and video are sequentially multiplexed into the output stream. All other pairs of C are completely removed from the multiplexed serial bit stream, but reconstructed from the adjacent C information on the receiving side.

【0022】

US Pat. No. 5,043,810, granted August 27, 1991, describes an improved way of transmitting or recording improved video signal processing equipment and improved television signal receivers. This processing method consistently controls the selections associated with parts of spatially and / or temporally adjacent images spatially and / or temporally. The determination process includes adjacent parts of this image, which may or may not affect the parts being processed. Pixels of one block construct a part of the image and are sampled according to a sampling pattern that does not correspond to the sampling operation, and every block is adjacent to the sampling pattern that corresponds to the sampling operation, that is, the block that is sampled by the corresponding sampling pattern. ..

【0023】

U.S. Pat. No. 4,720,745, granted January 19, 1988, describes methods and devices for improving video display. The NTSC composite video signal is divided into matrices, and its RGB components are digitized into a 512 x 512 frame pixel array. One high resolution frame is generated from each input frame. The sub-pixel value of a given pixel is obtained by inspecting the closest adjacent pixel and using an enhancement algorithm displayed by the data in the Ruskup table. The signal-to-noise ratio changes the value of a pixel based on the value of the closest adjacent pixel, or the value of a pixel based on the value of this closest adjacent pixel to this value and the value of its adjacent pixels. It is processed by comparing and determining to replace the median value of.

【0024】

U.S. Pat. No. 4,858,026, granted August 15, 1989, describes how to encode the image to be displayed. The image is encoded using data compression, which consists of the first acquired pixel information, as the first high resolution matrix. The second matrix with a lower resolution is created via a low-pass filter like the first high-resolution matrix. The third matrix is the difference between these two matrices. The fourth matrix is generated by further sampling the second matrix (not all matrices are used). The third and fourth matrices are encoded. Complementary decorating returns the second matrix to its original state, and the second matrix that has been returned to its original state is combined with the decoded fourth matrisk by interpolation filtering, and this original state is used. Combine the returned second matrix with the decoded third matrix. Although this method has applications such as compact disk image encoding, the communication processor of the present invention does not work according to the principles described in this patent.

【0025】

U.S. Pat. No. 4,733,299, granted March 22, 1988, describes how to convert an interlaced scan video signal into a non-interlaced scan video signal. This applicant has learned that optimal processing of videos is not required. The low resolution information comes from the currently interlaced field, and the rest of the detailed information comes from the stored signal containing the previous field. Only this detailed signal is obtained from the previous field, and human vision is not as fast as high spatial frequency identification motions, so only a few can be identified.

【0026】

U.S. Pat. No. 4,551,755, granted November 5, 1985, describes a method for bandwidth correction systems for television Jyonchuna. The bandwidth control voltage applied to the bandwidth throttling circuit provides a pass region determined by the relative levels of the video and audio intermediate frequency signals. This patent is believed to mean that the transmission process can be overmodulated, i.e., beyond the allocated band, and the correction signal allows the receiver to recognize and adjust for this condition.

【0027】

The communications processor of the present invention has applications, processes and methods different from those described in US Pat. No. 4,551,755. However, the concepts of bandwidth correction and bandwidth allocation must be distinguished. There is no need to allocate bandwidth from video to audio or audio to video, and when allocating bandwidth, the goal is not to correct malfunctions such as overmodulation, but to manage bandwidth in both formats. is there.

【0028】

U.S. Pat. No. 4,792,993, granted December 20, 1988, describes an improved method of TVRO (Television Receive Only, usually referring to satellite TV).

【0029】

This improvement is to automatically filter the audio signal over the frequency range of the outer band of the modulated video signal and combine the audio and video signals applied to this filter into a limited band of such signals. It is transmitted by a width channel. In remote areas, the reverse process is performed to separate these signals into audio and video.

【0030】

US Pat. No. 4,849,811 granted July 18, 1989 simultaneously sends audio and video signals through a standard telephone line or other channel with limited bandwidth to obtain a video image, this video image. To digitize, modulate the signal containing this digitized image, obtain the audio signal, filter this audio signal to a frequency range outside the frequency range of the band of the modulated video signal, and apply this filter. Describes how to combine audio and video signals and transmit these signals over a channel with a limited bandwidth. At the remote edge, the reverse process is performed to separate these signals into audio and video.

【0031】

Although the communication processor of the present invention has some elements in common with this patent that digitize a video image and process it in a digital manner, the communication processor of the present invention is a continuous sequence separated only by a software protocol. This is improved by combining audio and video together in the channel frame. U.S. Pat. No. 4,849,811 separates audio and video into frames with limited bandwidth by separating and modulating audio and video. The communication processor of the present invention does not use a telephone channel. Instead, the communication processor of the present invention uses a digital trunk, the smallest of which is 64,000 [bps] in series. Long-haul carriers can provide this digital trunk as a 64,000 [bps] trunk capacity or leased line, and 64,000 [bps] in series can provide minimal video conferencing services. Best performance for high bandwidth network carriers.

【0032】

U.S. Pat. No. 4,425,642, granted January 10, 1984, describes a method for common channel communication systems. This common channel communication system can simultaneously transmit a digital data signal in a communication medium signal such as a telephone voice or a television image. The digital data signal is transformed into a very low multi-frequency signal consisting of fundamental frequencies and fundamental harmonics over the communication bandwidth. Since this digital data signal is modulated by a spectral diffusion method and its energy capacity extends over the entire bandwidth, the signal-to-noise ratio is reduced by adding what appears to be pseudo-noise to the audio or video signal. Decrease. Since the digital data signal also produces pseudo-noise, the pseudo-noise is detected together and removed from the audio or video on the receiving side.

【0033】

The communication processor of the present invention does not perform spectral diffusion modulation and does not perform common channel data with audio or video. Communication processor of the present invention in contrast to straight using a protocol to hold audio, video and data in the column transmission channel.

【0034】

US Pat. No. 3,873,771 granted on March 25, 1975 describes a system that simultaneously transmits video and audio signals over the same transmission line using an FM slow-scan TV, but the audio signal is AM. It is transmitted by a single sideband technology. Both video and audio simultaneously occupy the channel in the separated frequency domain. The communication processor of the present invention does not need to use any analog modulation technology, frequency multiplexing technology or slow scanning FM-TV technology.

【0035】

U.S. Pat. No. 4,797,750, granted January 10, 1989, describes methods and devices for transmitting audio and / or video signals at the same time as transmitting recordable computer-generated displays. The computer may generate a series of codes and obtain an image from this series of codes, and the resolution of this image does not depend on the recording transmission medium. These codes are supplied to the first modem via the RS-232 communication line. The first modem converts these codes into image-generating audio tones. The audio tone is input to the audio channel on the left side of the video recorder. At the same time, the auditory information is picked up by the microphone and input to the audio channel on the right side, and the video camera sends a video signal to the video channel of the video recorder. During playback, the left channel audio is decoded by the first modem and reconverted into a computer-generated display. The communication processor of the present invention does not perform spectral diffusion modulation and does not use an RS-232 communication line. Instead, the present invention uses the modem in a different way.

【0036】

U.S. Pat. No. 4,736,407, granted April 5, 1988, describes a method for voice graphic conferencing systems. In this system, two or more users are either directly connected via a voice band telephone line or through a bridge device via a voice band telephone line. Each user has a personal computer, software and smart modem, cassette player / recorder and speaker horn. These are connected as shown in Figure 1. The smart modem receives the bauded signal, immediately decodes it and sends it to the computer so that the user can hear the audio through the speakerphone while under the control of the software. The cassette trecoder / player is used for unmanned operation.

【0037】

The communication processor of the present invention differs from this patent in that the video conferencing conducted by this patent feels more like computer data and audio being discussed.

