JPH05292330A - Picture data coding method - Google Patents

Abstract

PURPOSE:To improve the compression when a change is less between pictures by selecting only a block discriminated to be a change between a preceding picture and a current picture and coding the block. CONSTITUTION:A block change discrimination section 12 reads a picture element of the same block position as a relevant block of a reference frame latched in a reference frame latch section 11, compares the picture element with a current picture element block, calculates a change, obtains a block change from a picture element unit change, the change is compared with a discrimination threshold level by a threshold level latch section 13, in which whether or not the change is valid is discriminated and the result is latched by a block information latch section 14. When the block is a valid block, an ADCT coding section 15 reads block picture data in the block buffer 10 and encodes the data. In the case of an ineffective block, the block is not coded. A multiplexer 16 outputs sequentially block information of the latch section 14 and variable length data from the coding section 15. That is, when there is a change in an object block, a change discrimination threshold level is set small and when no change is in existence, the threshold level is set large to discriminate a succeeding block change.

Description

Detailed Description of the Invention

(Table of Contents) Industrial Application Field of the Prior Art (FIGS. 12 to 13) Problems to be Solved by the Invention Means for Solving the Problems (FIGS. 1 to 4) Working Example (a) First Example (FIGS. 5 to 8) (b) Description of second embodiment (FIG. 9) (c) Description of third embodiment (FIG. 10) (d) Description of fourth embodiment (FIG. 11) (e) Description of another embodiment Effect of the invention

[0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a screen data encoding method for dividing image data into m * n blocks and encoding each block.

Image data has an enormous amount of data as compared with code data. In particular, in order to store halftone image data and color image data, or to transmit at high speed and high quality, the level of each pixel is high. It is necessary to encode the key value with high efficiency.

An example of such a highly efficient compression method for image data is an adaptive discrete cosine transform coding method.
Adaptive Discrete Coding
sine Transform) divides an image into blocks of 8 × 8 pixels, transforms the image signal of each block into a coefficient of spatial frequency distribution by two-dimensional discrete cosine transform (hereinafter referred to as DCT), and a threshold adapted to the visual sense. The quantized coefficient obtained by the above is encoded by the Huffman table obtained statistically.

In such a multi-valued image coding method, it is required to efficiently code continuous images.

[0006]

12 and 13 are explanatory views (No. 1) and (No. 2) of the prior art. In a conventional still image coding apparatus, each image is coded independently even if there is a large correlation between the images and a small change, such as an animation image considered as a set of still images.

For example, in the adaptive discrete cosine change coding system shown in FIG. 12, as shown in FIG. 12 (A), the block buffer 10 holds one block (8 × 8 pixels) of data and a two-dimensional DCT transform unit. Set the image signal of each block to 20
It is converted into a coefficient of spatial frequency distribution by the three-dimensional discrete cosine transform, and is quantized by the threshold adapted to the visual sense in the quantizer 21,
The quantized coefficient is variable length coded by the Huffman table statistically obtained by the variable length coding unit 22 and output.

Similarly, on the restoration side of FIG. 12B, the variable length decoding unit 30 decodes the coded data into a fixed length, and the dequantization unit 31 dequantizes the data to restore the DCT coefficient. , The inverse DCT conversion unit 32 converts it into a block image, and the in-block pixel writing control unit 34 controls the image memory 33.
The restored image is written in the block address position of the block address generating unit 35.

Further, in the moving picture coding technique for a continuous scene such as a video conference apparatus, each picture is coded independently. That is, as shown in FIG. 13A, the difference between the block image of the image memory 26 holding the previous image and the block image of the current image is calculated by the difference image generation unit 27 and held by the block buffer 10. Two-dimensional DCT transform unit 20
Then, the image signal of each block is converted into a coefficient of the spatial frequency distribution by the two-dimensional discrete cosine transform, quantized by the quantizer 21 with a threshold adapted to the visual sense, and the quantized coefficient is statistically calculated by the variable length encoder 22. Variable-length coding is performed using the obtained Huffman table and output.

Further, the quantized coefficient is inversely quantized by the inverse quantizer 23 and inversely transformed into a DCT coefficient, and then inverse DCT transformed by the inverse DCT transformer 24 to be converted into a difference block image, which is then stored in the image memory. The previous image of 26 and the addition unit 25 add,
The image memory 26 is updated.

