JPH03250865A - Compression data quantity control system - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] (Industrial Application Field) This invention provides a method for compressing still image data using a variable length code and transmitting or recording the data so that the amount of data after compression becomes the required amount of data. This invention relates to a method for controlling the amount of compressed data to be controlled.

[Conventional technology]

In order to standardize the natural image encoding method, “Ba5eli
ne System” or “Extended System”
Various international standardization methods such as "em" have been proposed.

Figure 7 shows “Ba5elineS”, one of the international standardization methods.
This system divides one input image into multiple blocks of 8×8 pixels, and performs two-dimensional discrete cosine transformation (DCT) on each block. CoCo5
1neTransform is performed (processing P1), and quantization is performed by dividing the obtained DCT coefficient by each threshold of a quantization matrix consisting of 8×8 thresholds (processing P2).

FIGS. 8 and 9 are examples of quantization matrices for luminance signals and color difference signals.

The difference between the direct current (DC) component of the quantized DCT coefficient and the DC component quantized in the previous block is taken, and the number of bits of the difference is Huffman encoded.

The alternating current (AC) component is converted into a one-dimensional sequence by performing zigzag scanning within the block, and then subjected to two-dimensional Huffman encoding using the number of bits of effective coefficients and the number of consecutive zeros (invalid coefficients) (Process P3 and P4).

FIG. 10 shows an example of a zigzag scan table.

Note that during quantization in process P2, each threshold value of the quantization matrix is multiplied by a certain coefficient (scale factor), and then the DCT coefficient is divided. The image quality and compression rate of the compressed image are adjusted by this scale factor.

The data compressed in this way is decompressed by the reverse process of processes P1 to P4. That is, Huffman decoding in process P5, decoding of DC and AC components in process P6, inverse quantization in process P7, and inverse DCT (IDCT) in process P8.

[Problems to be Solved by the Invention] By the way, in the above-mentioned system, data compression is performed using a Huffman code, which is a variable length code, so the total amount of data after compression is the same as that at the end of the compression process (processes P1 to P4). You won't know until you do. Therefore, if it is necessary to perform encoding within a preset data amount range, some kind of data amount control is required.

An object of the present invention is to provide a compressed data amount control method that can control the amount of data after compression to be the required amount of data.

[Means to solve the problem]

In this invention, one digital image is divided into one block n×
Divide into multiple blocks each consisting of n pixels, perform discrete cosine transformation for each block, and apply the resulting n×n transform coefficients to each threshold of a quantization matrix consisting of n×n thresholds. Perform division and quantization, convert the transform coefficients after discrete cosine transform or quantization into a one-dimensional sequence in a fixed order from the DC component to the high frequency component, and then convert the resulting sequence into a one-dimensional sequence. This is a data compression method that encodes the number of zeros in a one-dimensional sequence ap (p=0.1
．． ..., n''-1), set the coefficient ap that satisfies the condition ``k≦P≦n2-IJ'' in the sequence ap to zero, and then In addition to encoding the number of data and compressing the data, A of the discrete cosine transform coefficient
The cumulative distribution of the sum of the squares of the C component is determined for each block, and the function G is proportionally converted to the desired amount of compressed data from this cumulative distribution.
Determine this function G and the cumulative value of the actual compressed data amount■
is compared for each block, and the value of the variable is adjusted for each block so that the cumulative value ■ follows the function G, thereby controlling the amount of data after compression to be the desired amount of data.

[Operation] In this invention, the adjustment of variables for each block is performed as follows. First, block 5 (s=12,...
2 of the AC component of the discrete cosine transform coefficient F uv for each T)
Find the sum of the products. Next, calculate the cumulative distribution Bs for all blocks of the sum of squares of the AC components, and then calculate the data amount function G, which is proportionally converted so that the maximum value B↑ of the cumulative distribution B3 corresponds to the desired data amount after compression. Calculate using the following formula.

The data amount function G obtained in this way is compared with the cumulative value ■3 of the actual compressed data amount for each block, and if G<V, the value obtained by subtracting the adjustment value Δk from the variable k is used as the variable for the next block. If V<G, the value obtained by adding the adjustment value Δk to the variable k is set as the variable of the next block, and if V s =G s, it is left as is.

