JPS63272272A - Picture compression system - Google Patents

Abstract

PURPOSE:To attain efficient transmission by assigning a specific code of a variable length code to a zero code so that a part with consecutive zero codes with zero amplitude difference at a value or over is subject to compression coding to a zero code corresponding to the number of consecutive zeros thereby reducing the quantity of codes with excellent picture. CONSTITUTION:A data is subjected to variable sampling density coding as to a first- order difference in the direction of line by a variable sampling density coding circuit 4, and a secondary difference being the predicted residue between the first-order difference of a picture element to be compression coding and the predicted value of the succeeding first-order difference based on the restored picture element after compression coding is obtained, and the quantizing code corresponding to the combination of the time difference and amplitude difference is obtained based on the quantization characteristic from the second-order difference. In the 2nd-order differential prediction quantization, the variable length coding is applied by a variable length coding circuit 27 so that in general the length becomes shorter sequentially from higher frequency of occurrence. The specific code of a variable length code is assigned to the zero code to apply variable length zero code compression coding so that the part having consecutive zeros in the amplitude difference is to be subjected to compression coding into the zero code corresponding to the number of consecutive zeros.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to an image compression method used for S-image monitoring in which image information from a TV camera or the like is transmitted via a telephone@R or the like by narrowband transmission.

[Background II Technique 1 Is the telephone line currently in use a voice telephone line? +? (0.3 to 3.4 KHz) is common, and a modem (modulator/demodulator) can transmit codes at a communication speed of 9,600 bits/second, but in the case of image monitoring and security systems, monitoring N As the number of images increases, there is a need for a low-cost image compression method with a high compression ratio.

Generally, an I 1' V camera or the like is used as a TV camera for image monitoring, and therefore still images are often obtained using multi-level dual image signals. In addition, vector quantization and motion compensation used in TV conference systems and 64K bit digital line TV telephones (JP-A-58-101581, JP-A-58-
No. 133088) A video transmission system using A4 is expensive and difficult to use due to cost constraints. On the other hand, still image transmission systems such as facsimiles are currently being developed into multi-level and color formats, and the international standardization of 1 bit/pel (pixel) compression methods for color images for 15DN 64 Kbit lines has been established by ISO and CCr TT. Vector quantization (Reference: Institute of Electronics and Communication Engineers technical report IE8
4-18 p, 9), block coding (Reference: Electronic Communication T Complete Journal '87/I Vol, 170-B No.
, i p, 66, Journal of the Institute of Image Electronics Engineers, Vol. 15, No. 4, p. 2
25), sequential reproduction encoding (Reference: Journal of the Institute of Electronics and Communication Engineers '87/I Vol., +70-B No.

i p, 105), cosine transformation (Reference: Journal of the Institute of Electronics and Communication Engineers '86/10 Vol. 169-B No. 10
) etc. are being considered.

Among these, 10-tsukukari coding and sequential reproduction coding are said to be methods with a small amount of calculation, and they are coded in a hierarchical manner by sequentially changing block rises and sampling intervals to produce a clear image. Each layer is compressed and variable length encoded.

The present inventors developed variable sample density encoding (Electronic Communication T Zenshu '7) as a method suitable for simpler microcomputer processing with a compression rate of about 2 to 3 bits/pet.
5/2 Vol, 58-^No2. p.97) to develop a predictive coding method for second-order differences and a variable pixel compression transmission method combined with zero-code compression, which are described in Japanese Patent Laid-Open No. 61-296866.

Since this conventional method is applied as part of the configuration of the present invention described later, it will be described in detail here.

PtS7 and FIG. 8 show a configuration for performing variable sample density compression/expansion of inter-line prediction residuals or DPCM compression/expansion of inter-line prediction residuals, and both image compression section A shown in FIG. The difference circuit 1 calculates the inter-line prediction residual using the sample image data (pixels of the current line) and the prediction data from the prediction circuit 2 (the restored pixels of the previous line that has been compressed and encoded), and calculates the inter-line prediction residual. The difference is passed through a line buffer 3 to a variable sampling density encoding circuit 4, which compresses the quadratic side residual into a quantization characteristic code in a direction perpendicular to the inter-line prediction residual, that is, in the line direction. The quantization characteristics are the combination of the amplitude difference value (quantization level) and time difference value (inter-sample VA) of the secondary side residuals as shown in Table 1, and the corresponding 4 1) own This is a table for mutual conversion with codes.

