JP2004254341A - Compression method of dynamic picture information and system thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a compression method of dynamic picture information, which can perform compression processing for a data permitting a forecasting code of an image information at a high compression rate and a high speed, as well as helps improvement in image quality, and to provide its system. <P>SOLUTION: A first mode of the present invention performs block dividing of an image in a frame, before an inter-frame compression processing and approximates (replaces) the each split block as a single plane defined by three elements of a pixel in the block. A second mode mutually compares a pixel, in an original image with a pixel in an image extended, after compression to output a difference information on each pixel; and if a pixel element which generates a larger difference than a parameter exists, a smaller block size is applied. A third mode, when making each spatially block-split I block distributed along a time base between each frame, does not insert the I block into a block position in a frame updated due to the occurrence of the difference between the frames with in a specified time period. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

  The present invention relates to a method and a system for compressing moving image information, which can compress data capable of predictive coding of image information at a high compression rate and at a high speed, and also improve the image quality.

  In the related art, it is customary to temporarily convert an image signal into another signal, and then perform coded transmission by assigning an appropriate code using the statistical properties of the converted signal. In this case, in the redundancy within one frame, for example, in an image of a regular pattern or a flat image, since the correlation between adjacent pixels is strong, the value of the pixel to be encoded next is reduced to some extent from the value of the already encoded pixel. So-called predictive encoding is performed, in which significant information compression is performed by extracting and encoding only the components that can be predicted and that cannot be predicted.

  In addition, for example, in a moving image such as a videophone, images of successive frames are very similar, and temporal changes are often limited, and such temporal redundancy uses prediction over frames. It can be removed by the inter-frame prediction coding. At this time, in general, a block code in which one code word is assigned to one symbol is adopted, one frame is divided into smaller pixel blocks, and the difference in luminance in each block is reduced. A so-called block encoding process, which is used for information compression, is employed.

  Further, a so-called entropy coding for realizing data compression by allocating an efficient code to a converted signal and a Huffman coding method as an efficient code generation method are known. A typical example is an arithmetic code, which divides a probability number straight line according to the appearance probability of a symbol sequence, and uses a binary decimal value indicating a position in the divided section as a code for the sequence. That is, code words are sequentially formed by arithmetic operations.

  In addition, conventionally, a three-step block coding system for efficiently coding an image signal consists of sampling, transforming and quantifying. In order to maintain the planar resolution and high-frequency components of the image signal at this time, it is usually necessary to perform sampling at twice the speed of the highest frequency component of the frequency.

On the other hand, in MPEG, in order to compress a moving image having a large amount of information, it is preferable to increase the encoding efficiency as much as possible. For this reason, in the related art, forward prediction (processing using a P frame) using only the encoded temporally past image signal as a prediction signal, and temporally future image signal in addition to the past image signal. Also, there is bidirectional prediction (processing using a B frame) that is used as a prediction signal. Further, in the conventional inter-frame prediction coding, a difference signal between an input image signal and a prediction image signal is transmitted, and on the restoration side, the transmitted difference signal is added to the already restored prediction image signal. Restore the original image. In this way, on the restoration side, restoration between frames becomes impossible unless a predicted image signal is prepared. Therefore, instead of using past and future image signals as prediction signals, an I-frame (a reference frame in which an image is completed only with frames) processed by intra-frame coding is adopted as a predicted image between frames. Is inserted into the frame sequence at regular intervals to cope with reproduction from the middle of an image or a data error.
None

  However, since the conventional image signal information compression method uses a complicated block coding method, it is necessary to perform high-speed compression processing on data that can be predicted and encoded, such as image information and audio information, at a high compression rate. Was difficult. In the moving image compression process, when normal difference information is compressed, that is, when A1 and A2 are expected to be close to each other and the value of A1 can be known before the information of A2, A2-A1 is set to a value close to 0. Since it is assumed that the occurrence probability is high and a conventional compression method using a Huffman code or an arithmetic code is used, if the possible values of A1 and A2 are 0 to n, the possible range of the difference A2-A1 is There are 2n + 1 ways, and 2n + 1 ways of Huffman code words must be prepared. However, there are actually n possible values for A2, and the n codes are not used locally when viewed locally, and are therefore redundant codes. Further, when the difference between the frames is made large, the image quality is greatly deteriorated, and there is a problem that a high quality image cannot be obtained.

  In addition, when the block size is increased, the compression ratio is improved, but the detail of the original image is lost and the image quality is deteriorated. In particular, in the case of an original image composed of a background of a fixed color and a thin line having a greatly different luminance, a phenomenon has occurred in which the thin line is completely lost.

  Furthermore, in MPEG, I-frames that are periodically inserted into a frame sequence are processed by intra-frame coding, so that the coding efficiency is lower than that of inter-frame coding, which is performed by taking the difference between frames. However, since the amount of generated information increases, for example, when a high-speed communication line cannot be used, there is a limit to the frequency of I-frame insertion. In addition, since the data amount of the I frame is 2 to 10 times as large as that of the other difference frames, this is a method that is contrary to the constant bit rate required for communication. That is, in the related art, since the I frame is periodically inserted into the frame sequence, the processing time becomes considerably long, and the display of the restored image is greatly delayed. Moreover, since the data amount of the data itself is large, the probability that an unrecoverable data error occurs in the I frame increases. Also, if the error makes it impossible to reproduce (decode) the I frame that is the "reference", the reproduction is interrupted until the next I frame unless special measures are taken. For example, when a data error occurs for some reason in image reproduction, the small effect is enlarged and affects many frames as a whole, and in the worst case, the reproduction is interrupted. Further, in the conventional method of inserting I frames at regular intervals, when starting reproduction from a frame at an arbitrary time position, first search for the nearest I frame by some means, reproduce an image therefrom, It is necessary to display the playback screen after reaching the frame at the target time position, and this I frame search is very troublesome. On the other hand, if special measures are taken to cope with the various adverse effects described above, the load on the decoding processing system naturally increases. In addition, in the reproduction process, the reproduction of the I-frame has a high load, and a function for satisfying the processing capability of the I-frame is required of the processing system.

  In view of the above, the present invention has been created in view of the above-described conventional problems, and can perform high-speed compression processing on data capable of predictive coding such as image information and audio information at a high compression rate. It is a first object of the present invention to provide a method and a system for compressing moving image information that can improve the image quality.

  It is also an object of the present invention to provide a method and a system for compressing moving image information in which details of an original image are not lost and image quality is not degraded even when the compression rate is improved by increasing the block size. The purpose of.

  Further, in order to prevent the influence of the occurrence of a data error at the time of reproduction from affecting the entire frame, thereby preventing the reproduction from being interrupted, and to start the reproduction from a frame at an arbitrary time position, as in the related art, First, a time-consuming processing operation such as searching for the nearest I-frame by some means and reproducing an image from the I-frame is omitted, and moving image information that can easily display a reproduction screen at an arbitrary time position is omitted. A third object is to provide a compression method and a system thereof.