In this way, the value of the variable is set for each block, and the value of "k≦p≦n!" in the one-dimensional sequence ap of the transform coefficient after quantization is set.
1. The coefficient ap that satisfies the following relationship is set to zero, and the compressed data amount cumulative value 3 is made to follow the function G.

[Embodiment] FIG. 1 is a schematic diagram showing an embodiment of the processing procedure of the compressed data amount control method according to the present invention, and the same parts as in FIG. 7 are given the same reference numerals and will be described.

First, input image data is divided horizontally and vertically into a block of n×n pixels, for example, a plurality of p×q blocks each consisting of 8×8 pixels, and a two-dimensional discrete cosine transform (DCT) is applied to each block ( Processing PL).

DCT is a type of orthogonal transformation in the frequency domain, where the transform coefficients are Fuv (u, v=o, 1.-, n-1), and the input image data for one block is f47 (i, j=0.I,・
-, n-1), where G=1/72 (w=o) =1 (W≠0). The obtained transformation coefficient F uv indicates a component obtained by decomposing one block of input image data into spatial frequencies.

In the conversion coefficient F uv, the coefficient F0° indicates a value (DC component) proportional to the average value of n×n pixels of the input image data f ij, and as u and v become larger, the component with a higher spatial frequency (AC component) represents.

The DCT coefficients F uv obtained as in step 2 are stored in the memory for each screen (process P10). The stored coefficient F uv is the sum of squares of AC components b s (s=1, 2.-, pXq) expressed by the following equation for each block (processing pH).

Next, for all blocks on one screen, the cumulative distribution Bs of the sum of squares expressed by the following formula is determined for each block (processing P
12), Furthermore, the maximum value of the cumulative distribution B, that is, the AC component 2
B7 (T=pXq
) is proportionally converted so that it corresponds to a predetermined data amount after compression (process P13).

FIG. 2 shows an example of the sum of squares 5, the cumulative distribution B, and the data amount function G.

The DCT coefficients for one screen stored in the memory are read out block by block, divided by the value obtained by multiplying each threshold value of a quantization matrix consisting of n×n threshold values by a scale factor, and quantized ( Processing P2). Multiplication processing using a scale factor is performed by bit-shifting each threshold value of the quantization matrix.

Next, for the DC component, a difference is taken from the component quantized in the previous block (processing P30), and the number of bits of the difference is Huffman encoded (processing P4).

For the AC component, zigzag scanning is performed in the order shown in Figure 10, and a one-dimensional sequence ap (p=1+2+...,
63) (processing P31). FIG. 3 shows a matrix of DCT coefficients after quantization, and FIG. 4(a) shows a one-dimensional sequence ap after zigzag scanning.

Next, a process is performed for this sequence ap in which a coefficient a9 that satisfies the relationship "k≦p≦63" is set to zero (process P32).

FIG. 4(b) shows an example in which rk=8J.

As is clear from FIGS. 3 and 4, the process P32 is a process for zeroing out the high frequency components after the second in the sequence ap, and corresponds to a type of low-pass filter process.

Therefore, by changing the value of k, the number of consecutive zeros can be changed, and the amount of data after compression, that is, the compression ratio can be adjusted.

Next, the number of consecutive zeros is run-length encoded (process P33), and two-dimensional Huffman encoding is performed using the run-length encoded data on the number of consecutive zeros and the number of bits of the effective coefficient (process P4).

Huffman encoding does not use the quantized coefficient values themselves for both DC and AC components, but performs Huffman encoding on the number of bits necessary to express the values.

In addition to the Huffman code, the value of the number of bits is added as additional information. For example, if the quantized coefficient is 2(1
0 base), if expressed in binary, it would be “000...
・010'', but the number of bits required to express this, 2, is Huffman encoded as a value that represents this value,
Add 2-bit data "10" as an additional bit.