As shown in FIG. 9, the points plotted on the coordinate axes of the amplitude difference value and the time difference value are quantization points, and these are connected by straight lines in the figure.

In Table 1, 15 quantization points are taken as shown in Figure PtSi2, and these quantization points are expressed by quantization characteristic codes encoded into 4b1 (.

These quantized characteristic codes are combined into two 8-bit codes that are modulated and transmitted by a modem, but there are consecutive zero codes indicating that the secondary side residual of the quantized characteristic code is zero. Therefore, the zero code compression circuit 5 compresses the zero code (
A transmission code (12 to FE) that indicates the loss of zero codes or the number of lines containing only zero codes when zero codes are continuous due to the compression characteristics (zero lines), and a 4-bit quantization characteristic code other than zero codes ( In the case of an 8-bit transmission code (0 to EF), it is compressed into a zero-compressed code and transmitted. This zero code zero line compression characteristic depends on the number of zero codes and 8
This is a table of mutual conversion with bit code.

The quantization characteristic code is decoded by the variable sample density decoding circuit 6, taken in to the interpolation synthesis circuit 8 via the line buffer 7, and synthesized with the value of the line buffer 9 of the feedback loop inside the prediction circuit 2, The signal is input to the prediction circuit 2 for prediction of the next line. Here, when the time difference values (sample intervals) are thinned out at regular intervals, the interpolation synthesis circuit 8
When each pixel is thinned out to 64 x 64 pixels, 256
This is for restoring to an image of ×256911 elements, 2
Even if only the changed area is thinned out every 4 pixels and sent to a 56x256 pixel background image, only the changed area is interpolated to restore a coarse image and combined with the 256x256 pixel background image.

On the receiving expansion unit B side, code data is decoded into a quantized characteristic code by the zero code decoding circuit 10 of the expansion unit shown in FIG. The interpolation synthesis circuit 1 restores the inter-line prediction residual and sends this inter-line prediction residual via the line buffer 12.
3, and a decoding loop using the prediction circuit 15 and line buffer 14 restores one line of image data.

Here, if all the time difference values of the above quantization characteristics are set to 1, the variable sampling density coding becomes the same as DPC,M, so the variable sampling density coding method for the inter-line prediction residual (Table 1) quantization characteristics Table 2 Zero-code compression DCPM for inter-fin prediction residuals or variable sample density encoding method for inter-line prediction residuals (variable sample density predictive encoding method) described above requires most of the calculation amount per pixel. DP
Since it uses quantization similar to CM, the quantization characteristics can be fixed, and it can be said to be simpler than other compression methods.

Furthermore, since the prediction method uses orthogonal second-order differences between lines and within lines, the prediction residuals are concentrated to zero compared to the first-order difference prediction method or the planar prediction method, making it possible to reduce the quantization error and reduce the amplitude. Even if the dynamic range of the difference value is small, in practice,
It has the characteristic that overload distortion is not noticeable. In addition, since the second-order difference is concentrated at zero, the effect of zero code compression becomes large, and since the prediction residual is calculated for each line with respect to the sample (current image), errors propagate in the direction between lines. It has the characteristics of good image quality with less double-headed business that is unique to variable sample density encoding.

Furthermore, the effect of equal-sign compression is limited to cases where zero codes are consecutive, so the effect is small when compressing the entire image, and the effect is large when compressing and encoding only changing pixels, which will be described later. Then it becomes 8.

=r The quantization characteristics of variable sample density encoding are as shown in the ts9 diagram, by making the quantization points overlap in the range of quantization values with different time difference values (sample intervals), the quantization value and time difference value ( By encoding into a code that corresponds to the combination with sample interval), the width of the quantization value in the important part of the time difference value is expanded, improving tracking of contours and edges, and reducing edge business. There is a characteristic that. Now, when quantizing with variable sample density using the quantization characteristics shown in Figures 9 and 10, when the amplitude difference value is less than a constant value apn, samples at a position with a constant interval τpn+ are taken in. When the time difference value is increased by n times, the time difference value Στpn+ and the prediction residual signal corresponding to the predicted value by the previous restored pixel are quantized into an amplitude difference value. When the time difference value is 2, the quantized amplitude difference value is determined by the combination of the time difference value and the amplitude difference value (also called quantization level) so that the quantized amplitude difference value can be selected as a value larger or smaller than the constant value αpn. encoded into a characteristic code.