Similarly, on the decompression side of FIG. 13B, the variable length decoding unit 30 decodes the encoded data into a fixed length, the dequantization unit 31 dequantizes the decompressed data, and reconstructs the DCT coefficient. , The inverse DCT conversion unit 32 converts the difference block image,
The in-block pixel writing control unit 34 writes the image obtained by adding the difference image and the previous image in the adding unit 36 to the block address position of the block address generating unit 35 of the image memory 33.

[0012]

However, in the conventional technique, the conventional still image coding technique has the following problems, but although the device can be configured with a small-scale circuit, an image with little change between images can be obtained. On the other hand, since each image is encoded independently, like an image of a videophone with a fixed animation or camera,
Even if there is very little change between consecutive images, the compression rate cannot be increased.

When a moving picture coding technique is used,
Due to the differential encoding, the compression rate is increased, but the device becomes large-scale. In order to solve this, the present applicant first compares the block image and the previous image, encodes a block having a change of a certain threshold value or more as an effective block, and transmits only the effective block to obtain a compression ratio. We have proposed a method for improving this (Japanese Patent Application No. Heisei 4 years No. 77957, March 31, 1992).
Date application).

However, with this proposed technique alone, when the rate of change of the block is changing gently, the effective block appears prominently, resulting in a problem that the screen becomes discontinuous.

Therefore, the present invention provides an image data encoding method capable of efficiently encoding a continuous image having little change between images with a small-scale circuit and improving the image quality of a restored image. To aim.

[0016]

FIG. 1 shows the principle of the present invention. According to a first aspect of the present invention, in an image data encoding method for encoding image data by using m × n pixels as blocks, a variation amount of pixels at the same position of a block at the same position of the current image and the previous image is detected. Then, a step of comparing the amount of change of the block with a determination threshold value to determine whether there is a change in the contents of the block, a step of changing the determination threshold value according to the determination result, and an effective change with the determination result. Encoding the block determined to be a block.

According to claim 2 of the present invention, in claim 1,
The determination threshold is changed according to the determination result to determine a change in a block adjacent to the block. A third aspect of the present invention is characterized in that, in the second aspect, when it is determined that the effective block is the valid block, the determination threshold value is changed from a large value to a small value and a change of the next block is determined. ..

A fourth aspect of the present invention is based on the second aspect.
When it is determined to be the valid block based on the determination result, the determination threshold is changed from a large value to a small value, and the change of the previous block is determined again.

A fifth aspect of the present invention is characterized in that, in the third or fourth aspect, the determination threshold value is returned to a large value when it is determined as an invalid block having no change based on the determination result.

A sixth aspect of the present invention is based on the second aspect.
It is characterized in that the block is judged to be an effective block from the judgment result of the block and the result of judging the adjacent block by changing the judgment threshold value.

According to claim 7 of the present invention, in claim 6,
When it is determined as a valid block by the determination threshold,
The determination threshold is changed from small to large, the next block is determined, and when the next block is a valid block, the block is determined as a valid block.

According to claim 8 of the present invention, in claim 6,
When it is determined as an invalid block by the determination threshold,
When the determination threshold is changed from large to small, and the block is re-determined and determined to be a valid block, the determination threshold is changed from small to large to determine the previous or next block, and the previous or next block is determined. When the block is a valid block, the block is determined to be a valid block.

[0023]

According to the first aspect of the present invention, as shown in FIGS. 1 to 4, of the blocks at the same position of the previous image and the current image, only the image block having a change from the previous image is selected and encoded. Therefore, in an image with many background portions that do not change, the code data required for the entire image can be largely omitted.

Further, the block determination is performed by changing the determination threshold value which is the determination reference based on the change determination result, so that even from the start of the change to the end of the change, it can be determined as the effective block area even if the block change rate changes gently. , It is possible to prevent discontinuous movement in which an effective block appears to project.

According to the second aspect of the present invention, as shown in FIGS. 1 to 4, since the adjacent block is judged by changing the judgment threshold value, the adjacent block can be judged and controlled according to the judgment result, and the effective block is projected. Can be prevented from appearing.

According to the third aspect of the present invention, as shown in FIG. 1, when the threshold value is large and it is determined that the block is an effective block, the threshold value of the next block is set to be small. It can be determined that the effective block area is from the start to the end.