On the other hand, if the quantized coefficient is negative, data obtained by subtracting 1 from the additional bit is added. For example, if the quantized coefficient is −
2 (decimal number), when expressed in binary number (2's complement representation), it becomes "'111...110", and the lower 2
The bit becomes an additional bit, and "01", which is obtained by subtracting "1" from "10", is added as an additional bit. By doing this, when the quantized coefficient is positive, the additional focus is 1.
If it is negative, it starts with 0, making it easy to distinguish between positive and negative.

Next, the amount of compressed image data of one block of compressed image data is counted and processed P14), and the cumulative amount of data V is calculated.
, and the data amount function G obtained in process P13 are compared (process P15), and the variables in the low-pass filter process are adjusted (process P16).

That is, if G3<V, the variable k is set to "k-Δk" to increase the number of consecutive zero data, and the amount of data after compression is decreased, and vice versa.

If <G, then the variable k is set to "k+Δk" to reduce the number of consecutive zero data and increase the amount of data after compression; if Vs''Os, leave it as is and perform data compression processing for the next block.

This series of processing is repeated for all blocks, and the fifth
As shown in the figure, the variable k is controlled for each block so that the cumulative data amount after compression, ■, follows the data amount function G, and becomes the desired data amount.

To decompress compressed data, conventional data decompression processing may be performed. That is, first Huffman decoding is performed (processing P5), differential decoding is performed for the DC component, and run-length decoding is performed for the AC component, and then the data is rearranged in the order of zigzag scan to obtain the transform coefficients for one block. (processing P6), inverse quantization is performed by multiplying this transform coefficient by the value obtained by multiplying each threshold value of the quantization matrix by the scale factor (processing P7), and further, inverse discrete cosine transform (ID CT) is performed. (process P8), and the decompression process ends. In this case, the variables used at the time of data expansion are no longer needed at the time of data expansion, so that the conventional method can be used for expansion processing.

Next, the operation of the data compression process will be explained in detail with reference to the flowchart of FIG.

First, initial settings of the system are performed (step S1). That is, each value of the total sum Br and the data amount function G for all blocks of the cumulative distribution B5. Set the value of S to the initial value "1".

Next, the first block of image data is input (step S2), and DCT is performed (step 33). DCT
The conversion coefficient F uV obtained by is temporarily stored in memory (step S4). Next, the sum of the squares of the AC components of the block is calculated (step S5), and its cumulative distribution B is obtained (step S6). In this case, since the first block (S=1) is being processed, Bs = bt.

Next, determine whether all blocks have been processed (
Step 37). In this case, since the processing of all blocks has not yet been completed, "1" is added to the block number S (step S8), the image data of the next block is input (step S2), and the steps S3 to S7 described above are performed. Repeat the process.

When all blocks have been processed, a data amount function G is calculated using the block number S as a variable (step S9), and an initial value ko is set as a variable for low-pass filter processing (step 510). The DCT coefficients of the first block are read out from the DCT coefficients of the screen and quantized (step 511).

If the coefficient after quantization is a DC component (step 5L2)
, and the DC component quantized in the previous block, and the difference value is encoded (step 513).

If the coefficient after quantization is an AC component (step 512)
, a zigzag scan is performed to convert it into a one-dimensional number sequence ap (step 514), and a low-pass filter process is performed to zero out the high frequency components after the number sequence ap of this number sequence ap (step 315) to compress the number of consecutive zeros. Run-length encoding is performed (step 516).

Next, for the DC component, the encoded difference value is Huffman encoded, and for the AC component, the run length encoded data on the number of consecutive zeros and the number of bits of the effective coefficient are
Huffman encoding of the dimensions is performed (step 517).

Next, it is determined whether all blocks have been processed (step 318), and if all blocks have been processed, the compression process is ended. In this case, only the first block has been processed, so the cumulative amount of data after compression is
The process proceeds to step 519. and,
The obtained cumulative data amount ■s and the data amount function G obtained in step S9.

(step 520), if cs < v, set variable k to "k-Δk" and increase the number of consecutive zero data (step 521); if Vs < Gs, set variable k to "k+Δk" and increase the number of consecutive zero data (step 521). Reduce the number of zero data (step 522), v,=c.