In particular, in the case of Table 1, constant value αp=3, constant interval τp=
Therefore, when the time difference value is 1 and the amplitude difference value is smaller than αp=3, the time difference value is set to 2. For example, when the time difference value is 2 and the amplitude difference value is 16 or more, the quantization code B(
hexadecimal), the time difference value is 3, and the amplitude difference value is 2~
When it is 7, a quantization code of 8 is output. Here, the amplitude difference value of 16.8 is multiplied by m in cases where the time difference value is 1, 2 or 3, but it is characterized in that it is a different quantization code.

Here, for a signal whose time difference value is 1 and is less than the constant value ffp, if the amount of information is regarded as the time difference value x the amplitude difference value, the probability is that if the time difference value is 2, the amplitude #& difference value will be even smaller than the constant value αp. Although the possibility is high, it is highly likely that the time difference value will be large for steep edges or irregularities such as a person's face, so the amplitude difference value 4 or smaller than the constant value αp will correspond to the time difference value 2. It can be said that an amplitude difference value 16 greater than a certain value Qp is provided.

As a result, the frequency of occurrence of codes with large time difference values can be increased, the compression rate can be increased to nearly 2 bits/pel compared to DPCM of 4 bits/pet (pixel), and deterioration in image quality can be reduced. As a result, the image quality is reduced to 8 bits (25 bits).
The SNR at 6 gradations) is about 35 to 37 dB.

Now, in variable sampling density coding, quantization 2 with a large time difference value
The compression ratio is increased by α, but in the conventional method described above, the time difference value is kept within a small range of about 1 to 2 in order to reduce the deterioration of image quality, like other block encoding and cosine transformation. It is difficult to make it a large value of 3 or more.

Therefore, the present inventors proposed a compressed transmission method for monitoring images when a change has already been detected (Patent Application No. 60-600012).
This method transmits the image information of the part that has changed between frames based on the interframe prediction residual between the previous image frame or predicted image frame and the current image frame. The circuit configuration of the image compression unit according to the method is shown. In this configuration, the change detection circuit 19 first compares the current image frame imported from the frame memory 16 to the current image frame x 7T17 and the reference image frame of the reference image frame x 7118 to detect a change.
If a change is detected, the changed pixel detection circuit 20 detects the error between the current image frame of the current image frame x 7a 17 and the predicted image (previous image) frame of the predicted image (previous image) frame x 7a 21. The magnitude of the inter-frame residual (absolute value of the difference) divides the pixels into large change pixels and small zero pixels.

Furthermore, if necessary, by propagation and degeneracy of changed pixels, one or two zero pixels sandwiched between changed pixels are made changed pixels, and one or two changed pixels sandwiched between zero pixels are made zero pixels. , a zero pixel isolated among changed pixels and a changed pixel isolated among zero pixels are removed.

The second-order difference molecular measurement encoding circuit 22 includes a quantization characteristic selection circuit 23
Therefore, the inter-frame residual or the value of the changing pixel of the current image frame can be compressed using the inter-line and intra-line quadratic difference molecular landscape conversion method (second-order DCPM, or differential variable sampling density encoding) shown in Figure 6. This allows efficient transmission of changed regions or changed pixels.

As shown in Table 2, when moving from a zero pixel to a changed pixel, the zero code compression circuit 24 transmits the value of the changed pixel as an initial value for prediction using a changed pixel code (FD, X), and converts the value from the changed pixel to a zero. When moving to a pixel, a zero pixel code (for example, FE) is transmitted as the end value of prediction, and the number of zero codes with a prediction residual of zero and the number of lines where all pixels are zero codes are encoded and compressed. Transmit. Also, by setting the value of the changed pixel in the current image frame to a value other than zero, it is possible to compress the changed pixel without using the changed R''1krF code or the zero pixel code, but in this case, even if the value of the changed pixel is zero, There is a need to.