According to the fourth aspect of the present invention, as shown in FIG. 2, when the threshold value is set large and it is determined that the block is a valid block, the threshold value is set small and the previous block is re-determined. The block can be a block, the effective block can be continuously determined, and the effective block area can be determined from the start to the end of the change.

In claim 5 of the present invention, as shown in FIGS. 1 and 2, when it is determined that the block is an invalid block that does not change according to the determination result, the determination threshold is returned to a large value, so the determination threshold remains small. It is possible to prevent the invalid block from being determined as the valid block.

In claim 6 of the present invention, as shown in FIGS. 3 and 4, it is determined whether the block is a valid block from the block determination result and the determination result of the adjacent block by changing the determination threshold. Therefore, continuous valid blocks can be effectively extracted, and it can be determined as a valid block area from the start to the end of the change.

According to claim 7 of the present invention, as shown in FIG. 3, when it is determined that the block is a valid block by the determination threshold,
Change the judgment threshold from small to large, judge the next block,
When the next block is a valid block, the block is determined to be a valid block, so even if the start of a valid block is detected by reducing the judgment threshold value, false detection can be prevented, and continuous valid Blocks can be detected.

According to the eighth aspect of the present invention, as shown in FIG. 4, when the block is judged to be an invalid block by the judgment threshold value,
When the judgment threshold value is changed from large to small and the block is judged again to be a valid block, the block is set as a candidate block, and the judgment threshold value is changed from small to large to change the previous or next block. When the determination is made and the previous or next block is a valid block, the block is determined to be a valid block, so that the valid block area from the start to the end of the valid block can be effectively detected.

[0032]

Embodiments (a) Description of the First Embodiment FIG. 5 is a block diagram of the first embodiment of the present invention, and FIG. 6 is an operation explanatory view of the present invention (No. 1; when the block change is gradual). 7 is an operation explanatory diagram of the present invention (No. 3; when the block change rate is not constant).

In FIG. 5 (A), 10 is a block buffer for holding image data of one block unit input from the input terminal, 11 is a reference frame holding section, and reference frame image data of one screen. Is stored.

A block change determination unit 12 compares the block image data of the block buffer 10 with the block image data of the same block position of the reference frame holding unit 11 on a pixel-by-pixel basis to calculate the amount of change in pixel value. , The amount of change in the pixel value within the block (the maximum value of the pixel values within the block, or the cumulative value of the absolute values of the pixel values within the block, or the cumulative value of the squared values of the pixel values within the block) and the threshold value holding unit 13 The change in contents of each block is determined by comparing with the threshold value of.

Reference numeral 13 is a discrimination threshold value holding section, which holds a large threshold value and a small threshold value and switches them by a threshold value change signal,
An output to the block change determination unit 12, 14 is a block information holding unit, which holds change information (block information) of each block from the block change determination unit 12.

Reference numeral 15 denotes an ADCT encoder, which determines the block buffer 1 according to the determination result of the block change determination unit 12.
The block image data of 0 is encoded as shown in FIG.
As shown in (B), the transmitted block image is subjected to a two-dimensional discrete cosine transform to transform into a coefficient of the spatial frequency distribution, and the quantization threshold holding unit 12a uses a threshold adapted to the visual sense. The quantization unit 21 that quantizes the coefficient of the spatial frequency distribution to obtain the quantized coefficient and the Huffman code table holding unit 22a are statistically obtained. However, the Huffman table is used to perform variable length coding of the quantized coefficient. Encoding unit 2
It consists of 2.

Reference numeral 16 denotes a multiplexer, which selectively outputs the block information of the block information holding unit 14 and the encoded data of the ADCT encoding unit 15. The operation of the configuration shown in FIG. 2 will be described with reference to FIG.

The current image data of one block is input from the input terminal and temporarily held in the block buffer 10. The reference frame holding unit 11 holds a reference signal for detecting the presence / absence of change in the current block. There is.

The block change determination unit 12 reads out the pixel of the block at the same position as the block of the reference frame held in the reference frame holding unit 11, and at the pixel at the same position, the pixel of the read block and the current image. The pixel of the block is compared, the amount of change is calculated, the amount of change of the block is calculated from the amount of change of each pixel (either the maximum value in the block, the cumulative value, or the cumulative value of the squared values), and the threshold value is held. By comparing with the discrimination threshold of the unit 13, it is determined whether the block is an effective block (the amount of change is greater than or equal to the threshold),
The determination result (whether it is a valid block or an invalid block) is output to the block information holding unit 14 and held therein.