【0038】

U.S. Pat. No. 4,955,048, granted September 4, 1990, describes how to multiplex the transmission of video and audio signals, separating the video signals into luminance (Y) and color (C) signals. Will be done. The Y signal is then modulated and the C signal is modulated in equilibrium using a low frequency carrier. This synthetic C-modulated signal is converted to a low frequency. Audio signals, frequency-modulated Y signals and frequency-converted C signals are increased by frequency division in order to transmit these signals across the telephone cable.

【0039】

U.S. Pat. No. 4,999,831, granted March 12, 1991, describes how wideband video, narrowband audio, and digital information can be digitally transmitted via an information network. This patent describes synchronously quantized sub-carrier multiplexing, electronically delivering audio, data and multiple channel full bandwidth NTSC video for digital transmission through communication lines and recovery processing. It is to multiplex. The channels to be multiplexed must be carefully selected in frequency bands so that they do not interfere with each other, then these signals are passed through a lowpass filter and the local reference signal (Double Side Band Suppressed). It is modulated using Carrier, DSBSC), resulting in the formation of baseband, middleband and highband channels. These channels are combined and input to the D / A converter. The result is a series-bit stream known as quantization-SCM. This patent is different from the present invention.

【0040】

US Pat. No. 5,027,400, granted January 25, 1991, describes a multimedia bidirectional broadcast system. The main controller receives information from the subscriber terminal via the network. The software of the main controller provides this service by decoding the request from the subscriber terminal and controlling all transmission processes. This service may be a video or commercial database. Transmission also includes bandwidth compression of the video signal. This system is described as providing a wideband ISDN broadcast system, providing CATV with a means of using a limited number of cable channels.

【0041】

The communication processor of the present invention differs from this patent in that this patent is designed to be interactively advertised on an ISDN wideband network, in which shoppers can view video images of the product. You can execute a transaction by seeing and interacting with the calling terminal. The patent is also applicable to CATV systems, where customers can selectively order video programming services over expert wiring, rather than customer-requested programming services. The communication processor of the present invention intends to hold a video conference on a workstation.

【0042】

U.S. Patent No. 1 granted on September 10, 1985 Nos. 4,541,008 describe a television signal transmission system that integrates circuits that process and encode iterative reduction signals. This system separates the video components, generates the sampled color digital values and luminance components, and inputs them to the storage buffer. The data processor compares a continuous sample of component video data, from which the data processor represents a slow-changing signal or a fast-changing signal by generating a variable-length block of video data. A circuit that encodes and multiplexes audio and synchronizes data with a signal stream, a circuit that encodes signals and control data to be transmitted to the receiving side, and a circuit on the receiving side that executes this process in reverse. Exists. The process of reducing iterations is to use variable speed scanning by using codes that indicate when the color information, luminance information, and luminance information can be repeated. When the information can be repeated, only a total of 8 samples are transmitted. Timing information that controls the scanning rate is extremely important. Regular sequential scanning of lines now requires reliance on circuits for creating rate tags with interpolated data. On the other hand, the communication processor of the present invention reduces the repetition of video data.

【0043】

U.S. Patent No. 1 granted on July 19, 1983 Nos. 4,394,774 describe methods for digital video compression and expansion systems and methods for compressing and decompressing digital video signals in real time at rates comparable to NTSC color broad carriers. The system compressor receives digital frames, divides these digital frames into subframes, and executes a single-pass spatial region to transform the region transformation into two-dimensional pixels. The composite coefficients are normalized and compressed using a predetermined rate. There is an optimal rate buffer that controls the field backup for compression. The system compressor optimally determines the field back component that controls the capacity of the rate buffer and controls the absolute amount of data resulting from the normalized step with respect to the momentary data content of the rate buffer memory related to the capacity of the rate buffer memory. Because it does, Rate Bashua will never be empty or full. In fact, color pictographs are classified into luminance and I-axis and Q-axis color components. The luminance component is compressed and extended using a rate buffer feedback technique that optimally encodes the background. The I-axis and Q-axis components are easily spatially passed through the low-pass filter and then spatially further sampled by dimensional interpolation on the receiving side of the system. The audio is sampled at a fixed rate through a filter, spoiled with a bitscreen synchronization code, and transmitted as a series bitstream.

【0044】

The other products will be further described below.

【0045】

Published April 1, 1991, PC Week Journal has a 43-page article on "Video Conferencing Expecting Analysts to Boom in the 1990s." Experts predict that video conferencing will grow as a form of communication.

【0046】

The Stanford Computer Optical Laboratory at 3530, Schugerbury Lane, P.O. Boxes 31266, Walnut Creek, 94598, Calif Ornia, makes a product called "4 Quitsk". This product is an image capture device that creates 512 x 512 low-light images by creating 30 to 60 frames per second using variable delay. This device operates between wavelengths of 130 to 920 [nm] and can create one frame in 5 [ns]. This device can be applied to create image frames in a time-division array for the user.

【0047】

The Welshyu Allen Inspection Systems Bureau, located at 4619 Jyodan Road, Scanny Teres Folds, 13151, NY, makes a product called the "VP3 Imaging Endscope." A small video probe-type video camera, which is slightly larger than a fountain pen, creates an image and sends the image back to the device along the communication line. The device compresses and modulates the image and transmits it over the telephone line at speeds of 14,400 [bps] to 2,400 [bps]. This device indicates that it can be integrated into a display bezel.

【0048】

Dialology, located at 300, Littleton Road, Percypany, NY, has created "Call Processor" and "Audio Texts" information programs for voice messages and automated relay telephone network services. It uses interactive voice response as an important feature for centralized dictation services. This product is representative of various voice compression and voice processing that can be realized.

【0049】

92121, Calif., San Diego, Sorrento Valley Road 11722-D, Telefoot Conni Unications, Inc. makes an image compression product called "ALICE." This product compresses to 1/15 without any loss of resolution. This product is a software device and is designed to be integrated with other products such as conference calls, pictorial databases and surveillance. It claims that it can send full-color images in less than 10 [seconds] over a standard telephone line or can store 4,000 high-resolution color images on a 50 [Mbyte] hard disk. This product is representative of the various compression levels that can be achieved.

【0050】

Data Products New England, located in Burns Park North, Warinford, 06492, Connected, makes a product called the "DPMUX M-44." The device receives channel input from the telephone exchange using E & M signals or baseband data from the digital device to the port. The audio signal is then converted to the requested digital audio of the bit rate, or in the case of data, it is retimed to the network timing and a time slot is assigned on the collective network transmission side. This collective network transmission side is either a small service or all T1 services. With this device, the operator can construct the number of channels, the specification of the bit rate, the specification of the timing (asynchronous, synchronous, independent synchronization) and the type of voice service. The concept that an operator can build and the concept of being downloaded to a remote location by a set frame can be extended to the concept that it can be built automatically by a control algorithm, and a remote location by a set frame. Can be downloaded to.

【0051】

The network transmission device made by various manufacturers has a function applicable to the communication processor of the present invention. These features include Time Slot Exchange or Digital Access Crossconnect Service (DACS) equipment and, for example, Network Equipment Technology, Inc. at 800 Saginaw Drive, 94063, California, California. It is a voice insertion device (TASI, voice data that can change the bandwidth instantly) like the IDNX product of. This class of equipment makes efficient use of available bandwidth by adapting data and voice to adjustable formats. This class of equipment is described as a transmission resource manager. The concept of managing transmission resources can be applied to the communication processor of the present invention for video conferencing on workstations.

【0052】

Published in February 1991, there is a special article "Streamlined Protocol" in Telecommunications. This article describes changes to transmission protocols that are streamlined to improve transmission performance. The communication processor of the present invention uses an optimal protocol, which allows the communication processor to communicate with a multiple electronic digital transmission device, which facilitates transmission control for video conferencing on workstations.

【0053】

Published in February 1991, there is a special article "LAN Interconnect Technology" in Telecommunications. This article describes the differences between bridges, routers and switches as used in interface to local area networks and transmission networks. The communication processor of the present invention performs network functions for bridging, routing, repeating and switching.