For this reason, in the moving image information compression method according to the first aspect of the present invention, pixels in spatially adjacent frames or pixels in temporally adjacent frames are compared with each other to determine the pixel element. The difference information is output, and whether the output difference information is a part larger than the given parameter or the other part is stored in a bit map, and the parameter stored in the bit map is stored. Is a method of compressing moving image information in which redundant information is reduced by performing a compression process on information of a large portion.
Before the inter-frame compression processing, the image in the frame is divided into blocks, and each divided block is approximated (substituted) as a single plane defined by three elements of the pixels in the block. Has solved the above-mentioned problem.

  Before the inter-frame compression processing, the image in the frame is divided into blocks, and each divided block is approximated (replaced) as a single plane defined by three elements of pixels in the block. The above-described problem has also been solved by performing intra-frame compression processing using the plane as a parameter.

On the other hand, in the moving image information compression system according to the first aspect of the present invention, pixels in spatially adjacent frames or pixels in temporally adjacent frames are compared with each other to calculate the difference between pixel elements. Bit map information recording means for outputting information, and storing, in a bit map, whether the output difference information is a portion larger than a given parameter or a portion other than the given parameter; A moving image information compression system comprising: an information compression unit that reduces redundant information by performing a compression process on a portion of information larger than a parameter stored by the information recording unit;
Block approximation means for dividing an image in a frame into blocks and approximating (substituting) each divided block as a single plane defined by three elements of pixels in the block before the inter-frame compression processing. The above-mentioned problem has also been solved by performing intra-frame compression using the plane obtained by the block approximation means as a parameter.

  In the moving image information compression method according to the second aspect of the present invention, the intra-frame compression process compresses the entire image in blocks of nxm pixels (n and m are integers of 2 or more) according to the intra-frame compression. Then, the pixels in the original image and the image expanded after compression are compared with each other, and the difference information of each pixel element is output. If there is a pixel element in which a difference larger than the parameter is present, a portion including the pixel element The above-described problem has also been solved by repeating the operation of applying a smaller block size to the surrounding area and the designated minimum block unit.

  On the other hand, in the moving image information compression system according to the second aspect of the present invention, in the block approximating means, the intra-frame compression processing is performed by compressing the entire image in accordance with the intra-frame compression by nxm pixels (n and m are 2 or more Of the original image and the image decompressed and decompressed, and outputs difference information of each pixel element. If there is a pixel element having a difference larger than the parameter, The above-described problem has also been solved by repeating the operation of applying a smaller block to a portion including the pixel element and a portion surrounding the pixel element up to a designated minimum block unit.

  In the moving image information compression method according to the third aspect of the present invention, an I frame (a reference frame in which an image is completed only by a frame) processed by intra-frame encoding is used, and this I frame is spatially When distributing the divided I blocks in the time axis direction between the frames, a difference between the frames occurs within the designated period, and the I blocks are inserted into the updated block positions in the frames. By not doing so, the above-mentioned problem was solved.

  On the other hand, the moving picture information compression system according to the third aspect of the present invention employs an I frame (a reference frame in which an image is completed only by a frame) processed by intra-frame coding, and There is provided an I-block inserting means for spatially dividing the blocks and distributing the divided I-blocks in the time axis direction between the frames, and the I-block inserting means generates a difference between frames within a designated period. The above-described problem was also solved by not inserting an I-block at a block position in a frame to be updated.

  The basic structure of the moving image information compression method and the system thereof according to the first aspect of the present invention is as follows: an intra-frame image is divided into blocks in advance, and all of the divided blocks are divided into pixel element sizes of each block; This is to approximate (replace) a single plane defined by three data of the inclination of the block in the x direction and the inclination of the block in the y direction, and to efficiently perform the intra-frame compression processing using the plane as a parameter. .

  The basic structure of the moving image information compression method and the system thereof according to the second aspect of the present invention is as follows. The whole image is compressed into blocks of n × m pixels (n and m are integers of 2 or more) according to intra-frame compression. Compress and compare the pixels in the original image and the image expanded after compression with each other and output the difference information of each pixel element. If there is a pixel element having a difference larger than the parameter, the pixel element is included. The operation of applying a smaller block size to a part and its surrounding parts is repeated up to the specified minimum block unit, so that even if the block size is increased and the compression ratio is improved, the detail of the original image Is not lost and the image quality is prevented from deteriorating.

  Further, the basic structure of the moving picture information compression method and the system thereof according to the third aspect of the present invention is that an I-frame is spatially divided in advance, and the divided I-block is divided in the time axis direction between the respective frames. At the time of dispersing, no I-block is inserted at a block position in a frame that has been updated due to a difference between frames. Therefore, when an image is reproduced, a predetermined number of frames before one frame is completely completed when one image is completely reproduced. It is only necessary to start the playback from the desired time position and display the playback screen after reaching the frame at the target time position, and easily perform the playback screen at an arbitrary time position without performing a troublesome I frame search process. It can be displayed.

  In addition, at the time of moving image distribution, the distribution data amount becomes uniform over time in the distribution server and the data communication path, so that higher distribution performance can be obtained than in the case of distributing content using the conventional technology. Also, on the receiving / reproducing side, fluctuations in the amount of received data per unit time are small, so the amount of memory required for buffering can be reduced, and the reproduction load is constant, so that even systems with low capabilities can be stable. Playback that is possible. In addition, since the influence of the data error on the reproduction is small, it is possible to continue the reproduction ignoring the data error, thereby making it unnecessary to retransmit the data in the system on the distribution side and reducing the load on the distribution side. . Furthermore, moving image broadcasting by multicast distribution or the like can be easily realized.

  According to the method and system for compressing moving image information according to the first aspect of the present invention, by removing block conversion, data capable of predictive coding of image information is compressed at a high compression rate and at high speed. It is possible to reduce the deterioration of image quality and sound quality. In particular, in the related art, if the difference between frames is large, the image quality is severely deteriorated. However, according to the first aspect of the present invention, the deterioration of the image quality can be reduced. That is, according to the first aspect of the present invention, it is possible to give a linear change to the image quality without sharply deteriorating the image quality with respect to the intra-block threshold value. As a result, the communication bit rate can be easily adjusted without deteriorating the image quality, and the compression ratio can be improved by about -20% to 50% while obtaining the same image quality in appearance. . In addition, the adaptive Huffman compression processing and the adaptive arithmetic compression processing are to collectively process conventional difference information generation and Huffman coding or a series of predictive coding processing of difference information generation and arithmetic coding to improve codeword efficiency. Thus, data capable of predictive coding of image information can be efficiently compressed. In addition, the data reduced (compressed) according to the first aspect of the present invention defines a plane, and becomes a plane showing gradation when expanded.

  According to the moving image information compression method and system according to the second aspect of the present invention, even when the compression ratio is improved by increasing the block size, the detail of the original image is not lost, and the image quality is not deteriorated. Reduction can be achieved. In particular, even in the case of an original image including a background of a fixed color and a thin line having a greatly different luminance, a situation in which the thin line is completely lost can be reliably prevented.