When the determination result is a valid block, the block change determination unit 12 sends a reference frame rewriting signal to the reference frame holding unit 11, and the reference frame holding unit 1
1 updates the held image data of the block to the current image data of the block buffer 10.

Further, the block change judgment unit 12 outputs a threshold value change signal to the judgment threshold value holding unit 13 according to the judgment result,
When the determination result is a valid block, the threshold value is switched from large to small, and when the determination result is an invalid block, the threshold value is switched from small to large.

When the result of this determination is a valid block, the ADCT encoder 15 reads and encodes the block image data of the block buffer 10, and when it is an invalid block, it does not perform read encoding.

The multiplexer 16 sequentially outputs the block information of the block information holding unit 14 and the variable length code data from the ADCT coding unit 15. That is, when there is a change in the target block, the change determination threshold is set to a small value,
The change of the next block is determined, and when there is no change in the target block, the change determination threshold is set to a large value and the change of the next block is determined.

As described above, since the effective block area can be continuously extracted in the scanning direction of the block by switching the change determination threshold depending on the presence or absence of the change, the unnaturalness of the block caused by the intermittent continuation of the effective block. It is possible to suppress the unnatural appearance.

This will be described with reference to FIGS. 6 to 8.
In FIG. 6A, in the case where the block changes gently as shown in FIG. 6A, in the regions A, B, and C, the conventional method of comparing with a single threshold value
As shown in (B), the portion A at the start of change and the portion C at the end of change are recognized as invalid blocks, and the peak portion B
Only the effective block appears, and the effective block appears prominently.

On the other hand, in the first embodiment, in the scanning direction, as shown in FIG. 6 (C), when it is determined that the peak portion B is a valid block, the threshold value is lowered to a small value. Therefore, the effective block area can be made continuous.

Further, in FIG. 7, as shown in FIG.
When the block B having a small change is isolated in the effective block area, in the conventional technique, as shown in FIG. 7B, the effective block A, the invalid block B, and the effective block C are used.
And the effective block is intermittent, and it looks unnatural, but
In the first embodiment, as shown in FIG. 7C, the threshold value is set small by recognizing the effective block A.
Is also recognized as a valid block, and the valid block area can be made continuous from the start of change to the end of change.

Similarly, in FIG. 8, when the block change rate is not constant as shown in FIG. 8A, in the prior art, effective blocks are intermittent as shown in FIG. 8B. However, in the first embodiment, as shown in FIG.
The effective blocks are continuous and the image quality can be improved.

(B) Description of the Second Embodiment FIG. 9 is a block diagram of the second embodiment of the present invention. In FIG. 9, the same components as those shown in FIG. 5 are denoted by the same symbols, and 17 is a previous block holding unit, which holds the previous block image data of the block buffer 10.

The operation of the configuration shown in FIG. 9 will be described with reference to FIG. 2. One block of current image data is input from the input terminal and temporarily held in the block buffer 10, and the reference frame holding unit 11 stores the current block of the current block. A reference signal for detecting the presence or absence of a change is held, and the previous block holding unit 1
In 7, the image data of the previous block held in the block buffer 10 is held.

The block change determination unit 12 reads out the pixel of the block at the same position as the block of the reference frame held in the reference frame holding unit 11, and at the pixel at the same position, the pixel of the read block and the current image. The pixel of the block is compared, the amount of change is calculated, the amount of change of the block is calculated from the amount of change of each pixel (either the maximum value in the block, the cumulative value, or the cumulative value of the squared values), and the threshold value is held. It is determined whether or not the block is a valid block (the amount of change is equal to or greater than the threshold) by comparing with the determination threshold of the unit 13.

When the determination result (valid block or invalid block) is a valid block, the block change determining section 12 sends a reference frame rewriting signal to the reference frame retaining section 11, and the reference frame retaining section 11 retains it. The image data of the block concerned is updated to the current image data of the block buffer 10.

Further, the block change determination unit 12 outputs a threshold value change signal to the determination threshold value holding unit 13 according to the determination result,
When the determination result is a valid block, the threshold value is switched from large to small, and when the determination result is an invalid block, the threshold value is switched from small to large.