【0054】

There is the 11th General Assembly of the International Telegraph and Telephone Advisory Board (CCITT), Document 52, Research Group 7, Report R43, Recommendation X.200 Open Interconnect Reference Model for CCITT Applications. It is also called the 7-layer ISO model. The application process referred to in this document is considered to be the video conferencing performed by the communications processor of the present invention.

【0055】

[Problems to be Solved by the Invention]

The present invention can be used for remote communication networks and provides the requested video conferencing service by interfering with different channels and services of different charges by the communication processor of the present invention on behalf of the workstation user.

【0056】

The workstation needs to communicate with another workstation that has the audio and video information of the communicator. Video of a correspondent with pictorial or audio can significantly improve individual productivity by significantly improving person-to-person communication. An object of the present invention is shown below.

【0057】

An object of the present invention is to propose a low-cost and simple video interface method from a workstation to the communication processor of the present invention. There are two or more solutions for designing this approach, but only one solution needs to be shown.

【0058】

Another object of the present invention is to propose a simple video weighting algorithm that is highly recognizable without the need for high bandwidth. There are two or more weighting algorithms that satisfy the weighting requirements. Since the weighting method is proportional to the transmission bandwidth and inversely proportional to the display performance, the communication processor of the present invention selects an appropriate weighting method that gives the best performance under existing conditions. It is important to use an algorithm for choosing a weighting algorithm.

【0059】

Another object of the present invention is to propose a speech management algorithm that includes the use of several weighting standards. Due to this weighting standard, the communication processor of the present invention is compatible with some standards and can be connected to various types of network lines. This method takes into account the type of service requested and the type of service requested at a remote location. It is selected based on the alignment between these two services. This selection serves as a criterion for selecting the weighting method described above.

【0060】

Another object of the present invention is to propose a network management algorithm for dynamic multiplex transmission in optical fiber media standards such as hierarchical T-carriers, SONET and FDDI that descend to the DS0 level. The DS0 level is the minimum common term for audio / video channels because compatibility with existing standards is required for the development of communications, and DS0 is an outstanding form of interface. An important function of the communication processor of the present invention is to collect at least audio and video in a fragmentary T1 carrier. The communication processor of the present invention uses a higher carrier rate when the bandwidth allows a higher carrier rate.

【0061】

Yet another object of the present invention is to propose an artificial intelligence (AI) for this system that controls the final processing after using the transmission network as an input base for processing determination. Factors such as connectivity and activity are the criteria used to assess connectivity potential. These types of factors add to the user requirements for services, workstation capabilities and the availability of remote users.

【0062】

[Means for solving problems]

In order to solve such a problem, in the present invention, in an audio / video communication processor system that combines workstation units in a system that transmits audio information and video information by a carrier, a digital bus is provided and elements are combined. , A communication processor that interconnects elements, a workstation interface that processes video and audio information at the workstation level, a video processor, and an audio processor, which are connected to a digital bus and control communication via the digital bus. A channel frame processor and a static audio / video multiplexing processor that is connected to the digital bus and dynamically allocates bandwidth between the audio information signal and the video information signal on the digital bus are provided, and the communication processor is a net. It has multiple network interface ports, including a port for work carrier signals and a port for local loop carrier signals, and the communication processor is a port carrier signal from one network to another and a combined work station to network. It has a means to transmit the information conveyed by the port carrier signal to the work port, and by interconnecting the workstation interface, the video processor and the audio processor, a digital signal and an analog signal pass between them. , Send digital information via the digital bus to the communication processor.

【0063】

Further, in the present invention, in an audio / video communication processor system that combines work station units in a network system that transmits and receives audio information and video information by a carrier, a digital bus is provided and elements are combined. A communication processor that interconnects elements, a means for receiving different network carrier signals, a means for transmitting and receiving audio information and video information by a network system, and video frame information having audio information and control information. By adjusting, a means for interrelating the transfer of voice information and video information under the control of the user's request and communication processor is provided, and the means for transmitting and receiving the voice information and video information is a certain network carrier. Include means to convert the signal format to another network carrier signal format.

【0064】

Further, in the present invention, in an audio-video communication processor system that combines work station units in a network system that transmits and receives audio information and video information by a carrier, a digital bus is provided and elements are combined. A communication processor that interconnects elements, a means for receiving different network carrier signals, a means for transmitting and receiving audio information and video information by a network system, and video frame information having audio information and control information. A means of interconnecting and sharing different elements of an audio-video communication processor system by coordinating and correlating the transfer of audio and video information under the control of the user's request and communication processor. The means for transmitting and receiving audio and video information shall include means for converting from one network carrier signal format to another network carrier signal format.

【0065】

[Action]

According to the present invention, there is provided a system that combines workstation units in a system that transmits audio information and video information via a carrier. This system has a communication processor with a digital bus, which interconnects the elements coupled to the communication processor. The network interface port of the communication processor includes a port for the network carrier signal and a port for the local loop carrier signal. This communication processor interconnects various units on the network. By interconnecting the workstation interface, the video processor, and the audio processor, digital signals and analog signals pass between them, and digital information is transmitted via the digital bus of this system. When controlling this system, a channel frame is provided which is connected to the digital bus and controls communication via the digital bus. Further, in this system bus, a static audio / video multiplexing processor that is connected to the digital bus and dynamically allocates a bandwidth between the audio information signal and the video information signal on the digital bus is provided.

【0066】

The voice processor of the present invention has a voice compression / weighting subprocessor coupled to a communication processor system and shared by the user. In addition, the voice processor has a method prepared by the user of the communication processor system to perform compression and weighting based on the available bandwidth and the determination of the communication processor, thereby finalizing the requested bandwidth. Allocate bandwidth process control.

【0067】

The video processor of the present invention is coupled to a communication processor system and has a video compression / weighting subprocessor shared by the user. This processor was requested by compressing the video information and following the control from the communication processor based on the video activity and allottable bandwidth and weighting provided by the user of this system and the judgment of the communication processor. Allocate final bandwidth process control to bandwidth.

【0068】

Since the system of the present invention is provided with the static audio / video multiplexing processor shared by the users of the system, the communication channel and the usage status are modeled and the channel transmission in the system is performed. Calculate the parameters for. Combine the channel frame processors so that they are shared among users, and assemble the channel frames for insertion into the aggregate frame.

【0069】

Further, the present invention provides a composite spar frame processor common to all users, and this processor is provided for each port that implements the language of the remote communication interface.

【0070】

Audio-visual communication processor systems can run artificial intelligence (AI) software for a common user, and all other system processor elements provide parameters for performing digital language modulation conversions from one channel in a set frame. By synthesizing, the channels of other collective frames are regenerated to set up the conditions for using Faxis Milli, and the channels are threaded so that the operation of the workstation is connected between the system elements that connect to the above system. Subordinate to the communication processor system.

【0071】

[Example]

Before explaining the preferred embodiments in detail, it is important to explain, for example, how a communication processor works. In the workstation, user 1 decides to request a video conferencing session with another user 2. The user has the same equipment but is in another building. User 1 launches a menu on the screen. This menu uses a mouse to allow the user to "Who is" Click Who user directory. The user selects the name of the person he / she wants to set up from this directory. This name is associated with a list in the routing directory. The user also selects the window size for the received video. The name in the directory window is associated with the routing directory and is mapped within the routing directory. Since the communication processor is associated with the route to user 2, the communication processor is associated with the network port and the remote communication network interface (ISO level) for that port. The communication processor then tests the ports for available bandwidth, queues and lists the requests in the job, and returns to the user with a workstation with the recommended window size.

【0072】

A high fidelity symbol is added as a tag to the window size. This high fidelity symbol represents sound quality. Increasing the window size reduces the sound quality, and decreasing the window size improves the sound quality. As the user increases or decreases the size of the window, the high fidelity symbol changes from large to small. The video window consists of two elements. That is, one is size and the other is speed. The speed at which the image is updated is represented by the rotating wheel (reel of the video camera). The communication processor then controls the speed of the rotating wheel by communicating the window size and speed of a large range of workstations, which can be taken into account. The workstation uses application software to process the high fidelity symbol and make this high fidelity signal proportional to the low to maximum values. The faster the rotating wheel is spun, the closer the image is to the real thing, and the slower the spinning wheel is, the farther away it is from the real thing. Experience causes the user to reach an acceptable value.