Here, in order to reduce deterioration in image quality and to perform quantization within the range of changing pixels, it is not possible to increase the time difference value of the quantization characteristic. In addition, for multilevel images, the amount of code can be reduced by variable length coding called Huffman coding, but in variable sampling density coding using quantization characteristics with small time difference values, the frequency distribution of the occurrence of the code itself is such that the amplitude difference value is zero. Since it concentrates on the code of , it has not been possible to apply this to previous methods. In other words, in the conventional method, the frequency distribution of codes changes depending on the target image, so Huffman coding requires a large amount of calculation to optimally set the codebook of variable-length codes, and also requires zero-code compression to compress changing pixels. The problem is that it cannot be used effectively.

The code set collection circuit 26 in FIG.
A pufferfish is added to the above compression code for each frame to the number of transmission pixels, compression parameters, and code for indicating selection of all-one/variable-northern portion, and is transmitted as communication data. The compression parameters mentioned here include the quantization characteristics and the prediction coefficients of the interframe device prediction circuit, and if the prediction coefficients are optimal, the prediction residuals will concentrate to zero, the quantization error will become small, and the image quality will improve.

[Object of the Invention] The present invention has been made in view of the above-mentioned points, and its object is to provide an image compression method that efficiently transmits high-quality images by reducing the amount of code.

[Disclosure of the Invention] The present invention provides a linear prediction residual difference between the pixels of the current line to be next compressed and encoded and the predicted value of the next line based on the restored FjM elements of the previous line of the compression encoded fossil. Calculate the difference, and calculate the second difference, which is the predicted residual between the first difference of the pixel to be compressed and encoded next in the line direction and the predicted value of the next first difference based on the restored pixel of the compression encoder fossil. In an image compression method that uses second-order difference molecular measurement to obtain the quantization code corresponding to the combination of the time difference value and the amplitude difference value using the quantization characteristic, When variable-length encoding is performed so that the average code length of the obtained quantized code is shortened, the part where consecutive zero codes of zero values are equal to or greater than a certain value is converted into a zero code corresponding to the number of consecutive zero codes. The present invention will be described by way of embodiments.

Figure 1 shows the basic circuit configuration of the image compression ?'ISA of this embodiment, Figure 2 shows the basic circuit configuration of the decompression section B of this embodiment, and Figures 7 and 8 show the basic circuit configuration of the image compression section B of this embodiment. Components with the same numbers as in the figure perform the same operations.

First, compression encoding will be explained based on the @1 diagram configuration. The image data used here is for monitoring. The video signal of the V camera is separated into horizontal and vertical synchronization signals and image signals, and this image signal is YC-separated into a luminance signal (Y) and a color difference signal (C). Then, each signal is converted into a 6- to 8-bit digital value by an A/D converter "a" (not shown), and then read into the current image frame memory. In this example, the luminance signal (Y) of one image is 8 bits (256 x 256 pixels).
6th floor +1! ri/pixel.

The difference circuit 1 subtracts the inter-line predicted value of the prediction circuit 2 from the image data of the current image, and stores the first difference in the line buffer 3.
Lines are written, in other words, the line buffer 3 contains a prediction between the pixels of the current line to be compressed and encoded next and the predicted value of the next line based on the restored pixels of the previous line by the compression encoder. The first-order difference, which is the residual, is stored. This first-order difference is variable-sample-density encoded in the line direction by the variable-sample-density encoding circuit 4, and then the next first-order difference is predicted based on the first-order difference of the pixel to be compression encoded and the restored pixel by the compression encoder. A quantization code corresponding to the combination of the time difference value (sample interval) and the amplitude difference value (quantization level) is calculated from this second difference using the quantization characteristics. demand.

This quantization code has a fixed length (4 bits) as shown in Table 1.
). In reality, variable length encoding may be performed at one time.

The process of performing variable sampling density decoding from the quantized code to obtain the predicted value of the next line is similar to the conventional method described above.

In this second-order difference molecular measurement quantization (second-order DPCM, or differential variable sample density coding), a variable-length encoding circuit generally uses a variable-length encoding circuit so that the code length of this quantization code becomes shorter in the order of occurrence frequency. It is characterized by variable length encoding using No. 27. .

Considering Or variation encoding for zero-code compression, Table 3 shows an example of Huffman encoding using 15-level DPCM.This Huffman encoding has a low code generation rate. Since the code is assigned from the end of the object, when zero code compression is applied to a part with consecutive zero codes "0000" of zero amplitude difference value (quantization level), the extension code is 12 bits.
This makes me think that zero-code compression is not suitable.