If the result of this determination is that the block is valid, the previous block in the previous block holding unit 17 is read out, and the block change determination unit 12 reads out the block in the same position as the previous block in the reference frame holding unit 11. It is determined again whether the previous block is a valid block.

When the block change determination unit 12 determines that the previous block is a valid block, the block change determination unit 12 determines the retained determination result (change information) of the previous block as a valid block (for example, "1").
To the block information holding unit 14 and the coding unit 15, and when the previous block is determined to be an invalid block, the determination result of the held previous block is stored in the block information holding unit 1.
4 and the encoding unit 15.

When the block change determination unit 12 outputs the determination result of the effective block, the ADCT encoding unit 15
The block image data of the previous block holding unit 17 is read and encoded, and when it is an invalid block, read encoding is not performed.

The multiplexer 16 sequentially outputs the block information of the block information holding unit 14 and the variable length code data from the ADCT coding unit 15. That is, when there is a change in the target block, the change determination threshold is set to a small value,
The change of the previous block is re-determined, and when it is a valid block, the previous and current blocks are set as valid blocks, and when there is no change in the target block, the change determination threshold is set to a large value and the next block is set. Determine the change in.

As described above, the change determination threshold value is switched depending on the presence or absence of a change, and when it is determined that the block is a valid block, the previous block is checked again with a low threshold value.
Even if the area changes gently, the effective block can be effectively set.

This will be described with reference to FIGS. 6 to 8.
In FIG. 6A, in the case where the block changes gently as shown in FIG. 6A, in the regions A, B, and C, the conventional method of comparing with a single threshold value
As shown in (B), the portion A at the start of change and the portion C at the end of change are recognized as invalid blocks, and the peak portion B
Only the effective block appears, and the effective block appears prominently.

On the other hand, in the second embodiment, in the scanning direction, as shown in FIG. 6E, when it is determined that the peak portion B is an effective block, the threshold value is lowered to a small value and the previous block A is determined again. Therefore, the area A can be determined as an effective block, and the effective block area can be continuous from the start to the end of the change.

Further, in FIG. 7, as shown in FIG.
When the block B having a small change is isolated in the effective block area, in the conventional technique, as shown in FIG. 7B, the effective block A, the invalid block B, and the effective block C are used.
And the effective block is intermittent, and it looks unnatural, but
In the second embodiment, as shown in FIG. 7C, the threshold value is set small by recognizing the valid block C.
Is also recognized as a valid block, and the valid block area can be made continuous from the start of change to the end of change.

Similarly, in FIG. 8, when the block change rate is not constant as shown in FIG. 8A, in the conventional technique, effective blocks are intermittent as shown in FIG. 8B. However, in the second embodiment, as shown in FIG.
The effective blocks continue from the start of change to the end of change, and the image quality can be improved.

(C) Description of Third Embodiment FIG. 10 is a block diagram of the third embodiment of the present invention. In FIG. 10, the same components as those shown in FIGS. 5 and 9 are denoted by the same symbols, and 18 is a previous block information holding unit, which is the determination result of the previous block determined by the block change determination unit 12. The previous block information is stored.

The operation of the configuration of FIG. 10 will be described with reference to FIG. 3. One block of the current image data is input from the input terminal and temporarily held in the block buffer 10, and the reference frame holding unit 11 stores the current block of the current block. A reference signal for detecting the presence / absence of a change is held, and the previous block holding unit 17 holds the image data of the previous block held in the block buffer 10.

The block change determination unit 12 reads out the pixel of the block at the same position as the block of the reference frame held in the reference frame holding unit 11, and at the pixel at the same position, the pixel of the read block and the current image. The pixel of the block is compared, the amount of change is calculated, the amount of change of the block is calculated from the amount of change of each pixel (either the maximum value within the block, the cumulative value, or the cumulative value of the squared values), and the threshold value is held. By comparing with the discrimination threshold (small) of the unit 13, it is determined whether or not the block is an effective block (the amount of change is greater than or equal to the threshold).

When the determination result (valid block or invalid block) is a valid block, the block change determining unit 12 holds the block information (valid block) in the previous block information holding unit 18 and then sets the threshold value. The change signal is output to the determination threshold value holding unit 13 to switch the threshold value from small to large.