【0073】

Next, the communication processor starts the session. It begins by opening a channel to other communications processors and forwarding requests for services. The first communication processor determined the available bandwidth. The second communication processor must determine if the second user 2 is available.

【0074】

By improving the video processing of the hardware system, as in U.S. Pat. No. 4,862,264, the video is processed by the block to identify the activity of the block, and the transmission is completely accurate or the accuracy is reduced. Judge whether it is transmission. With these improvements, transmission can always be performed with a reduced charge system, which can be convenient for the next influential user. This reduced tariff does not necessarily affect video quality when redundancy, rather than information, is removed in the transmission selection process. Thus, U.S. Patent No. 1 Unlike 4,862,264, the communication processor system makes decisions based on the dynamic range of pixel numbers. Video is always transmitted as a binary number. The number of binary symbols per pixel is fixed by the A / D process, but the number of bits transferred and their position within the numeric field are first determined by the communication processor based on the available bandwidth. Will be decided. The communication processor may choose not to transmit video, but this bandwidth may be devoted to audio or data. The user may mention any change in video quality as its purpose is to reduce bandwidth conditions on the boot side with a controlled field of view and a controlled illuminance of the field back. It may or may not be mentioned.

【0075】

Note that the communications processor of the present invention may be improved to cover areas covered by the communications processor of US Pat. No. 3,795,763. The memory is used to hold consecutive video frames and execute a difference algorithm between these frames. The communications processor of the present invention then performs a new process different from that described in US Pat. No. 3,795,763. As an example of this difference, there is an example in which pulse synchronization information is not transmitted because the video hardware of the workstation at the other end has different video characteristics from the transmission workstation. However, it should be understood that the communication processor of the present invention processes workstation video rather than a normal television signal.

【0076】

The second communication processor determines that user 2 can use it, and then proceeds.

【0077】

Each workstation can set the reception status. When the workstation is running and in use and the user sets the receive status to "NO", the communication processor denies access to the first communication processor because it is in use. User 1 is also notified. When user 2 sets the reception status to "YES", an inquiry statement appears through the screen. If the user proactively responds to this query before a predetermined amount of time has elapsed, the sensation is initiated. Since the parameters of workstation 2 are the parameters of workstation 1, they both have the same fidelity and window size. Also, by customizing this session, User 2 with a higher performance workstation can have a larger window or a higher quality video window. The customized parameters can be set again by the user through the menu. The response to the receive state also includes information about the window size defined in the customized list, if available.

【0078】

This is a simple case because the second user is in a different building but has the same equipment. In such situations, it is expected that remote communication networks will not be used, but rather optical cables or wire cables with FDDI-type bandwidth may be used. When the second user is not in the same location as the first user, the telecommunications network is the means of communication for transporting the traffic. In this second situation, there are more interactions with more types of equipment. In most cases, the communication processor communicates through a relay device. The communication processor must be able to use the appropriate language, such as 2B1Q or AMI, through the physical layer (see ISO, CCITT Recommendation X200, etc.) level and the network layer.

【0079】

The communication processor can act as a relay station between two other communication processors that you want to communicate with. This is a necessary function because there is no transmission device that can interact with the communication processor and hold a video conference so far. In situations where the communication processor acts as a network device, the communication processor may be called to perform routing or bridging between two separate transmission lines. For example, a communication processor is required to convert from 2B1Q to AMI. The rules of conversion are embodied in artificial intelligence software. If the rule does not exist, the communication processor will not attempt to root or bridge. This area is very complex and will not be discussed here.

【0080】

Cost is an important determinant for implementing new technologies. The cost of communicating video and audio from a workstation can be significantly reduced by providing a common communication processor to serve the user. This cost is reduced by using the display and control processing power present in the workstation. Rather than competing with Token Ring or Ethernet Local Area Network (LAN), the communications processor performs different services by communicating audio and video within the functional area at low cost and preparing for long-distance transmission. To do. The purpose of the communication processor is to provide excellent voice and communication, to inform as much as possible of the time-varying constraints of the transmission medium and the time-varying degree of loading or usage. Communication processors do not rely on specific standards for audio or video, but rather adjust bandwidth, resolution and transmission rate to meet this constraint when service is required.

【0081】

The workstation initiates a service request. The user does this by launching a software-driven menu on the screen. Menu parameters are determined by the communication processor. When the parameters are selected, these parameters summarize the service request. The service request contains data about the nature or form of the service and the destination of the signal. This information is sufficient for the communications processor to attempt some routing threads before making a positive decision. If a positive decision cannot be made, the communication processor determines what can be done and suggests to the user what can be done.

【0082】

The main concept of this design is to integrate several disciplines that use system engineering techniques. A low-cost communication processor is possible through the development of high-speed signal processing, optical processing, and light wave processing that adheres to multiple communication standards.

【0083】

An embodiment of the present invention will be described in detail below with reference to the drawings.

【0084】

FIG. 1 shows a preferred embodiment of a communication processor specified to process the audio and video of a workstation using artificial intelligence technology to use a remote communication network. This embodiment uses a telecommunications network, which is the first to provide a simple and effective means for personal video conferencing. Communication processors are a collection of existing technologies that have been improved and modified to work together for normal purposes. The most important function is the network type as shown in Fig. 1. Networks have evolved in the direction of increasing bandwidth. This was true at one point, but it has also evolved towards providing digital bandwidth-style services. With this service, customers interact with networks at various levels. Generally referred to as the ISO (International Organization for Standardization) 7-layer structure, the lower layers basically control the network equipment to access the physical layer, data link layer, network layer and transport layer. It was divided like this. The ISO specifications for access to the physical layer, data link layer and network layer are incorporated into the communication processor, especially AI, and communicated to the network consisting of these elements that provide these functions. Starting from the transport layer and moving to the upper layers, the communication processor communicates only to other communication processors to provide this function. Currently, it may be possible to structure a part of the transport layer of the communication processor to operate with the remote communication network device in the upper layer, but the physical layer, data link layer and network layer currently in the lower layer. The function of is incorporated into the artificial intelligence part of the communication processor. These three lower layers are designed within the communication processor using the CCITT definition to communicate to the remote communication network device. Furthermore, the functions assigned to these layers are used by artificial intelligence software.

【0085】

The communication processor consists of the following eight basic components.

【0086】

(1) Audio compression / weighting subprocessor shared by multiple users. The algorithm chosen is based on the bandwidth that can be allocated. Such sharing is important as it provides key bandwidth process control along with the communication processor.

【0087】

(2) Video compression / weighting subprocessor shared by multiple users. This compression algorithm and weighting process follows control from the communication processor based on voice activity and available bandwidth. This subprocessor works with the audio compression / weighting subprocessor described above.

【0088】

(3) Video imaging device (may be incorporated in the bezel). It works with a communication processor through an interface board set for workstations and related software.

【0089】

(4) Voice-encoded processor (may be incorporated in bezel). It provides digital audio capabilities in either 64,000 [bps] PCM or 32 [Kbps] or 16 [Kbps] ADPCM or LPC bidirectional dual or full duplex.

【0090】

(5) A static audio / video multiplexing processor shared by multiple users. This feature is driven by AI software within the communication processor. This processor essentially performs communication channel and usage state modeling. This processor calculates the parameters of the channel and the aggregate processor. This processor utilizes collected data, such as error free seconds, to minimize the overhead by changing the error correction code.

【0091】

(6) The channel frame processor may be shared or shared among users. It is part of a communication processor and has the ability to collect channel frames for insertion into aggregate frames. This implements the results of a static audio / video multiplexing processor. This feature is driven by AI software within the communication processor.