  According to the moving image information compression method and system according to the third aspect of the present invention, when an I frame is spatially divided into blocks in advance, and when the divided I blocks are dispersed in the time axis direction between frames, Since no I-block is inserted at the block position in the frame that has been updated due to the difference between the frames, the reproduction is started before the predetermined number of frames at which one screen is completely completed when reproducing the image. Then, it is only necessary to display the reproduction screen after reaching the frame at the target temporal position, and it is possible to easily display the reproduction screen at an arbitrary temporal position without performing a troublesome I frame search process. In addition, at the time of moving image distribution, the distribution data amount becomes uniform over time in the distribution server and the data communication path, so that higher distribution performance can be obtained than in the case of distributing content using the conventional technology. Also, on the receiving / reproducing side, fluctuations in the amount of received data per unit time are small, so the amount of memory required for buffering can be reduced, and the reproduction load is constant, so that even systems with low capabilities can be stable. Playback that is possible. In addition, since the influence of the data error on the reproduction is small, it is possible to continue the reproduction ignoring the data error, thereby making it unnecessary to retransmit the data in the system on the distribution side and reducing the load on the distribution side. . Furthermore, moving image broadcasting by multicast distribution or the like can be easily realized.

First, a description will be given of a moving image information compression method according to a first embodiment of the present invention.
That is, according to the present invention, pixels in a spatially adjacent frame or pixels in a temporally adjacent frame are compared with each other to output difference information of pixel elements, and the output difference information is given. Whether a portion larger than the parameter P or a portion other than the parameter P is stored in a bit map, and information of a portion larger than the parameter P stored in the bit map is compressed to perform redundancy processing. A method of compressing moving image information that reduces unnecessary information,
Prior to the inter-frame compression processing, an image in a frame is divided into blocks, and each divided block is approximated (substituted) as a single plane defined by three elements of pixels in the block. .

  Before the inter-frame compression processing, the image in the frame is divided into blocks, and each divided block is approximated (replaced) as a single plane defined by three elements of pixels in the block. Intra-frame compression processing is performed using the plane as a parameter P.

  Further, information of a portion that is not larger than the parameter P stored in the bit map is processed (deleted) as a pixel having no change.

  Further, as an approximation method for forming a single plane defined by three elements of a pixel, an average or a least square method is used.

  In addition, in the intra-frame compression processing, the plane is defined by three data items: the size of the pixel element of the pixel in the block, the inclination of the block in the x direction, and the inclination of the block in the y direction.

  The information stored in the bit map is compressed by at least one binary image encoding method selected from the group consisting of run length, modified READ (MR, MMR), modified Huffman (MH), and JBIG. Is what is done.

  In addition, information of a portion larger than the parameter P is information compressed by an adaptive Huffman encoding process having a Huffman table of the number of predicted information.

  In addition, redundant information between frames is further reduced by entropy coding.

  This entropy coding has a Huffman table of the number of prediction information, and has an adaptive Huffman coding process of coding using one table selected based on the prediction information from among them, or an arithmetic table of the number of prediction information. Then, it is performed by an adaptive arithmetic coding process in which coding is performed using one table selected based on prediction information from among them.

  Then, the difference information of the pixel elements is used.

  The difference information is a difference output by comparing the pixel t and the pixel t−1 between frames.

  Further, difference information output by comparing each pixel obtained by treating n × m pixels (n and m are integers of 2 or more) between frames as one block is used.

  Further, n × m pixels (n and m are integers of 2 or more) in a frame are treated as one block, and information of a difference output by comparing the pixel t and the pixel t−1 between the frames is used. It is.

  In addition, in n × m pixels between frames, n is 2 to the Kth power (K is a natural number) and m is 2 to the K′th power (K ′ is a natural number).

  Further, before the inter-frame compression processing, an intra-frame compression processing in which the divided block size changes within the same frame is performed.

  Next, a moving image information compression system according to a first embodiment of the present invention will be described.

That is, according to the present invention, pixels in a spatially adjacent frame or pixels in a temporally adjacent frame are compared with each other to output difference information of pixel elements, and the output difference information is given. A bit map information recording means 4 for storing in the bit map whether the part is larger than the parameter P or a part other than the parameter P; and the parameter P stored by the bit map information recording means 4 An information compression means for reducing redundant information by performing compression processing of a large portion of information;
Block approximation means for dividing an image in a frame into blocks and approximating (substituting) each divided block as a single plane defined by three elements of pixels in the block before the inter-frame compression processing. And performs intra-frame compression using the plane obtained by the block approximation means as a parameter P.

  The information compressing means 5 processes (deletes) information of a portion not larger than the parameter P stored in the bit map information recording means 4 as a pixel having no change.

  Further, in the block approximation means, an average or a least square method is used as an approximation process for forming a single plane defined by three elements of a pixel.

  In the block approximation means, the plane is defined by three data items: the size of the pixel element of the pixel in the block, the inclination of the block in the x direction, and the inclination of the block in the y direction.

  In addition, the information stored by the bit map information recording means 4 is at least one binary image code selected from the group consisting of run length, modified READ (MR, MMR), modified Huffman (MH) and JBIG system. The information is compressed by the conversion process.

  The information compressing means 5 for compressing information of a portion larger than the parameter P performs adaptive Huffman encoding having a Huffman table of the number of predicted information.

  In addition, an entropy encoding unit 6 for reducing redundant information between frames is provided. The entropy encoding unit 6 uses one table selected based on the prediction information from a Huffman table of the number of prediction information. Huffman coding or adaptive arithmetic coding using a single table selected based on prediction information from an arithmetic table of the number of prediction information.

  The difference information stored in the bit map information recording means 4 handles nxm pixels (n and m are integers of 2 or more) in a frame as one block, and the pixel t and the pixel t- This is a difference output by comparing with 1.

  Hereinafter, an embodiment of a method and system for compressing moving image information according to the first aspect of the present invention will be described.

  FIG. 1 is a block diagram schematically showing a compression path of moving image information. For example, a composite analog signal output from an NTSC system device such as a video camera, a disc player, or a video cassette player is converted into a digital signal by the analog-to-digital converter 1 and represents one line of a video frame. And is stored in the buffer 2. It is described that an analog signal output from an NTSC system device is converted into a digital signal by an analog / digital converter 1 as a video signal, digitally output, and stored in a buffer 2. Is not limited to this. That is, the present invention efficiently compresses all video signals including general video signals output from a predetermined device.

  Further, as shown in FIG. 1, the encoder compressor 3 sequentially compares the pixel t and the pixel t−1 between frames, and determines whether the difference is larger than the given threshold value, the parameter P. Or a bit map information recording means 4 for storing the other part in a 1-bit bit map. The comparison between the pixel t and the pixel t-1 is performed based on a pixel element (luminance, pigment, or the like). Here, t means time, and a pixel of the current t frame (pixel t) and a corresponding pixel (pixel t) of the temporally immediately preceding t-1 frame (at a position in the frame). -1). In addition, information of a portion where the difference between the two pixels t and t-1 stored by the bit map information recording means 4 is larger than the parameter P is subjected to compression processing, and information of other portions is not changed between frames. Process (delete) as pixels. The information of the portion larger than the parameter P is compressed by the information compressing means 5 by the adaptive Huffman coding process having a Huffman table of the number of predicted information, for example. The encoder compressor 3 compares spatially and temporally adjacent pixels and outputs difference information to reduce redundant information between frames. For example, the encoder compressor 3 calculates prediction information from an arithmetic table of the number of prediction information. And entropy encoding means 6 for performing an adaptive arithmetic encoding process for encoding using one arithmetic table selected based on. The difference between the pixel t and the pixel t-1 can be regarded as an absolute value when the difference is not larger than the given parameter P. In addition, regardless of whether the difference is larger than the parameter P or not, the difference can be taken as an absolute value.