When variable-length encoding is performed after zero-code compression, the extension code also changes each time it is encoded, and the zero-code code is not known at the zero-code compression stage, so it is difficult to decide how many consecutive zero-codes to compress. Can not.

(Table 3) Therefore, in the present invention, a specific code of the variable length code is compressed and encoded so that the part in which the zero codes of the amplitude difference value are consecutive at a certain value or more is compressed and encoded into a zero code corresponding to the number of consecutive zero codes. is assigned to the zero code code and b (variable length zero code compression coding is performed. After determining this, code assignment is performed together with other quantization codes.

In the embodiment of the second-order differential DPCM, in order to simplify the encoding calculation, variable length codes as shown in Table 4 are fixed in advance, and the code of the zero code is set as '111', for example.
1". In this case, it is not necessarily optimal like Huffman encoding, but it is not much different from this, and moreover, it is suitable for microcomputer processing. This zero code code is 4b i
t, but the code length of the variable length code other than this is 2 or 3.
Since there are two of each, when assigning a zero code to a positive quantization code, a 2bit variable length code is assigned to one of the zero code and the positive quantization code. I am using this to compensate for the zero sign code.

This makes it possible to add the effect of zero code compression regarding secondary difference and change pixel compression to variable length encoding.

The zero code code 1111'' corresponds to the code F (hexadecimal number) in Table 2, and follows “1111” (4 bits follow F <
Corresponds to 4bit. Since the zero sign code is 2 bits, the meaning of the zero sign in @2 table is Fl is 5 sign zero, F
2 uses 6 code zeros, . . . , F8 uses 12 code zeros for variable length encoding, and up to 4 zero codes can reject zero codes.

Table 5 is an example of the frequency distribution of codes for quadratic difference DPCM encoding and variable length zero code compression encoding.
I RL (monochrome). Here, a specific code "1111" of the variable length code is used to compress and encode the part where five or more consecutive zero codes of zero in the amplitude difference value are compressed into zero codes F1 to F8 corresponding to the number of consecutive zero codes. As a result of assigning to the zero code code, 28578 zero codes are reduced by 7961 to 20617, resulting in a zero code compression effect of 1154 bytes, 796JX2 1]54X8=669
This means that there was a zero code compression effect of 0 bits = 836 bytes.

As a result, the amount of code is 6 for headers, fin end codes, etc.
Add 40 bytes and do 2. 5b i L/pet SNI'? =38.7dB. Conventional example 11 formula 2
3.7 bit/pel for differential DPCM and zero code compression
, 5NR=: (8.7dB, so 1.2bit
This means that the amount of code is reduced.

In the case of Town V sample density coding, for simplicity, the correspondence between the variable sample density codes and the variable length codes in Table 4 is set using the target FIi (
*It is necessary to change it depending on the quantization characteristics. Table 6 shows an example of quantization characteristic ratio variable sample density encoding for a time difference value of 2, and in this example, 2゜0biL/pel+5NR=3
The average code length of the variable length was 5.8 dB, and the average code length of the variable length was 3°4 bits for a 4-bit quantized code. 2.36 it/pel of the conventional method.

This is a reduction of 0.3 bit for 5NR=35.8 dB.

Now, in the decompression unit B of FIG. The zero-code compression code is decoded into a 4bit quantization code, and then the conventional method shown in tItJ8 (Table 4) (Table 5) (Table 6) The figure shows the image compression ffl5A and decompression unit B of this embodiment. In addition to the configuration of the first embodiment, this embodiment also stores the current image frame to be compressed and encoded next and the frame number 7a29. The difference circuit 32 calculates the inter-frame difference between the restored previous image or the predicted image frame from the prediction circuit 30 based on the current image, and calculates this inter-frame difference using the second-order difference numerator in FIGS. 1 and 2. Measurement quantization (2
It performs variable-length zero code compression encoding using differential DPCM or differential ei (variable sample density encoding). Frame x 7a 29,
The prediction circuit 30 creates a predicted image 7 layers l. In the 4th I71 decompression unit B, corresponding to the image compression unit A, the interpolation and synthesis circuit 33. One frame worth of image data is restored by a frame cross 7a 34 and an f-measurement circuit 35.