If the result of this judgment is that the block is valid, the next block is read from the block buffer 10 and the block change judgment unit 12 makes the reference frame holding unit 1
The block at the same position as 1 is read and it is determined whether the next block is a valid block.

When the block change determination unit 12 determines that the next block is a valid block, the block change determination unit 12 determines the retained determination result (change information) of the current block as a valid block (for example, "1").
When the next block is judged to be an invalid block, the judgment result of the held current block is changed to an invalid block,
It outputs to the block information holding unit 14 and the encoding unit 15.

When the block change determination unit 12 recognizes the current block as a valid block, it sends a reference frame rewriting signal to the reference frame holding unit 11, and the reference frame holding unit 11
Updates the reference frame holding unit 11 with the image data of the block held in the previous block buffer 17.

When the block change determination section 12 outputs the determination result of the effective block, the ADCT encoding section 15
The block image data in the previous block buffer 17 is read and encoded, and when it is an invalid block, read encoding is not performed.

The multiplexer 16 sequentially outputs the block information of the block information holding unit 14 and the variable length code data from the ADCT coding unit 15. That is, when the change determination threshold is small and there is a change in the target block, this is set as a temporary valid block, the change determination threshold is set to a large value, the change in the next block is determined, and the next block also changes. If so, the target block and the next block are recognized as valid blocks, and if there is no change in the next block, the current block is determined as an invalid block.

As described above, the change determination threshold value is switched depending on the presence or absence of change, and when it is determined that the block is a valid block, the next block is checked with a high threshold value, and the current block is determined to be the valid block according to the determination result of the next block. Even if the area is changing slowly,
Effective blocks can be set effectively.

This will be described with reference to FIGS. 6 to 8.
In FIG. 6A, in the case where the block changes gently as shown in FIG. 6A, in the regions A, B, and C, the conventional method of comparing with a single threshold value
As shown in (B), the portion A at the start of change and the portion C at the end of change are recognized as invalid blocks, and the peak portion B
Only the effective block appears, and the effective block appears prominently.

On the other hand, in the third embodiment, in the scanning direction, as shown in FIG. 6D, when the block A is determined to be an effective block with a small threshold value, the threshold value is increased to the next block B. Therefore, the area A can be determined as a valid block, the start of change can be effectively detected, and the valid block area can be made continuous.

Further, in FIG. 7, as shown in FIG.
When the block B having a small change is isolated in the effective block area, in the conventional technique, as shown in FIG. 7B, the effective block A, the invalid block B, and the effective block C are used.
And the effective block is intermittent, and it looks unnatural, but
In the third embodiment, as shown in FIG. 7 (D), the threshold value is set to be large by recognizing the effective block B to determine the next block C. Therefore, the block B is also recognized as the effective block.
The effective block area can be made continuous.

Similarly, in FIG. 8, when the block change rate is not constant as shown in FIG. 8A, in the conventional technique, effective blocks are intermittent as shown in FIG. 8B. However, in the third embodiment, as shown in FIG.
The start of change can be effectively detected, the effective blocks continue, and the image quality can be improved.

(D) Description of the Fourth Embodiment FIG. 11 is a block diagram of the fourth embodiment of the present invention. In FIG. 11, the same parts as those shown in FIGS. 9 and 10 are shown by the same symbols, and 19 is a rear block buffer, which holds the next block data from the input terminal. is there.

The operation of the configuration of FIG. 11 will be described with reference to FIG. 4. One block of image data is input from the input terminal and temporarily held in the rear block buffer 19, and the reference frame holding unit 11 stores the current block of the current block. A reference signal for detecting the presence or absence of a change is held, the current block buffer 10 is provided with the block data from the subsequent block buffer 19 and is held as the current block data, and the previous block holding unit 17 is The image data of the previous block held in the block buffer 10 is held.

The block change determination unit 12 reads out the pixel of the block at the same position as the block of the reference frame held in the reference frame holding unit 11, and at the pixel at the same position, the pixel of the read block and the current block. Compare with the pixels of the current image block in buffer 10,
The change amount is calculated, the change amount of the block (either the maximum value in the block, the cumulative value, or the cumulative value of the squared values) is calculated from the pixel-by-pixel change amount, and the determination threshold (large) of the threshold value holding unit 13 is calculated.
It is determined whether the block is a valid block (the amount of change is equal to or more than a threshold value) by comparing with.