【0092】

(7) A composite superframe processor common to all users. There is one processor for each port. This feature is driven by AI software within the communication processor. This processor uses, for example, FDDI, 2B1Q, AMI or such a form of telecommunications interface language.

【0093】

(8) A communication processor that runs artificial intelligence (AI) software that is common to all users. All other processors are subordinate to this processor. This processor synthesizes the execution parameters of the digital language modulation conversion from one channel of the set frame and reproduces the channel in the other set frame. This also sets up the conditions for using fax machines.

【0094】

The devices described in (2)-(5) are independent (not related), but these devices define the specifications between them. However, it is not defined in the sense of a unique device, but in a specific functional sense. These devices represent what can be achieved with current technology. Given what is available, communication processors can evolve from these types of equipment or similar types of equipment.

【0095】

(6) and (7) are clearly transmission bandwidths by allowing more voice channels to be assigned to the digital carrier than the channels allowed by the TASI device's fixed multiplexing system. It is related in its own way, except that it is to increase. For example, T-1 carrier is 64,000 [bps] ] 24 fixed channels, but when the T1 carrier is voice-inserted (TASI), more than 32 channels can be transported. The quality of service is degraded due to TASI and data transmission is severely compromised. The extra capacity of the audio channel is transmitted using the period when the audio is not flowing, that is, the period when the audio is inactive. The collective frame structure contains information in remote areas and reconstructs the audio channel in the proper order. The difference is that the fixed 8 bits used when encoding the audio signal dynamically drops from 8 bits to 4 bits, which depends on the dynamic cleanliness and activity. When the encoded data is replaced with voice, the TASI algorithm produces enough signal distortion noise to destroy the data. Therefore, it is the user's responsibility to know how to handle these data lines with the remote communication carrier. [0096]

You do not necessarily have to use the line purchased for a specific fee for other purposes. Communication processors fit within this category. AI coding is specific to each toll line and cannot be replaced. Even when the nominal speeds are the same, the port of the communication processor designed for the ISDN (2B1Q) B channel cannot be replaced with a 128,000 [bps] channel from the T-1 programmable multiplexer.

【0097】

(1) The characteristic of the first element, the network architecture.

【0098】

Several workstations are connected to a communications processor to perform video conferencing. The video conference may have the starting point and the other party on the premises, or the other party may be remote. The remote destination requires a set-up of similar communication processors and workstations, but not the same as the parent unit. The workstation of the same communication processor can also hold a video conference. The workstation runs the software to perform some tasks with the communication processor. These tasks are as follows.

【0099】

(1) Instruct the user on the format of the service to be executed using the menu. There are two types of information to collect. This includes general types of required video services such as window size and low to high reflection speed, and required audio services such as good sound quality from bass to treble and no delay of 1 second. There are general types.

【0100】

A high fidelity symbol is added to the window size. The high fidelity symbol is a record player icon. The high fidelity symbol represents sound quality. Increasing the window size reduces the sound quality, and decreasing the window size improves the sound quality. As the user increases or decreases the size of the window, the high fidelity symbol changes from large to small. The video window consists of two elements. One is size and the other is speed. The speed at which the video is updated is represented by the rotating wheel (the reel of the video camera). The video symbol is a video camera with a reel on the top of the box and a lens on one side of the box. The side surface of the camera is a box whose size can be changed, and the received image is arranged in this box. The communication processor then controls the speed of the rotating wheel by communicating the high range window size and speed of the workstation, which can be considered. The workstation uses application software to process the high fidelity symbols and make these high fidelity symbols proportional to the low to maximum values. The faster the rotating wheel is spun, the closer the image is to the real thing, and the slower it is spun, the farther it is from the real thing. Experience causes the user to reach an acceptable value. If sufficient bandwidth is available, there is little or no interaction between high fidelity symbols, reels and window sizes. When the bandwidth is small, the movement is small and very large dialogue occurs immediately. The communication processor always controls the size of the high fidelity symbol and the user cannot request or generate an impossible service. All impossible services are video and no audio, or all are audio and no video. That is, no situation is programmed into the artificial intelligence software.

【0101】

Also, voice services require some caution. Fidelity comes at the expense of video quality as part of the price of sound quality. In this case, the synchronization of audio and video is unique. The communication processor always tries to synchronize the audio with the video. But at reel speed, synchronization of audio and video is unique. The communication processor allows a one second difference in synchronization between audio and video.

【0102】

(2) Instruct the user on the telephone number of the remote endwork station. Video conferencing routing uses telephone numbers with specific names and passwords and times based on call acceptance criteria. The communication processor establishes a connection instruction table by communicating this information in the superframe between these processors. Phone numbers are important because they make a huge investment in phone number databases. The communication processor is programmed with the ID of each user service. This information is actively shared between communication processors. Part of this information is the reception status of the user. If this reception status is negatively set, the communication processor will not attempt to call. Since the communication processor network is an important part of the user directory, "Who is" The Who directory first embarks on a map showing where users are concentrated. A software control foot in one of these locations will launch the user's name and phone number at that location. In addition, there is an indicator that shows the reception status so that the user does not make unnecessary efforts. This information is shared between communication processors using the total capacity of the idle superframe.

【0103】

Features of the second element. The video of the workstation will be explained.

【0104】

A method of separating the optical function and the electronic function of the camera and providing the electronic function of the camera in the communication processor may be a low-cost method. A basic camera that converts an image into a signal contains two essential elements: focusing, which belongs to optical functions, and image conversion, which belongs to electronic functions. Such a camera may be embodied by using focusing with an optical fiber focal plane image collector located on the communication processor. The purpose of this approach is to serve a large number of users using one set of high speed electronics. If the imaging device is fast, the device can serve the user at the same time. In the communication processor, the weighting algorithm and the multiplexing algorithm can be applied at the same time. Figure 2, titled "Workstation Video Input," shows an interface between video collection lines located where the workstation is located. An image lens and a video collection system are a video endoscope or a modified form of the system. To keep the depth small, the primary lens can be designed to interface with a right angle mirror. With this right angle mirror, the required focal length can be obtained along the width of the bezel. By this method, the depth of the bezel can be reduced. The image device or electronic focal plane is then perpendicular to the image. Also, the electronic focal plane does not have to be in the bezel, but may need to be in the communication processor. Next, the image lens and the collection system transmit the image existing in the optical format in the optical guidance device to the communication processor. Both techniques work, but considering the cost of implementation, the method of providing an electronic focal plane in the bezel seems to be more preferable, but the second method further develops into an alternative method, sexion.

【0105】

In a preferred method of providing an electronic focal plane in the bezel, consider the possibility of a conventional method using a small camera device such that the video imaging device in the bezel is placed on a workstation display. In this case, a single wireline cable is all that is required for the interface between the workstation and the communication processor. The bandwidth of an average RF cable is sufficient to support frequency division multiplexing for transmission and reception of video, audio and data. Figure 11 shows that when this technique is used, the interface to the communication processor is simplified to other opposite types of signals that share this cable.

【0106】

One key point of the present invention is the weighting algorithm of the video processor required to maintain the image quality and the transmission bandwidth. Compression is required in addition to weighting. Compression and weighting are not mutually exclusive, but must be designed to be compatible with each other. Consider the following. That is, if the image is divided into scanning zones using a unique central zone defined as an area that spends more than 50% of the time required to focus, then the time required to focus A unique concentric peripheral zone is defined as an area that spends 35 [%]. The rest of the concentric peripheral zone is the remaining 15 [%]. FIG. 11 is a graph showing the applicable weighting methods.