  Then, as shown in FIG. 2, after encoding (encoding), the block data for each frame is sent to the memory 10 of the bit map information recording means 4, where it is delayed by one frame time and is present as the immediately preceding frame. Then, it is filtered by a temporal filter 11 which is a time variable impulse response filter. After the filter, the current frame data 13 and the immediately preceding frame data 14 are checked for redundancy between frames by a compressor, and the difference is calculated. That is, each block encoded by the comparing means 12 is compared with the corresponding block of the immediately preceding frame. Each block is marked with a single bit that defines whether it is a changing block or a block that has not changed with respect to the previous block. This process creates a frame bitmap with one bit per block. At this time, the bit map for each frame is distinguished by comparison between frames.

  In the present embodiment, the basic principle is intra-frame compression without changing the block size. That is, as shown in FIG. 5, the intra-frame image is divided into blocks in advance, and all the divided blocks are subjected to the pixel element size z of each block, the inclination of the block in the x direction, and the inclination of the block in the y direction. Is approximated (substituted) with a single plane defined by the three data items. That is, in intra-frame compression, an image is first divided into blocks, and these divided blocks are replaced with a single plane for approximating the divided blocks. This plane can also be defined by three pixel elements of each block, for example, z: magnitude of luminance, x: inclination in x direction, and y: inclination in y direction. Further, it can be defined by the size z of the pixel element of the block, the inclination of the pixel element of the block in the x direction, and the inclination of the pixel element of the block in the y direction. Furthermore, it can be defined by the size z of the pixel element of the pixel in the block, the inclination of the pixel element between blocks in the x direction, and the inclination of the pixel element between blocks in the y direction. As an approximation method, for example, an average or least square method is applied. The data thus reduced (compressed) defines a plane, and when expanded, becomes a plane showing gradation. At this time, if the number of pixels constituting the block is s, the compression ratio in the frame is 3 / s. As the s increases, the compression ratio improves, but the image quality decreases. The size and shape of the block are nxm pixels, and n and m are arbitrary as long as they are integers of 2 or more. In the n × m pixels between frames, n may be 2 to the power of K (K is a natural number), and m may be 2 to the power of K ′ (K ′ is a natural number).

  Next, the basic principle of inter-frame compression in the present embodiment will be described.

  That is, as a first method of the inter-frame compression, the above-mentioned intra-frame compression is performed on the block at the same position in the next frame (t + 1) of the current frame t, z: the size of the pixel element, x: x of the block Slope in direction, y: z (t + 1), x (t + 1), and y (t + 1) are obtained in three elements of the slope in the y direction of the block. Then, a mean square error between z (t), x (t), y (t) and z (t + 1), x (t + 1), y (t + 1) is calculated and compared with the threshold k. As a result, if it exceeds the threshold value k, it is determined that there is a difference. Alternatively, z (t), x (t), y (t) and z (t + 1), x (t + 1), y (t + 1) are compared with respective thresholds kz, kx, ky, and the difference is determined by the threshold. If it exceeds k, it is determined that there is a difference. If it is determined that there is a difference, the block is marked on a bit map indicating the position of the block in the frame. One bit map is obtained in the former case and three bit maps are obtained in the latter case. This bit map is a sequence of 0s and 1s (binary data) and is compressed using run-length compression or the like. Also, the difference data Δz (t) = z (t + 1) −z (t), Δx (t) = x (t + 1) −x (t), Δy (t) = y (t + 1) −y (t) are Entropy compressed. Note that the first method does not perform decompression, so that the calculation is light, but a calculation error may accumulate.

  As a second method of inter-frame compression, data compressed according to the basic principle of intra-frame compression is decompressed and data of each pixel element constituting a block is restored. Then, the root mean square error between the data of the pixel element at the same position in the block forming the same position of the next frame (t + 1) and the data of the restored pixel element is calculated and compared with the threshold value k. . As a result, if it exceeds the threshold value k, it is determined that there is a difference. If it is determined that there is a difference, the block is marked on a bit map indicating the position of the block in the frame. This bit map is a sequence of 0s and 1s (binary data) and is compressed using run-length compression or the like. Also, the difference data Δz (t) = z (t + 1) −z (t), Δx (t) = x (t + 1) −x (t), Δy (t) = y (t + 1) −y (t) are Entropy compressed. In the second method, since the expansion is performed, the calculation is heavy, but the calculation error is not accumulated.

  As a third method of inter-frame compression, a root-mean-square error of data of pixel elements at the same position in a block constituting a block at the same position in the current frame (t) and the next frame (t + 1) is calculated, and a threshold is calculated. Compare with k. As a result, if it exceeds the threshold value k, it is determined that there is a difference. If it is determined that there is a difference, a difference ΔP from the data of the pixel element at the same position in the block constituting the block at the same position in the next frame (t + 1) is calculated. Do. Then, it is marked on a bit map indicating the position of the block in the frame. This bit map is a sequence of 0s and 1s (binary data) and is compressed using run-length compression or the like. The difference data ΔP is subjected to entropy compression. In the third method, since the compression is performed after the difference determination is performed, the amount of calculation is the smallest, and no calculation error is accumulated.

    The 1-bit bit map information stored by the bit map information recording means 4 is a binary image encoding such as run-length, modified READ (MR, MMR), modified Huffman (MH), JBIG, etc. Information is compressed by the method. Specifically, in the case of run-length encoding, in general, a binary document image handled by facsimile or the like often has a certain number of white pixels or black pixels appearing to a certain extent, and white or black is continuous in one-dimensional direction. A so-called run, which is a lump of pixels, is used as an encoding unit, and the length of the number of consecutive pixels is encoded as a run length. For example, in a digital facsimile using a public telephone network, it is customary to use a modified Huffman code configured separately for black and white for a run length model.

  Further, in the case of the modified Huffman coding (MH), this is adopted as a one-dimensional coding method for facsimile transmission which reads at, for example, an image density of 8 pixels / mm and obtains black and white pixel information of 1728 pixels per scanning line. The MH code represents a run length which is the length of the continuous cluster of white pixels (white run) or the cluster of black pixels (black run), and the occurrence of a white run or black run of a certain length. The principle of data amount compression is to assign a variable length code using the probability that the probability has a statistical bias.

  In the case of the modified READ (MR, MMR), this is adopted as a standard of the two-dimensional encoding system in addition to the one-dimensional encoding system, and in the case of the MR, one-dimensional encoding is performed. Later, up to one continuous line at a standard resolution and up to three continuous lines at a high resolution are two-dimensionally encoded. In the case of MMR, the MR encoding method is infinite for both the standard resolution and the high resolution. It is set.