If so, data compression processing for the next block (steps 311-317) is performed using the same values.

In this way, data compression is performed while adjusting the values of variables for each block, and as shown in Figure 5, after compression, the cumulative data amount ■, is controlled to follow the data amount function G, so that the amount of data becomes the desired amount of data.

Note that instead of using a constant value as the amount of change Δk in a variable, for example, if Δk=a (Vs Gs ) (a: constant), the data amount function Gs and the actual data amount ■.

This will feed back the error between the two, further improving the tracking performance.

In addition, in this embodiment, the low-pass filter processing (processing P32) during the data compression processing is replaced by the quantization processing (processing P2).
Although this is performed after the quantization process, it may also be performed before the quantization process.

[Effect of the invention] According to this invention, a matrix of two-dimensional DCT coefficients is
Convert the DC component to a one-dimensional sequence in a fixed order from the DC component to the high frequency component, and calculate the DC of the high frequency component after the arbitrary number.
Using a method of adjusting the amount of data by setting the T coefficient to zero, the amount of data after compression is determined in advance for each block of the image, a data amount function is determined, and this determined data amount function is used to adjust the amount of data when actually compressed. Data compression is performed while changing the value of the variable for each block so that the actual compressed data amount follows the data amount function. It becomes possible to control the amount of data to a desired amount of data.

Also, when determining the data amount function, the DC of each block
Since it is based on the sum of the squares of the AC components of the T coefficient, it is possible to allocate a large amount of data to areas where the image signal changes rapidly, and a small amount of data to areas where the change is gradual. Natural compression can be performed according to the situation.

Furthermore, since variables used during data compression are not required during data expansion, data expansion can maintain compatibility with conventional methods.

In addition, if the variable k is changed not by a fixed range of change but by a value corresponding to the error between the actual compressed data amount and the data amount function, the actual compressed data amount can follow the data amount function. Performance is further improved.

[Brief explanation of drawings]

FIG. 1 is a diagram showing an embodiment of the processing procedure of the compressed data amount control method according to the present invention, and FIG.
Sum of powers bs, cumulative distribution bs! Figure 3 is a matrix table showing DCT coefficients after quantization; Figure 4 is a table in which DCT coefficients are converted into a one-dimensional sequence; Figure 5 is a data volume function. Fig. 6 is a flowchart for explaining the data compression operation of Fig. 1, Fig. 7 is a diagram showing the processing procedure of conventional compression/expansion processing, Fig. 8 9 is a table showing the quantization matrix of the luminance signal, FIG. 9 is a table showing the quantization matrix of the color difference signal, and FIG. 10 is a table showing the zigzag scan table. Figure (a) (b) (0 Figure block number p^q

Claims (2)

[Claims] (1) Divide one digital image into multiple blocks each block consisting of n×n pixels, perform discrete cosine transformation for each block, and convert the resulting n×n transform coefficients into n Quantization is performed by dividing by each threshold of a quantization matrix consisting of x n thresholds, and the transform coefficients after the above discrete cosine transform or after the above quantization are linearly arranged in a fixed order from the DC component to the high frequency component. A data compression method that converts the one-dimensional number sequence a_p(p
= 0, 1, ..., n^2-1), and set the coefficient a_p that satisfies the condition "k≦p≦n^2-1" in the above sequence a_p to zero. After that, data compression is performed by encoding the number of consecutive zeros, and the cumulative distribution of the sum of squares of the AC components of the above-mentioned discrete cosine transform coefficients is obtained for each block, and from this cumulative distribution, the desired amount of compressed data is determined. A function G proportionally converted to is determined, and this function G and the cumulative value V of the actual compressed data amount are compared for each block, and the value of the variable k is determined so that the cumulative value V follows the function G. A compressed data amount control method characterized in that the amount of compressed data is adjusted for each block. (2) Let the adjustment value Δk of the variable k for each block be Δ
2. The compressed data amount control method according to claim 1, wherein k=a(V-G) (a: constant).