【0107】

An image frame composed of N × J vertical scan lines and horizontal scan lines is digitized to M-bit resolution. For example, let M be 12, and represent it with N = J = 512 pixels. Next, each pixel is represented by a resolution of 12 bits. The total number of bits is (N, J, M), that is, 3,145,728 bits. At this rate, a series channel running at a rate of 1.544 [Mbps] compresses this number of bits to 1/15 and 209,715. It takes more than 2 seconds to transmit even if it can be bitten. The first video frame is compressed and transferred so that there is no previous frame for comparison. The next frame is weighted relative to the first frame by taking the difference between the first frame and the previous frame compared pixel by pixel. When the composite value is less than the predefined delta amount, the composite value is zero for that pixel. The position of the first non-zero pixel is indicated by (N, J) as (x, y). Also, the position of the next pixel at the position (x ±, y + 1) must be non-zero. Sigma is the deviation calculated by comparing the next r scan line consisting of non-zero pixels. Here, r × s is about 50 [%] of the central portion of the pixels, and when calculated in this way, 50 [%] of the total number of pixels is less than the maximum. Therefore, (x, y) to (x + s, y + r) is the central part of the pixel. But to calculate the starting point when transmitting in 12-bit numbers, use x as the starting row value and y as the column value. The last values in this sequence are x + s and y + r. If a zero value occurs by comparison before 50% of the central part of the pixel occurs first, then the total number of bits is (256, 256, 12) or less than the maximum of 786,432 bits. The next time you compress, divide the maximum number of bits by 15 to reduce it to 52,428 (which is the average compression possible). Since DS0 is selected as the base carrier in the present invention, a basic carrier rate of 11,570 [bps] can be used for voice and overhead. FIG. 12 shows possible weighting algorithms.

【0108】

In the digital format, video weighting is applied. In this algorithm, the video is assumed to be 640 pels x 480 pels, but can be any number. The combined number of pels is 307,200. By making the definition arbitrary, 50 [%] of this number is given a resolution of 12 bits, 35 [%] is given a resolution of 8 bits, and 15 [%] is given a resolution of 4 bits. Given the resolution. We chose 12, 8 and 4 for the number of bits to exaggerate and clearly explain the concept of weighting. Communication processors could use such values. The picture is transmitted in grayscale with 4,096 shades, 256 shades and 16 shades as the rest. Each area is compressed using a boundary-dependent image compression algorithm.

【0109】

One of the 50 to 100 compression factors is possible in real time, but communication processors do not need real time. The video is output as a series of images updated at a rate proportional to the requirements, bandwidth costs and system loading conditions of the service. The signal-to-noise ratio of the video is very important for the user to accept. In order to increase the signal-to-noise ratio on the activation side, a technique for illuminating an object from a short wavelength region to a medium-short wavelength IR region using a low-power laser is shown.

【0110】

Illumination with Fresnel lens dispersion is a good way to keep photometric measurements at high and low emissivity analysis levels. It can be seen that if this low emissivity analysis level can be kept low by the Fresnel dispersion lens, the safety of the target eye can be guaranteed. At the same time, the frequency response of the detector should be as wide as possible, from visible wavelength IR to medium and short wavelength IR. Figure 6 shows the detector frequency integration scale, showing the effect of integrating all the wavelengths shown. In fact, the imaging system does not attempt to control the lighting, except in the broad sense that it has enough light to produce a good image. Spectral components and short-term behavior play important roles in image processing. The purpose of image illumination is to remove dispersed illumination such as AC characteristics of power lines and blue / green spectral components that shine in all directions as if they were open. Due to the blue / green spectral components, the communication processor performs more image processing work. The illuminance of the office depends on the AC power supply for the primary power supply and the light that spreads in all directions like a flower.

【0111】

Integrating the amplitude for the frequency bandwidth yields a signal with a high signal-to-noise ratio. This signal is reproduced in a remote location as a black and white grayscale image. In other words, white is not a single color, but white is a function of a single value of only amplitude (the area below the curve in FIG. 13), so in this case it can be considered in this way. This method has been simplified for spatial and temporal reasons. It can be seen that the last signal is the illumination function on the activation side, the optical fiber bandpass response, and the convolution of the detector bandwidth.

【0112】

Features of the third element. Voice management will be described.

【0113】

The communication processor may be connected to the workstation via a high quality coaxial cable. The coaxial cable is divided into several frequency management units that use frequency division multiplexing. As a result, the coaxial cable can be loaded with double audio, received simplex video with a field back, double data, and double network data. Figure 15 shows how speech is processed at the pso level.

【0114】

FIG. 7 shows the general configuration of the entire unit network of the system for the audio-visual communication processor shown in FIG. Typically, the system provides video conferencing by having one or more communication processors to serve a set of workstations that process audio and video transmission.

【0115】

The communication processor uses artificial intelligence software to know the connection status. Since the conversion rules are contained in the table, the system can react to the communication environment. The combined system processes optical signals for low-cost communication video conferencing and long-distance transmission that transmit audio and video within the functional area. The communication processor provides audio and video communication under time-varying transmission medium constraints and time-varying loading degrees or conditions of use. Bandwidth, resolution and transmission rate can be adjusted to meet this constraint when service is requested. The workstation initiates a service request. The service request contains data about the nature or form of the service and the destination of the signal. This information is sufficient for the communications processor to attempt a thread before making a positive decision. If it cannot be affirmed, the communication processor determines which is possible and outputs to the user a request that can be changed. With respect to FIG. 8, the audio-visual communication processor has a network port, a local port and a communication processor unit in the system, and this system is composed of an additional unit connected to the system digital bus. A composite superframe processor, a static audio / video multiplexing processor, a channel frame processor, and a workstation subprocessor are connected to this system digital bus. The workstation subprocessor includes a video processor having a weighting function and a compression function, an audio processor having a compression function and a weighting function, and a workstation interface line. Both of these workstation subprocessors are interconnected in both analog and digital styles. The system has a telephone-style interface port for connecting workstations as part of the overall system.

【0116】

The telephone-style interface is the only way of thinking that this interface allows a stand-alone device such as a fax machine to interact with a workstation. The workstation can run the fax emulation software, which converts an electronic document into a fax signal and sends it to a remote fax machine. The telephone-type interface can receive telephone signal information and outgoing signal information.

【0117】

The workstation audio signal is received by the communication processor. Received by multiplexing one of the N voice input lines for processing. Sampling is performed at 64 [Kbps] and is compatible with what is currently being done (first level DS0 carrier). Weighting is either the mu-method or the a- method. The communication processor automatically converts between the mu method or the a method when the routing table description instruction conversion is required for video conferencing across the Atlantic Ocean. After weighting, the signal is further processed to reduce the transmission rate and assemble the composite data into a numerical sequence suitable for the packaged network. The rate at which the packet is transmitted depends on the loading of the network. The communication processor network does not try to compete with the telephone network, but adds audio to the video. However, it includes options for choosing to include only audio.

【0118】

Be careful about phone-style voice bandwidth. The 64,000 [bps] rate is the Nyquist sampling rate, which samples 8,000 samples per second. Each sample is 8-bit binary data. 2 to the 8th power is 256 in decimal notation. Since the electrical audio signal must be symmetric about the zero [v] axis, 256 is split into +128, -127 and 0. The total number reserved by these positions is 256. This means that the total signal bandwidth of 4,000 [Hz] can be used for voice. The nominal bandwidth of a Type 500 handset commonly used in telephone equipment has a nominal bandwidth of 3,000 Hz. The first restricted filter is this handset. The handset or headphone that is part of the communication process group is not thus limited and has a nominal high fidelity voice bandwidth of 4,000 Hz. FIG. 15 shows how an audio packet is formulated for video conferencing. Linear Predictive Coded Compression (LPC) increases audio data bandwidth by 1,200 [bps] ] Serial rate can be reduced. This voice is inadequate with respect to factors such as speaker recognition and word loss, which causes the voice to repeat when the transmission medium does not perform well. The quality of this transmission medium is expected to be good, so the algorithm only reduces the audio to 1,200 [bps]. Audio is processed with 2,400 [bps] LPC and 16,000 [bps] or 32,000 [bps] Optimal Differential Pulse Code Modulation (ADPCM) and 64,000 [bps] PCM. Which of these to choose depends on the available bandwidth, user requirements and remote compatibility. These factors are recorded by artificial intelligence software running on the communication processor. Figure 4 shows how variable splitting is used. Variable partitioning reduces redundancy and saves bandwidth space.