  The basic structure of the moving image information compression method and system according to the present invention is to reduce redundant information between frames by comparing spatially and temporally adjacent pixels (pixels) and outputting difference information. It is in. That is, as shown in FIG. 2, the pixel t and the pixel t−1 between the frames are sequentially compared, and the difference between the pixel t and the pixel t−1 is larger than a given parameter P which is a given threshold value, or at other portions. Is recorded as 1-bit bit map information. Then, the entropy coding unit 6 predicts a code appearing in each of the frames and between the frames, and outputs a slight deviation from the prediction to reduce redundant information. At this time, it is well known that, when coding transmission is performed by performing coding allocation, the average code length per pixel does not become less than the average information amount (entropy).

  Hereinafter, the algorithm of the adaptive Huffman coding method will be described. Adaptive Huffman coding aims to increase the efficiency of codeword generation by performing a series of predictive coding processes of difference information generation and Huffman coding. In the conventional Huffman encoding process, a codeword is usually generated using one Huffman table, and in the dynamic Huffman encoding process, the Huffman table is updated every time one word is encoded. On the other hand, in the adaptive Huffman coding, a Huffman table (code table) of the number of prediction information is provided, one table is selected from a plurality of tables by a table selector based on the prediction information, and coding is performed using the table. Perform As a result, it is possible to efficiently compress data that can be predicted and encoded, such as image information and audio information.

  Next, the algorithm of the adaptive arithmetic coding method will be described. Adaptive arithmetic coding aims to increase the efficiency of codeword generation by processing a series of predictive coding processes, namely, difference information generation and arithmetic coding. In a conventional arithmetic coding process, a codeword is usually generated using one occurrence probability table, and in a dynamic arithmetic code, the occurrence probability table is updated every time one word is coded. On the other hand, in the adaptive arithmetic coding, an arithmetic table (decoding table) of the number of prediction information is provided, one table is selected from a plurality of tables by a table selector based on the prediction information, and coding is performed using the table. I do. As a result, data capable of predictive coding of image information can be efficiently compressed.

  As the adaptive transform coding for a moving image, a method is employed in which a transform coefficient is scaled by feedback control using the sufficiency of a transmission path buffer memory and then coded. In this case, a threshold for determining an insignificant coefficient not to be encoded based on a histogram of transform coefficients for a representative image, a run-length code for encoding the continuity of the insignificant coefficient, and an adaptive Huffman encoding table for encoding a significant coefficient are obtained. , An encoding method based on this is adopted.

  As shown in FIG. 3, the specific configuration of the predictive encoding circuit is such that the analog-to-digital converted image input data is delayed halfway, and the value of the previous data (the data with the optimal delay) is stored in a table. At the same time as being sent to the selector and encoded, the difference is obtained by comparing the image input data with that transmitted directly to the encoder. In the table selector, a code table is selected based on the prediction information in accordance with the image input data and sent to the encoder, where the image input data is subjected to information compression to obtain an adjusted code word.

  Then, as shown in FIG. 4, the specific structure of the predictive decoding circuit is that the codeword is transmitted to the decoder, and at the same time, the directly transmitted codeword is once sent to the table selector, where the prediction information is transmitted. A decoding table is originally selected and sent to the decoder, and the difference between the decoded table value and the already decoded pixel value is obtained to obtain adjusted image output data.

  Further, a description will be given of a moving image information compression method according to the second embodiment of the present invention.

That is, according to the present invention, pixels in a spatially adjacent frame or pixels in a temporally adjacent frame are compared with each other to output difference information of pixel elements, and the output difference information is given. Whether a portion larger than the parameter P or a portion other than the parameter P is stored in a bit map, and information of a portion larger than the parameter P stored in the bit map is compressed to perform redundancy processing. A method of compressing moving image information that reduces unnecessary information,
Before the inter-frame compression processing, an intra-frame compression processing in which the divided block size changes within the same frame is performed.

  The intra-frame compression process compares the pixels of each block with each other while changing the divided block size, and outputs difference information of pixel elements. When the difference information is larger than the parameter P, the pixel element The smaller block size is applied to the portion including.

  Further, when the difference information of the pixel element is larger than the parameter P, the application of the smaller block size is repeated.

  Further, an image in a frame is divided into blocks, each divided block is approximated (substituted) as a single plane defined by three elements of pixels in the block, and the plane is used as a parameter P. It is.

  In addition, in the intra-frame compression processing, the plane is defined by three data items: the size of the pixel element of the pixel in the block, the inclination of the block in the x direction, and the inclination of the block in the y direction.

  In the intra-frame compression process, the whole image is compressed in blocks of nxm pixels (n and m are integers of 2 or more) in accordance with the intra-frame compression, and pixels in the original image and the image expanded after compression are mutually separated. An operation of outputting difference information of each pixel element by comparison and applying a smaller block size to a part including the pixel element and a part around the pixel element when a pixel element having a difference larger than the parameter P exists. Is repeated up to the designated minimum block unit.

  In addition, if there is no change in the block size as a result of the intra-frame compression processing, the inter-frame compression processing is performed.

  When the block size changes in a direction to increase, the data of the block is output as it is without taking the difference.

  Further, when the block size changes in a direction to decrease, a difference from the previous decompressed data is obtained for each part, and the difference is compressed by the corresponding block size.

  Next, a moving image information compression system according to a second embodiment of the present invention will be described.

  That is, according to the present invention, in the block approximation means, the intra-frame compression process compresses the entire image in blocks of nxm pixels (n and m are integers of 2 or more) according to the intra-frame compression, and compresses the original image and the compressed image. The pixels in the image expanded later are compared with each other, and the difference information of each pixel element is output. When there is a pixel element in which a difference larger than the parameter P exists, a part including the pixel element and a part around the pixel element are included. , The operation of applying smaller blocks is repeated up to the designated minimum block unit.

  If the block approximating means does not change the block size as a result of the intra-frame compression processing, the inter-frame compression processing is performed.

  Further, when the block approximating means changes in a direction in which the block size increases as a result of the intra-frame compression processing, the data of the block is output without taking the difference.

  If the block approximation unit changes in a direction in which the block size decreases as a result of the intra-frame compression processing, a difference from the previous decompressed data is obtained for each part, and the difference is compressed using the corresponding block size. Things.

  Hereinafter, a method and system for compressing moving image information according to a second embodiment of the present invention will be described with reference to FIGS.

  As described above, when the block size is increased, the compression ratio is improved, but the detail of the original image is lost, and the image quality is deteriorated. Particularly, in the case of an original image composed of a background of a fixed color and a thin line having a greatly different luminance, a phenomenon occurs in which the thin line is completely lost. In order to solve this, the following method is used. For the sake of simplicity, the description is based on a premise (example) of a 16 × 16 pixel monochrome image.