Friend (Example) - Figures 5 and 6 show the image compression section A and decompression section B of this embodiment.
In this embodiment, in addition to the configuration of the first embodiment, prediction is performed based on the current image frame to be compressed and encoded next and the restored previous image or current image stored in the frame box 7a29. A change pixel detection circuit 36 detects a pixel with a large change above a threshold value and which has an inter-frame difference with the image frame, and converts the change pixel between the current image or frames into second-order difference molecular measurement quantization (second-order difference DPCM, or difference difference sample). Density encoding) and Of variable length code compression encoding. Note that the code editing circuit 38 in Figure #45 corresponds to the code editing circuit 26 in Figure 11, and the quantization characteristic selection circuit 37! 1r
This corresponds to the quantization characteristic selection circuit 23 in Figure J11. Also#
The code distribution circuit 39 in the IJG diagram distributes code data into zero codes and quantization characteristics, and the quantization characteristics are manually input to the quantization characteristics selection circuit 40.

In the case of the present invention, as in the conventional method, the second-order difference predictive coding method and zero code compression are not limited to intra-frame coding, but also inter-frame coding as in the changing pixel compression described in It can also be used for encoding. Therefore, by selecting the quantization characteristics and the number of transmission pixels, the information is compressed using a compression rate high enough to allow the outline of the changing part of the image to be recognized, or the image information of the sampling grid (coarse image) is compressed and transmitted in a narrow band. Transmit with a narrow belt,
The receiving side (decompressing unit B side) interpolates and displays the data, and the transmitting side (image compressing unit A side) also interpolates and displays the transmitted i'
A predicted image of the next image frame to be transmitted is created based on the image frame, and the current image is updated to update the current image based on a sharpening (fine) command or transmission mode setting command from the transmitting side or receiving side. Writing to the frame buffer 7a is prohibited, the image data in the current image frame 7a is set as the current image frame, and the interframe residual ( By repeating the process of compressing the information (difference) using quantization characteristics with a low compression rate and transmitting it in a narrow band, the predicted value converges and becomes clearer sequentially. With predicted image frame? It is also possible to compress and transmit the information of the inter-frame residual or the current image frame, which is the difference, and to sequentially sharpen the information.

Furthermore, the present inventors have already proposed (Patent Application No. 60-01276)
No. 8) Jittery l/-mul? ll 4fr position prediction The prediction coefficients of the jtl path can be selected, and the feedback loop outside the interframe j11f position prediction circuit includes a compression process and an expansion process using the town change sample evaluation method, and - the above-mentioned changed pixel The detection circuit 36 detects the amount of change in the interframe residual (the sum of the interframe residual for each pixel, the absolute value for each pixel, etc.) by the interframe residual detection circuit, the change amount detection circuit, and the averaging processing circuit. When the amount of change in the inter-frame residual is reduced by applying an averaging spatial filter (selection or repetition) to the inter-frame residual or the current image frame of the transmission 11f, the frame with changed pixels is The inter-frame residual, which is the error between the current image frame before transmission and the predicted value based on the transmitted Af image frame, is information compressed. Narrowband transmission? '? When applying an averaging spatial filter to the current image frame before transmission, first set the already transmitted +if image frame to zero and apply a spatial filter with a high degree of averaging to the current image frame before transmission ( selection or repetition'O
r) is applied to transmit the inter-frame residual of the pixel with a change in the low-frequency component, and a spatial filter with a low degree of averaging is applied to the next current image frame to transmit the inter-frame residual of the pixel with a change in the high-frequency component. By transmitting the difference, the predicted value can be converged and gradually sharpened.