【0119】

The voice can be replaced with a fax signal when the fax is requested. The bezel or workstation adapter has an input end for a flux signal. The communication processor does not generate flux signals, but one side of the workstation generates these flux signals regardless of the communication processor's interface adapter. The dynamic allocation of bandwidth is based on voice as the first priority when the signal is a flux signal, even in the case of network loading. The video is transmitted using the period during which the audio is inactive. Speech is dynamically compressed from 64,000 [bps] using a good predictive compression algorithm. Voice data rate predictions range from 2,400 [bps] to 1,200 [bps]. The audio compression may be zero, and in the case of a software signal, this audio data rate is DS0, or 64,000 [Bps]. The workstation packet control section is also dynamic. If the information is redundant, the iterative signal of the previous packet is used to carry this information.

【0120】

The communications processor needs to collect all the software packets before sending them to the workstation. The fruit package cannot be destroyed when it is possible with an audio package or a video package. This feature requires a cross. Speech processing also requires a cross. The video signal does not require a cross. The video packet may be discarded if necessary. The workstation will not assemble this frame unless the new frame is complete and damaged. Although U.S. Pat. No. 4,682,225 does not mention it, we believe it is necessary to improve the hardware for video images across multiple conference call stations over the telephone line.

【0121】

The communication processor of the present invention having a video processor discards pixels or pels. These pels are discarded because the receiving station has requested a window size for the video. Pells that do not have an address in the window, as determined by the addressing algorithm, are discarded. This is the first step to reduce unwanted bandwidth. All pels with addresses in remote windows are processed by compression or weighting. There is no single representative pel used to represent a group of pels.

【0122】

Features of the fourth element. Explain network management.

【0123】

The key to understanding the concept of communication processors is to incorporate standard compatibility of networks between various transmission systems (standard bridges). The T carrier system is common and price lists are available for this service. T-Caria may be used in ISDN networks. Synchronous optical networks (SONET) used different multiplexing systems and were also able to carry ISDN signals. The lowest level is It is OC-1 at 51.84 [Mbps]. This level may be used for local area transmission access (LATA) loops or as a transmit long-range network. This is a suitable carrier because of its large capacity to provide excellent communication. The communication processor must be compatible with signaling technologies such as DS2 / DS3. Although the communication processor does not interact with the telephone exchange at any level, the wide area network transmission device may search for this signaling pattern and can exchange network routing information instantly. Artificial intelligence based on the knowledge of the price list knows when the carrier can do this, so it is not possible to allocate the most important control information to the most sensitive bit position. This feature also applies to ISDN bear lattices.

【0124】

For local loop operations that are common to all functions, it is desirable to select the FDDI carrier. If you depend on the telecommunications network carrier, you do not need to select the FDDI carrier. FIG. 17 shows how the local loop and the telecommunications network are different. Also, intra-regional carriers can provide 64,000 [bps] carriers from the telephone system rather than from telephone lines. Such lines are probably inserted and dropped by the DACS machine before connecting the T-1 carrier to the telephone exchange.

【0125】

Consider the case of connectivity in Figure 17. The momentary constraints of the communication processor CP2 to the communication processor CP4 are quite different from the momentary constraints of the communication processor CP3 to the communication processor CP5. Available services from communication processor CP2 to communication processor CP4 are reduced accordingly. The communication processor CP2 may use a dedicated DS0 line. Another possibility is ISDN interface. Each has its own characteristics, sloops and protocols. Communication processors can interact with each other on the collective side of the network. Artificial intelligence software is important for this, and it records which price list services are connected to which network port by all their capabilities and limitations.

【0126】

In addition, the communication processor must select the error correction code by maintaining the attributes of the history file network such as line information, bit error rate, and error free time, and as a result, minimize the transmission overhead. The selected code is sent to the packet control part of the frame. Expert rules are created to take error statistics into account. If the bit error rate is at least 10-5 or less, or the error free time is 10 or less, then ECC should not be used. It should be noted that error distribution has a greater effect than the size of BER when choosing the right ECC. The communication processor is a convenient analogy similar to meteor burst communications, and the communication processor can utilize good propagation time up to the transmission rate by reducing the overhead. Figure 18 is an example of error data statistics that can be collected by a communication processor. Figure 18 shows what size block affects the quality of transmission. The auto-repeat request is a measurable parameter. This protocol can also record the number of times a given protocol must intervene to replace the error bit. The communication processor can record these statistics and can be modified to a protocol that selects the maximum data bandwidth and reasonably good transmission.

【0127】

Expert communication processing (AI) needs to take error statistics into account. The key factor in the performance of Slooput is the quality of the connection. Bit error rates alone are not enough to select the correct error correction code (ECC). In addition to BER, error free seconds (EFS) and error distribution through packet size are required. It is important that the communication processor measures these values and keeps a history of these for each network port. The choice of block size is also determined with ECC.

【0128】

Consider the following example to illustrate. A derivative key factor required for the calculation is the sloops per block size, which must include the factorization of dynamically grown block growth due to its addition to ECC. In FIG. 18, for example, the block size grows geometrically so that it falls within the candidate ECC (other ECCs grow mathematically or linearly). Figure 18 shows the history of the network port. This statistic averages 50% A block of size 116 [k] contains a sequence of one of seven consecutive error bits. If the ECC has the ability to correct 6 error bits in a sequence containing consecutive sequences, then the ECC is selected with a smaller block size of 48 [k] suitable for the current conditions. ECC applies only to control information. Audio / video data does not receive ECC. However, error statistics on control information apply to audio and video data as well when calculating block size. If ECC is no longer valid, the connection needs to be evaluated dynamically, so the call may be terminated before completing normal service due to poor performance. When bandwidth space is available, it may be called any number of times per hour for the purpose of collecting statistics. The user does not necessarily have to make multiple calls in this way, and may be based on previous usage patterns. Statistics about this call attempt are stored and used to calculate ECC-style success as a prospect for future call attempts.

【0129】

The outline of the network management plan shows the key elements of the plan.

【0130】

(1) Descriptor table for the characteristics of each network port side interface, all parameters and charge service.

【0131】

(2) A service table job that contains all the parameters that affect the service.

【0132】

(3) Service table requirements queued by arrival time, bandwidth requirements and connection complexity (suggested to service channel).

【0133】

(4) Quality performance table for each network workport (including measured performance statistics and calculated statistics).

【0134】

(5) Execute each request in the table (add a new channel) Calculation of the request service and customized request for each user in the performance table.

【0135】

(6) Calculation of the limit for the performance table of the set frame for each network interface port (the system is stopped).

【0136】

(7) A model of each current set port with data collected to plan bandwidth conditions and calculate bandwidth allocation. Develop all the primary parameters required by the subprocess.

【0137】

A feature of the present invention is that the communication processor of the present invention shares the video subprocessor in the communication processor with the user. The method of sharing is not important, for example, analog is switched to one A / D or discrete A / D for each connection between the workstation and the user. However, when a user processes video using weighting or compression, one communication processor is used, which has the function of being subordinately supported by the video processor for all users.

【0138】

The communication processor of the present invention uses a video subprocessor to perform work of selecting pixels from each user according to the specifications determined by the communication processor. The video subprocessor is always working to avoid maximizing bandwidth so as to leave the transmission capacity available to the next user. The video is compressed or weighted depending on the conditions. Weighting means that the pixel-allocated dynamic range is reduced to the active dynamic range, and in fact the active dynamic range has less bits than the total allocated dynamic range. For example, if the 12-bit position represents the entire range and the pel indicates a variation in the 4-bit position range, only 4 bits are transmitted. A special protocol is used to contact the receiver and 4 of the 12 allocated bits are transmitted to the receiver. Also, each pel has a specific address on the video memory side of the workstation. Therefore, no information is transmitted for vertical or horizontal synchronization. Communication processors need little to worry about compression or techniques for such compression. It can be compressed by hardware or by software, and the type of compression is not important in the present invention.