  That is, as shown in FIG. 6, the entire image is compressed (expanded) in blocks of 16 × 16 pixels according to the above-described intra-frame compression. The pixels in the original image and the image expanded after compression are compared with each other to output difference information of each pixel element, and are compared with the threshold value d1. As a result of the comparison, when there is a pixel element having a difference exceeding the threshold value d1, the portion including the pixel element is compressed (expanded) by a block of 8 × 8 pixels (large circle portion in FIG. 8). At this time, the surrounding area is also compressed by 8 × 8 pixels. Next, the pixels in the original image and the image decompressed and expanded are compared with each other to output difference information of each pixel element, and are compared with the threshold value d2. As a result of the comparison, when there is a pixel element having a difference exceeding the threshold value d2, a portion including the pixel element is compressed (expanded) by a block of 4 × 4 pixels (the middle circle in FIG. 8). At this time, the surrounding area is also compressed by 4 × 4 pixels. Further, the pixels in the original image and the image expanded after compression are compared with each other, and the difference information of each pixel element is output, and is compared with the threshold value d3. As a result of the comparison, when there is a pixel element having a difference exceeding the threshold value d3, a portion including the pixel element is compressed (expanded) by a block of 2 × 2 pixels (small circle portion in FIG. 8). At this time, the surrounding area is also compressed by 2 × 2 pixels. According to such a method, it is possible to perform compression without losing the details of the original image while maintaining a high compression ratio.

  Next, as a result of the above-described intra-frame compression processing, the inter-frame compression processing when the compressed image shown in FIG. 7A for frame (t) and FIG. 7B for frame (t + 1) is obtained. The method will be described. That is, for 1 → 1 ′ and 2 → 2 ′ in which the block size does not change, the difference is obtained by any of the methods described in the inter-frame compression processing, and inter-frame compression is performed. As for 4 → 4 ′, which changes in the direction in which the resolution becomes coarse, 4 ′ is treated as a key block (key frame) that can be independently expanded without depending on the previous frame. No difference is taken at this time. That is, the data of the block 4 'is output as it is without taking the difference. As for 3 → 3 ′, which changes in the direction in which the resolution becomes denser, a difference from the decompressed data of 3 is obtained for each part, and the difference is compressed with a corresponding block size.

  Next, a description will be given of a moving image information compression method according to a third embodiment of the present invention.

  That is, the present invention employs an I frame (a reference frame in which an image is completed only by a frame) processed by intra-frame coding, spatially divides the I frame in advance, and divides the divided I block into respective blocks. It is distributed in the time axis direction between frames.

  When each of the spatially divided I blocks is dispersed in the time axis direction between frames, a difference between frames occurs within a designated period, and an I block is updated at a block position in the updated frame. Is not inserted.

Further, the image in the frame is divided into blocks in advance, and all of the divided blocks are defined by three data of the pixel element size of each block, the inclination of the block in the x direction, and the inclination of the block in the y direction. In addition to approximation (replacement) with a single plane, it is possible to adopt an I frame (a reference frame in which an image is completed only with a frame) processed by intra-frame coding and insert the I frame into a frame sequence. A method for compressing moving image information,
When an I-frame is divided into blocks in advance and the divided I-blocks are dispersed in the time axis direction between the respective frames, a difference between the frames occurs within a designated period, and the difference between the I-blocks is determined in the updated block position in the frame. Does not insert an I block.

Further, pixels in spatially adjacent frames or pixels in temporally adjacent frames are compared with each other to output difference information of pixel elements, and the output difference information is compared with the given parameter P. Whether it is a large part or another part is stored in a bit map, and information of a part larger than the parameter P stored in the bit map is compressed to reduce redundant information. Moving image information compression method,
An I frame to be processed by intra-frame coding (a reference frame in which an image is completed only by a frame) is adopted, the I frame is spatially divided in advance, and the divided I block is divided into frames in the time axis direction. In the case of dispersing the data into the blocks, the I-block is not inserted at the block position in the updated frame where the difference between the frames occurs within the designated period.

  Furthermore, a moving image information compression system according to a third aspect of the present invention will be described.

  That is, the present invention employs an I frame (a reference frame in which an image is completed only by a frame) processed by intra-frame coding, spatially divides the I frame in advance, and divides the divided I block into respective blocks. It has an I-block insertion means 7 for dispersing in the time axis direction between frames.

  Further, the I-block inserting means 7 does not insert an I-block at a block position in a frame where a difference between frames occurs within a specified period and is updated.

In addition, the in-frame image is divided into blocks in advance, and all of the divided blocks are defined by three pieces of data of the pixel element size of each block, the inclination of the block in the x direction, and the inclination of the block in the y direction. A moving image information compression system having block approximation means for approximating (replacing) with a single plane,
I-blocking means 8 for spatially dividing an I frame to be processed by intra-frame coding (a reference frame in which an image is completed only by a frame), and distributing the divided I blocks in the time axis direction between the respective frames In this case, there is provided an I-block inserting means 7 for inserting an I-block into a portion other than the block position in the updated frame in which a difference between frames occurs within a designated period.

  Hereinafter, a moving image information compression method and a system thereof according to a third embodiment of the present invention will be described with reference to FIGS.

  That is, it is an encoding method for responding to image reproduction (decoding) from the middle or a data error occurring during reproduction. In this method, a compression algorithm that does not use a motion prediction / compensation technique for compressing three or more entire frames is a precondition. First, a frame coded by intra-frame prediction, that is, an I frame is spatially divided into one or a plurality of blocks, and these divided I blocks are dispersed in a time axis direction (I block formation). Note that the block size, division shape, and the like in such I-blocking can be arbitrarily changed and may be randomly selected.

  Specifically, as shown in FIG. 8, an I frame of 8 × 8 pixels is spatially divided into blocks of 2 × 2 pixels to form 16 I blocks, and these are arranged in a frame sequence at regular intervals. In this case, a portion where a difference output between frames occurs (a portion having a large amount of motion) and an I block (the amount of generated information of this block itself is larger than other difference frames) are overlapped. In such a case, useless I blocks are inserted and the amount of information is extremely increased, so that an unrecoverable data error occurs at the position of the inserted I frame. In order to avoid this, if there is no problem in the processing speed on the encoding side, an I block is not inserted into a block that has been updated (differential output) due to a difference between frames occurring within a specified period. It is.

  A specific encoding method will be described with reference to FIG. Here, as one example, an I frame of 8 × 8 pixels is spatially divided into 1 × 2 pixel blocks by the I blocking means 8 to form 32 I blocks. On the other hand, a frame of an image has a maximum block of 16 × 16 pixels, and one example is a moving image in which this block is composed of 8 × 8 pixels. Further, in FIG. 9, for convenience, the n + 11 to n + 32 are omitted.