When the block change determination unit 12 determines that the current block is a valid block, the block change determination unit 12 outputs valid block information to the block information holding unit 14 and the encoding unit 15, and causes the current block buffer 10 to be encoded.

On the contrary, when the determination result (valid block or invalid block) is an invalid block, the block change determination unit 12 outputs a threshold value change signal to the determination threshold value holding unit 13 to set the threshold value from large to small. Switch to current block buffer 1
The current block is read again from 0, and the block change determination unit 1
In step 2, the block at the same position as the block in the reference frame holding unit 11 is read out and it is determined again whether the current block is a valid block.

When the block change determination unit 12 determines that the current block is a valid block, the block change determination unit 12 determines the retained determination result (change information) of the current block as a valid block (for example, "1").
When the next block is judged to be an invalid block, the judgment result of the held current block is changed to an invalid block,
It outputs to the block information holding unit 14 and the encoding unit 15.

When the block change determination unit 12 recognizes the current block as a valid block, it gives a read instruction to the previous block buffer 17, first reads the previous block, and makes a determination with a threshold value (large).

Based on this determination, the block change determination unit 12
Determines that the current block is a valid block and determines that the valid block information is the block information holding unit 1.
4. Output to the encoding unit 15, and the encoding unit 15 encodes the block data of the current block buffer 10.

If the block change determination unit 12 determines that the preceding block is an invalid block by this determination, it gives a read instruction to the subsequent block buffer 19, reads the next block, and makes a determination with a threshold value (large).

Based on this determination, the block change determination unit 12
Determines that the current block is a valid block and determines that the valid block information is the block information holding unit 1.
4. Output to the encoding unit 15, and the encoding unit 15 encodes the block data of the current block buffer 10.

On the contrary, when the next block is determined to be an invalid block, the current block is recognized as an invalid block, the invalid block information is output to the block information holding unit 14, and the current block buffer 10 is not encoded.

Further, when the block change determination unit 12 determines or recognizes that the block is a valid block, it sends a reference frame rewriting signal to the reference frame holding unit 11, and the reference frame holding unit 1
1 updates the reference frame holding unit 11 with the image data of the block held in the current block buffer 10.

The multiplexer 16 sequentially outputs the block information of the block information holding unit 14 and the variable length code data from the ADCT coding unit 15. That is, when the change determination threshold value is large and the target block is an invalid block with no change,
This is set as a temporary invalid block, the change determination threshold is set to a small value, the change of the current block is determined again, and when it is determined that there is a change in the current block, it is set as a temporary valid block, and the front and rear blocks are set to the threshold ( If the previous or subsequent block is a valid block, the current block is determined to be a valid block.

As described above, the change determination threshold value is switched depending on the presence or absence of a change, and when it is determined that the block is an invalid block, the current block is checked with a low threshold value. Since it is determined whether the subsequent block is a valid block and whether it is a valid block or not is determined based on the results of the blocks before and after, it is possible to effectively set the valid block even when the region changes gently.

This will be described with reference to FIGS. 6 to 8.
In FIG. 6A, in the case where the block changes gently as shown in FIG. 6A, in the regions A, B, and C, the conventional method of comparing with a single threshold value
As shown in (B), the part A at the start of change and the part C at the end of change are recognized as invalid blocks, and the peak part B
Only the effective block appears, and the effective block appears prominently.

On the other hand, in the fourth embodiment, when the block A is determined to be an invalid block with a large threshold value in the scanning direction, as shown in FIG. If the effective block is a valid block, the preceding and subsequent blocks are determined with a threshold value (large). When the subsequent block B is a valid block, the current block A is determined to be a valid block. The start can be effectively detected and the effective block area can be contiguous.

Further, in FIG. 7, as shown in FIG.
When the block B having a small change is isolated in the effective block area, in the conventional technique, as shown in FIG. 7B, the effective block A, the invalid block B, and the effective block C are used.
And the effective block is intermittent, and it looks unnatural, but
In the fourth embodiment, as shown in FIG. 7C, the current block B is re-determined with the threshold value set to be small by recognizing the invalid block B, and when it is determined to be a valid block, the next block is set to the threshold value (large). Since the current block B is determined to be a valid block when it is determined to be a valid block, the block B is also recognized as a valid block and the valid block area can be made continuous.