【0139】

Also note that U.S. Patent No. 1 Unlike No. 4,733,299, the system of the present invention has the same type of traffic. The communication processor is not related to the interlaced scan video or the non-interlaced scan video, but to the video data. The workstation may use interlaced scanning images and other non-interlaced images. All users of the communications processor are running video conferencing regardless of scan format. The basic video conferencing structure consisting of combined side frames is the same for all users, except that one user has a much larger bandwidth than the other user has allocated. There are several reasons for this, but once connected, bandwidth is allocated during the session. The bandwidth of either audio or video may change momentarily, but this is not a factor for the user. Each user has a channel in the join-side frame that can be decoded by other communication processors. Communication processors ideally use high-speed digital carriers such as T1 and other types of carriers on the coupling side, but these carriers are DS1, DS2 by transmission devices such as digital access cross-connect, for example. Or, by being treated as a DS3 carrier, it cannot be connected to a voice switch or data switch. Nor can these be processed as variable frequency divider signals for the T1 service. If the service on the coupling side is a T1 frequency divider signal, this must be a fixed frequency divider service.

【0140】

Consider whether an image can be focused on the end of an optical fiber cable connected to a communication processor using a series of lenses. The quality of optical fiber cables is different from that of cables that carry high-speed binary data. Optical fiber cables need to maintain a linear spatial image of the focal plane without strict attention to time dispersion. There are fiber lenses in which the large focal plane gradually tapers to the diameter of the fiber lens to reach the diameter of the optical fiber cable. The optical fiber cable ends with a communication processor in the lens array and electronic shutter. The shutter is positioned to project an image onto a single planar out-lens (composite prism), a facet with multiple surface inputs. The output of the composite prism is a focal lens array into the focal plane detector. FIG. 12 shows the weighting.

【0141】

The video controller selects electronic shutters in multiple sequences, collects images from the focal plane detector (CCD) in analog format, and converts the signal to digital format. The high-speed performance of the CCD allows the user to be serviced asynchronously, which is advantageous in terms of cost, but it is not necessary for the performance of the communication processor.

【0142】

As described above, the present invention has been illustrated and described based on the optimum embodiment thereof, but various changes may be made to the detailed configuration without departing from the spirit and scope of the present invention.

【0143】

[Effect of the invention]

As described above, according to the present invention, there is a communication processor that has a digital bus and elements are combined to interconnect these elements, and a workstation interface and a video processor that process video information and audio information at the workstation level. And the audio processor, the channel frame processor connected to the digital bus and controlling the communication via the digital bus, and the bandwidth between the audio information signal and the video information signal on the digital bus connected to the digital bus. A static audio / video multiplexing processor that is dynamically allocated is provided, and a workstation interface, a video processor, and an audio processor are interconnected, and a digital signal and an analog signal pass between them via a digital bus. By sending digital information, the video interface from the work station to the communication processor can be simplified at low cost, so it is possible to communicate easily and reliably with other work stations in the audio-video communication processor system. This makes it possible to significantly improve communication between users.

[Simple explanation of drawings]

[Figure 1]

FIG. 1 is a Venn diagram showing various devices of a transmission device. The intersection of the circles indicates the transmission device, and the ellipse indicates the communication processor. Communication processors with various types of network devices are used by the network carrier to provide fee services, so the communication processor must be able to select the appropriate algorithm with all its advantages and constraints. Must be. The communication processor then formulates the channel packet according to the service requested by the user by processing within the constraints by taking advantage of these advantages. The communication processor must perform the same kind of work as the transmission device and at the same time add audio and video processing to these functions, which makes the communication processor a new class of device. In the field of telecommunications, DCE is a data circuit-terminating equipment, which is owned by a telecommunications company. DTE is a data terminal device and is owned by a user who receives services from a telecommunications company. The distinction between DCE and DTE applies to equipment as such. Since this is a new class of equipment, market usage determines whether this new class of equipment is DTE or DCE. This is because this new class of equipment has both DTE and DCE. Figure 1 shows a new machine-class audio / video communication processor.

[Figure 2]

Figure 2 is a schematic diagram showing the communication process sign tough ace at the workstation level. It consists of a bezel that includes an electronic device that creates forward video, processes transmitted and received audio, and receives field back and remote video. The communication processor sign Tough Ace communicates with a communication processor that has control data and receives data and session data for user interface such as menu selection. Software and interface adapter cards that allow workstations to interact with communication processors are not shown. The bezel directly interacts with the communications processor.

[Fig. 3]

FIG. 3 is a schematic diagram showing how the video weighting plan can be structured as shown in the display area of FIG.

[Fig. 4]

Figure 4 is a schematic diagram showing a channel frame in three data formats.

[Fig. 5]

FIG. 5 is a schematic diagram showing a data output structure on a set port of a communication processor.

[Fig. 6]

FIG. 6 is a functional diagram showing the high-level function and interface of the communication processor. As shown, this is a preferred embodiment. From the workstation side, the communication processor is considered to be a DTE. On the network side, there is a network connection called a collective port. Collective frames are transmitted from each port. Each assembly frame contains a channel frame from the workstation. The channel frame enters one port, which has no destination at this point, but is routed or bridge-backed to another port for another destination. The information needed to do this must be returned to the origin communication processor for proper set-up and adjustment.

[Fig. 7]

FIG. 7 is a block diagram showing the basic concept of how to connect the functions of the communication processor to the workstation group and the remote communication network. Give the concept and title of the network.

[Fig. 8]

FIG. 8 is a block diagram showing a preferred embodiment of the present invention, which is integrated into a larger system unit as described by connecting with other drawings.

[Fig. 9]

FIG. 9 is a block diagram showing a unique image processing configuration that can be applied to a communication processor.

[Fig. 10]

FIG. 10 is a block diagram showing the relationship between the termination of the fiber, the lens array, the electronic shutter, the multi-sided input lens, the single-sided output lens, the focal lens, and the focal plane detector.

[Fig. 11]

FIG. 11 is a schematic diagram provided for explaining the technique for interfacing between the workstation and the communication processor.

[Fig. 12]

FIG. 12 is an enlarged schematic diagram of FIG. 3 with the number of vertical lines and the number of horizontal lines added. This number comes from IBM's VGA graphics specification. However, the purpose of this figure is to emphasize, and each workstation must use a window, so the window is only a small part of the full screen. The bandwidth associated with the video description is not driven by this.

[Fig. 13]

FIG. 13 is a schematic diagram showing the ideal bandpass frequency response of the focal plane image detector. Part of the video processing integrates the frequencies from the low frequency limit to the high frequency limit into a single value representing a pixel. This image is called the detector frequency integration scale.

[Fig. 14]

FIG. 14 is a schematic diagram showing the format of the signal inserted between the communication processor and the workstation using the basic technique shown in FIG. The purpose of this figure is to show that no complex multiplexing or optical fiber configuration is required to perform this function. Simple frequency division multiplexing on RG cables with limited bandwidth is sufficient.

[Fig. 15]

FIG. 15 is a block diagram showing possible techniques for processing audio signals. This method first treats audio as 64 [kbps] PCM with either mu or a weighting before compressing or weighting the signal when the carrier is at speeds above T1 and the activity is light. That is. If the bandwidth does not need to be reduced, the audio signal may still be subject to TASI (voice insertion) processing, although the bandwidth of the voice signal does not need to be further reduced. TASI processes with video processing sex in exchange for the bandwidth allocated between audio and video.

[Fig. 16]

FIG. 16 is a schematic diagram showing how voice is replaced with a channel packet's flux signal or text. Focus represents the audio band of an image. Complex images are advantageously transmitted from workstation to workstation using the features of the present invention.

[Fig. 17]

FIG. 17 is a schematic diagram showing how to interconnect communication processors.

[Fig. 18]

FIG. 18 is a graph showing the change in error bit with respect to the block size.