  First, an I block indicated by a black portion in the figure is inserted in the horizontal direction in units of 1 × 2 pixel blocks. An object (a dark gray portion: a difference output block in which the object advances in the background) that needs to update (differential output) a block of up to 2 × 2 pixels from the upper right of the image moves to the lower left. Then, up to the (n + 3) th frame, normal I-block insertion processing is performed. On the other hand, the block is updated (differential output) by the object appearing at the upper right of the (n + 3) th frame, and as a result, the I block that is to be inserted into the (n + 4) th frame is not inserted (hatched portion). The light gray portion in the figure is a difference output block that returns to the original background because the object has moved. The processing in which the I block is not inserted appears in the (n + 7) th frame and the (n + 8) th frame in this case. That is, if there is one difference output block (light gray portion) that returns to the original background due to the movement of the object in the (n + 5) th frame in the (n + 7) th frame as a portion to be updated, a block unit of 1 × 2 pixels , Only one block on the right side is not inserted at that position. Also, a difference output block (light gray portion) that returns to the original background due to the movement of the object in the (n + 4) th frame and the (n + 5) th frame is a block of 1 × 2 pixels in the (n + 8) th frame in the horizontal direction as a portion to be updated. , An I block in units of 1 × 2 pixels is not inserted at that position. In this case, the time reference (near past) in which the I block is not inserted is the number of frames (8 × 8 ÷ 2 = 32 frames) necessary for inserting the I block at the positions of all the blocks. That is, the I block is not inserted into the block at the position updated (differential output) due to the motion of the object or the like within 32 frames. On the other hand, in order to start reproduction from a frame at an arbitrary temporal position, decoding is started from a predetermined number of frames before one screen is completely completed, and after a frame at a target temporal position is reached, the reproduction screen is started. Should be displayed.

FIG. 3 is a block diagram schematically illustrating a compression path of moving image information. FIG. 4 is a block diagram schematically showing a comparison path of information recorded in bitmap information recording means. FIG. 4 is an explanatory diagram schematically showing an encoding path. It is an explanatory view showing an outline of a decoding path. FIG. 9 is an explanatory diagram showing a plane defined by three data of a size z of a pixel element of a block, an inclination of the block in an x direction, and an inclination of a block in a y direction for approximating a divided image block. FIG. 13 is a plan view of an image illustrating an operation of applying a smaller block size to a pixel element and a surrounding area when a pixel element having a difference larger than a parameter exists. 5A and 5B show images for explaining inter-frame compression, in which FIG. 5A is a plan view of a frame (t), and FIG. 5B is a plan view of a frame (t + 1). FIG. 3 is a plan view showing a plurality of I blocks constituting an I frame. FIG. 11 is an explanatory diagram showing an insertion state of an I block between frames.

Explanation of reference numerals

DESCRIPTION OF SYMBOLS 1 ... Analog-to-digital converter 2 ... Buffer 3 ... Encoder compressor 4 ... Bit map information recording means 5 ... Information compression means 6 ... Entropy coding means 7 ... I block insertion means 8 ... I block formation means 10 ... Memory 11 ... temporal filter 12 ... comparing means 13 ... current frame data 14 ... immediate frame data

Claims (43)