Similarly, in FIG. 8, when the block change rate is not constant as shown in FIG. 8A, in the conventional technique, effective blocks are intermittent as shown in FIG. 8B. However, in the fourth embodiment, as shown in FIG.
The start of change can be effectively detected, the effective blocks continue, and the image quality can be improved.

(E) Description of Other Embodiments In addition to the above embodiments, the present invention can be modified as follows. The above embodiments may be applied in combination with each other.

The encoding unit and the restoration unit are shown as blocks, but these are represented by the program of the microprocessor.
You can run it. The amount of transmission can be further reduced by encoding and outputting the change information.

The present invention has been described above with reference to the embodiments.
Various modifications are possible within the scope of the invention, and these modifications are not excluded from the scope of the invention.

[0098]

As described above, according to the present invention,
It has the following effects. Only blocks that have been determined to have changed between the previous image and the current image are selected and encoded and output, so the compression rate when there is little change between images can be greatly improved, and the videophone with a fixed camera It is possible to increase the average compression rate of images such as images and animation images.

Even in this way, since the threshold is changed according to the determination result and the block is determined, even if the rate of change of the block is gradually changing, the area from the start portion to the end portion of the change is set as the effective block area. Because it can be set,
It is possible to prevent discontinuous movement of the screen in which the effective block appears to protrude.

[Brief description of drawings]

FIG. 1 is a principle diagram (1) of the present invention.

FIG. 2 is a principle diagram (2) of the present invention.

FIG. 3 is a principle diagram (3) of the present invention.

FIG. 4 is a principle diagram (4) of the present invention.

FIG. 5 is a configuration diagram of a first embodiment of the present invention.

FIG. 6 is a diagram for explaining the operation of the embodiment of the present invention (No. 1).

FIG. 7 is a diagram for explaining the operation of the embodiment of the present invention (No. 2).

FIG. 8 is an explanatory diagram (Part 3) of the operation of the embodiment of the present invention.

FIG. 9 is a configuration diagram of a second embodiment of the present invention.

FIG. 10 is a configuration diagram of a third embodiment of the present invention.

FIG. 11 is a configuration diagram of a fourth embodiment of the present invention.

FIG. 12 is an explanatory diagram (1) of the conventional technique.

FIG. 13 is an explanatory view (No. 2) of the conventional technique.

[Explanation of symbols]

 10 block buffer 11 reference frame holding unit 12 block change determination unit 13 threshold value holding unit 14, block information holding unit 15 ADCT encoding unit 16 multiplexer

Claims (8)

[Claims] 1. An image data encoding method for encoding image data as m × n pixels as a block, wherein a change amount of a pixel at the same position in a block at the same position in a current image and a previous image is detected, A step of comparing the amount of change of the block with a determination threshold to determine whether or not there is a change in the contents of the block, a step of changing the determination threshold based on the determination result, and a determination of an effective block with a change based on the determination And a step of encoding the generated block. 2. The image data encoding method according to claim 1, wherein the determination threshold is changed according to the determination result, and a change in a block adjacent to the block is determined. 3. When the valid block is determined from the determination result, the determination threshold value is changed from large to small,
3. The change of the next block is determined.
Image data encoding method. 4. When the valid block is determined based on the determination result, the determination threshold value is changed from large to small,
The image data encoding method according to claim 2, wherein the change of the previous block is determined again. 5. The image data encoding method according to claim 3, wherein when the determination result determines that the block is an invalid block that does not change, the determination threshold is returned to a large value. 6. The image data encoding according to claim 2, wherein it is determined whether the block is a valid block based on a determination result of the block and a result of determining the adjacent block by changing the determination threshold. Method. 7. When the determination threshold determines that the block is a valid block, the determination threshold is changed from small to large to determine the next block. When the next block is a valid block, the block is determined to be a valid block. The image data encoding method according to claim 6, wherein 8. When the determination threshold is determined to be an invalid block, the determination threshold is changed from large to small, and when the block is re-determined and determined to be a valid block, the determination threshold is decreased. 7. The image data encoding method according to claim 6, wherein the previous or next block is determined largely, and when the previous or next block is a valid block, the block is determined as a valid block.