In addition, in the process of transmitting the monitoring image (current image) information at low speed or sequentially sharpening, it is possible to transmit the monitoring image (current image) information at a predetermined interval by 11 ( <The prediction residual between frames of the thinned out pixels is divided into pixels whose magnitude is greater than the first setting value, and pixels whose prediction residual is smaller than the first setting value, and the prediction residual of zero pixels is calculated. Alternatively, the changed part of the ffN image, in which the pixel value is set to zero and isolated zero pixels and changed pixels are removed by a change residual shaping circuit, is compressed and encoded and transmitted, and the change residual is determined on the receiving side (expansion unit B side). The circuit and interpolation synthesis circuit interpolate and restore the interframe predictability i or pixel values of roughly thinned out pixels, and also transmit the same (on the FT side, the change residual judgment circuit and the interpolation synthesis circuit interpolate the predicted image) only when the neighboring grid of roughly thinned pixels (coarse grid) is a change 1lIIi element, the pixel between them or the grid adjacent to the change pixel (coarse grid) is
By interpolating and restoring the predicted difference between frames or pixel values between t pixels of a coarse grid), it is also possible to sequentially sharpen only the changed part without reducing the sharpness around the changed part. . In particular, when transmitting prediction residuals between frames, only when the adjacent grid (coarse grid) is a change pixel, the pixel between them or the change pixel and the zero pixel of the adjacent grid ($11 grid) When interpolating and restoring the value of the prediction residual between frames between frames, first the value of a zero pixel within the changed pixel interval of the restored image (predicted image) of the corresponding previous image is changed to the restored image of the previous image at the changed pixel position. By interpolating with the pixel values of It is also possible to reduce blurring of parts.

[Effects of the Invention] As described above, the present invention uses variable-length codes to compress consecutive zero codes of a quantized code of second-order difference 1) PCM or second-order difference variable sampling density coding. By assigning a specific code of the variable code code to the zero code code m5 in the short part, it is possible to achieve good image quality and a high compression rate using a quantization characteristic with a small time difference value.

[Brief explanation of the drawing]

FIG. 1 is a circuit configuration diagram of the image compression unit according to the first embodiment of the present invention, FIG. 2 is a circuit configuration diagram of the decompression unit shown above, and FIG. , f:lS4 Figure is a circuit configuration diagram of the same extension section as above, and Figure 5 is Embodiment 3 of the present invention.
6 is a circuit diagram of the decompression unit of the above, and FIG. 7 is a circuit diagram of the image compression unit of the conventional example.
FIG. 8 is a circuit configuration diagram of the decompression section same as above, FIGS. 9 and 10 are operation explanatory diagrams same as the above, and FIG. 11 is a circuit configuration diagram of another conventional image compression section. DESCRIPTION OF SYMBOLS 1... Differential circuit, 2... Prediction circuit, 3... Line buffer, 4... Variable sample density encoding circuit, 5...
Zero code compression circuit, 6...Machihen sample evaluation decoding circuit, 7
...Line pan 7a, 8...Interpolation synthesis circuit, 9...
・Line/in 7a, 10... Zero code decoding circuit, 1
1... Variable sampling density decoding circuit, 12... Line buffer, 13... Interpolation and synthesis circuit, 14... Line buffer, 15... Prediction circuit. Agent Patent Attorney Ishi 1) Ai Figure 7 Figure 9 Figure 10 Figure j1111

Claims (3)

[Claims] (1) Next, find the first-order difference, which is the prediction residual between lines, between the pixels of the current line to be compression-encoded and the predicted value of the next line based on the restored pixels of the previous line that has been compression-encoded. Regarding the first-order difference, calculate the second-order difference, which is the prediction residual between the first-order difference of the next pixel to be compression-encoded in the line direction and the predicted value of the next first-order difference based on the compression-encoded restored pixel, and calculate this second difference. In an image compression method that performs quantization using second-order difference molecular measurements, the quantization code corresponding to the combination of a time difference value and an amplitude difference value is determined using quantization characteristics. When variable-length encoding is performed so that the average code length of the encoded code is shortened, the part in which zero codes of the amplitude difference value are consecutive for a certain value or more is compressed and encoded into a zero code corresponding to the number of consecutive zero codes. An image compression method using variable length zero code compression encoding, characterized in that a specific code of a variable length code is assigned to a zero code code. (2) In the above-mentioned second-order difference molecular measurement quantization, the inter-frame difference between the current image frame to be compressed and encoded next and the restored previous image stored in the frame buffer or the predicted image frame based on the current image is calculated. An image compression method according to claim 1, characterized in that the image compression method includes a step. (3) In the above-mentioned second-order difference molecular measurement quantization, the inter-frame difference between the current image frame to be compressed and encoded next and the restored previous image stored in the frame buffer or the predicted image frame based on the current image is 2. The image compression method according to claim 1, further comprising a step of determining pixels with a change greater than a certain threshold value.