Pixels in a spatially adjacent frame or pixels in a temporally adjacent frame are compared with each other to output difference information of pixel elements, and in a portion where the output difference information is larger than a given parameter. Moving image information for reducing redundant information by storing, in a bit map, whether or not there is a certain portion or the other portion, and compressing information of a portion larger than the parameter stored in the bit map Compression method,
Prior to the inter-frame compression processing, the image in the frame was divided into blocks, and each divided block was approximated (replaced) as a single plane defined by three elements of pixels in the block. A method of compressing moving image information that is a feature.   Prior to the inter-frame compression processing, the image in the frame is divided into blocks, and each divided block is approximated (replaced) as a single plane defined by three elements of the pixels in the block. 2. The method for compressing moving image information according to claim 1, wherein the intra-frame compression processing is performed using the parameter as a parameter.   3. The moving image information compression method according to claim 1, wherein information of a portion not larger than the parameter stored in the bit map is processed (deleted) as a pixel having no change.   4. The moving image information compression method according to claim 1, wherein an average method or a least square method is used as an approximation method for forming a single plane defined by three elements of pixels.   5. The intra-frame compression process, wherein the plane is defined by three data items: a size of a pixel element of a pixel in the block, an inclination of the block in the x direction, and an inclination of the block in the y direction. Video information compression method.   The information stored in the bit map is information compressed by at least one binary image encoding method selected from the group consisting of run length, modified READ (MR, MMR), modified Huffman (MH), and JBIG. A method for compressing moving image information according to claim 1.   7. The moving image information compression method according to claim 1, wherein information of a portion larger than the parameter is compressed by an adaptive Huffman encoding process having a Huffman table of the number of prediction information.   8. The moving image information compression method according to claim 1, wherein redundant information between frames is further reduced by entropy coding.   Entropy coding has a Huffman table of the number of prediction information, and has an adaptive Huffman coding process of coding using one table selected based on the prediction information from among them or an arithmetic table of the number of prediction information. 9. The method for compressing moving image information according to claim 8, wherein the method is performed by an adaptive arithmetic coding process of coding using one table selected based on the prediction information from among them.   10. The moving image information compression method according to claim 1, wherein difference information of pixel elements is used.   The moving image information compression method according to any one of claims 1 to 10, wherein the difference information is a difference output by comparing a pixel t and a pixel t-1 between frames.   The moving image according to any one of claims 1 to 11, wherein n * m pixels (n and m are integers of 2 or more) between frames are treated as one block, and difference information output by comparing respective pixels is used. Information compression method.   2. An n.times.m pixel (n and m are integers of 2 or more) in a frame is treated as one block, and information on a difference output by comparing a pixel t and a pixel t-1 between the frames is used. 13. The method for compressing moving image information according to any one of claims 12 to 12.   14. In the n × m pixels between frames, n is 2 to the power of K (K is a natural number) and m is 2 to the power of K ′ (K ′ is a natural number). Compression method.   15. The moving image information compression method according to claim 1, wherein an intra-frame compression process in which a divided block size changes within the same frame is performed before the inter-frame compression process. Pixels in a spatially adjacent frame or pixels in a temporally adjacent frame are compared with each other to output difference information of pixel elements, and in a portion where the output difference information is larger than a given parameter. Moving image information for reducing redundant information by storing, in a bit map, whether or not there is a certain portion or the other portion, and compressing information of a portion larger than the parameter stored in the bit map Compression method,
A method for compressing moving image information, comprising performing an intra-frame compression process in which a divided block size changes within the same frame before the inter-frame compression process.   The intra-frame compression process compares the pixels of each block with each other while changing the divided block size, and outputs difference information of the pixel element. If the difference information is larger than the parameter, the portion including the pixel element is included. 17. The moving picture information compression method according to claim 16, wherein a smaller block size is applied to the moving picture information.   18. The moving image information compression method according to claim 16, wherein when the difference information of the pixel elements is larger than the parameter, application of a smaller block size is repeated.   17. An image in a frame is divided into blocks, each divided block is approximated (replaced) as a single plane defined by three elements of pixels in the block, and the plane is used as a parameter. 18. The method for compressing moving image information according to any one of 18.   20. The intra-frame compression process, wherein the plane is defined by three data items: a size of a pixel element of a pixel in the block, an inclination of the block in the x direction, and an inclination of the block in the y direction. Video information compression method.   In the intra-frame compression processing, the entire image is compressed in blocks of nxm pixels (n and m are integers of 2 or more) in accordance with the intra-frame compression, and the pixels in the original image and the image expanded after compression are compared with each other. Output the difference information of each pixel element, and when there is a pixel element having a difference larger than the parameter, specify an operation of applying a smaller block size to a portion including the pixel element and a portion around the pixel element 21. The moving image information compression method according to claim 16, wherein the method is repeated up to the minimum block unit.   22. The moving image information compression method according to claim 16, wherein if the block size does not change as a result of the intra-frame compression processing, the inter-frame compression processing is performed.   22. The moving image information compression method according to claim 16, wherein when the block size changes in a direction to increase, the data of the block is output as it is without taking a difference.   22. The moving image information compression according to claim 16, wherein when the block size changes in a decreasing direction, a difference from the previous decompressed data is obtained for each part, and the difference is compressed with the corresponding block size. Method.   An I frame to be processed by intra-frame coding (a reference frame in which an image is completed only by a frame) is adopted, the I frame is spatially divided in advance, and the divided I block is divided into frames in the time axis direction. 25. The moving image information compression method according to claim 1, wherein the moving image information is distributed.   When distributing each of the spatially divided I blocks in the time axis direction between frames, an I block is inserted at a block position in the updated frame where a difference between frames occurs within a specified period. 26. The moving image information compression method according to claim 25, wherein the compression is not performed. An image in a frame is divided into blocks in advance, and all of the divided blocks are simply defined by three data items, ie, the pixel element size of each block, the inclination of the block in the x direction, and the inclination of the block in the y direction. A moving image that can be approximated (replaced) by one plane, and employs an I frame (a reference frame in which an image is completed only with a frame) processed by intra-frame encoding and inserts the I frame into a frame sequence A method of compressing information,
When an I-frame is divided into blocks in advance and the divided I-blocks are dispersed in the time axis direction between the respective frames, a difference between the frames occurs within a designated period, and the difference between the I-blocks is determined in the updated block position in the frame. Is a method for compressing moving image information, wherein I blocks are not inserted. Pixels in a spatially adjacent frame or pixels in a temporally adjacent frame are compared with each other to output difference information of pixel elements, and in a portion where the output difference information is larger than a given parameter. Moving image information for reducing redundant information by storing, in a bit map, whether or not there is a certain portion or the other portion, and compressing information of a portion larger than the parameter stored in the bit map Compression method,
An I frame to be processed by intra-frame coding (a reference frame in which an image is completed only by a frame) is adopted, the I frame is spatially divided in advance, and the divided I block is divided into frames in the time axis direction. A method of compressing moving image information, wherein an I-block is not inserted at a block position in a frame updated when a difference between frames occurs within a specified period when the image data is distributed. Pixels in a spatially adjacent frame or pixels in a temporally adjacent frame are compared with each other to output difference information of pixel elements, and in a portion where the output difference information is larger than a given parameter. Bit map information recording means for storing whether there is or not a part in a bit map, and performing compression processing of information of a part larger than the parameter stored by the bit map information recording means And a information compression means for reducing redundant information in the moving image information compression system,
Block approximation means for dividing an image in a frame into blocks and approximating (substituting) each divided block as a single plane defined by three elements of pixels in the block before the inter-frame compression processing. Wherein the intra-frame compression is performed using the plane obtained by the block approximation means as a parameter.   30. The moving image information compression system according to claim 29, wherein the information compression unit processes (deletes) information of a portion not larger than the parameter stored by the bit map information recording unit as a pixel having no change.   31. The moving image information compression system according to claim 29, wherein the block approximation unit uses an average or a least squares method as an approximation process for forming a single plane defined by three elements of pixels.   32. The block approximation unit according to claim 29, wherein the plane is defined by three data items: a size of a pixel element of a pixel in the block, an inclination of the block in the x direction, and an inclination of the block in the y direction. Video information compression system.   The information stored by the bit map information recording means is obtained by at least one binary image encoding process selected from the group consisting of run length, modified READ (MR, MMR), modified Huffman (MH), and JBIG. 33. The moving image information compression system according to claim 29, wherein the compression is performed.   The moving image information compression system according to any one of claims 29 to 33, wherein the information compression means for performing the compression processing of the information larger than the parameter performs the adaptive Huffman encoding processing having a Huffman table of the number of prediction information.   An entropy encoding unit for reducing redundant information between frames, wherein the entropy encoding unit performs encoding using one table selected based on the prediction information from a Huffman table of the number of prediction information. The moving image according to any one of claims 29 to 34, wherein Huffman encoding processing or adaptive arithmetic encoding processing for encoding is performed using one table selected based on prediction information from an arithmetic table of the number of prediction information. Image information compression system.   The difference information stored in the bit map information recording means treats n × m pixels (n and m are integers of 2 or more) in a frame as one block, and determines a pixel t and a pixel t−1 between the frames. The moving image information compression system according to any one of claims 29 to 35, wherein the difference is a difference output by comparison.   In the block approximation means, the intra-frame compression process is performed by compressing the entire image in blocks of nxm pixels (n and m are integers of 2 or more) in accordance with the intra-frame compression, and pixels in the original image and the image expanded after compression. The differences are compared with each other, and the difference information of each pixel element is output. When there is a pixel element having a difference larger than the parameter, a smaller block is applied to a portion including the pixel element and a portion around the pixel element. 37. The moving picture information compression system according to claim 29, wherein the operation is repeated up to a designated minimum block unit.   The moving image information compression system according to any one of claims 29 to 37, wherein the block approximation means performs an inter-frame compression process if the block size does not change as a result of the intra-frame compression process.   The moving image according to any one of claims 29 to 38, wherein, when the block approximation unit changes in a direction in which the block size increases as a result of the intra-frame compression processing, the data of the block is output as it is without taking a difference. Information compression system.   In the block approximation means, when the block size changes as a result of the intra-frame compression processing, a difference from the previous decompressed data is obtained for each part, and the difference is compressed with the corresponding block size. 39. A moving image information compression system according to any one of 29 to 38.   An I frame to be processed by intra-frame coding (a reference frame in which an image is completed only by a frame) is adopted, the I frame is spatially divided in advance, and the divided I block is divided into frames in the time axis direction. 41. The moving picture information compression system according to claim 29, further comprising an I-block inserting unit for dispersing the moving picture information into a plurality of video data.   42. The moving picture information compression system according to claim 41, wherein the I-block inserting means does not insert the I-block at a block position in the frame where the difference between the frames occurs within a specified period and is updated. The intra-frame image is divided into blocks in advance, and all of the divided blocks are converted into a single unit defined by three data of the pixel element size of each block, the inclination of the block in the x direction, and the inclination of the block in the y direction. A moving image information compression system having block approximation means for approximating (substituting) with a plane of
I-blocking means for spatially dividing an I-frame (a reference frame for which an image is completed only by a frame) processed by intra-frame encoding, and distributing the divided I-blocks in the time axis direction between the respective frames. A moving image information compression system comprising: an I-block inserting unit that inserts an I-block into a portion other than a block position in a frame in which a difference between frames occurs within a designated period and is updated.

Images