KR100207405B1 - Variable length encoding method by using discriminator - Google Patents

Abstract

The present invention relates to a variable length encoding method using a delimiter, which enables the hardware to be encoded, real-time processing of data, and high-speed decoding of each code by using a delimiter in encoding. To this end, the present invention provides a variable length encoding apparatus having a delimiter table for a delimiter which is header information of each code, and a code table in which the respective codes are arranged in order of their probability of occurrence. Detecting a probability and accumulating the numbers in turn and calculating a cumulative sum by a power of two; Each time the cumulative value is calculated, the cumulative sum is calculated and the threshold for the total occurrence probability is repeated until the cumulative sum of the probability of occurrence for the newly calculated codewords exceeds the threshold of the total occurrence probability. Doing; Selecting the delimiter from the delimiter table when the comparison result calculation cumulative sum or the new calculation cumulative sum at this step is equal to or greater than a threshold for the overall occurrence rate; If a delimiter for the cumulative sum is selected in this step, assigning a fixed length code in the code table to correspond to the number of codewords; And repeating each step until allocating fixed length codes for all codewords is completed.

Description

Variable length coding method using delimiter

1 is a block diagram of a typical image coding system using MC-DCT suitable for applying a variable length coding method using a separator according to the present invention.

2 is a schematic block diagram of a variable length coding block shown in FIG. 1 for explaining a variable length coding method using a separator according to the present invention.

3 is a flowchart illustrating a process of performing variable length encoding using a separator according to a preferred embodiment of the present invention.

(A) of FIG. 4 is a code table showing an example of variable length codes according to a conventional method, and (b) is a code table showing an example of variable length codes using a separator according to the present invention.

* Explanation of symbols for main parts of the drawings

100, 107: frame memory 110: subtractor

120: video coding block 130: variable length coding block

140: buffer 150: video decoding block

160: adder 180: current frame prediction block

1310 variable length encoder 1330 delimiter table

1350: sign table

The present invention relates to a variable length coding technique for compressing and encoding a video signal. More particularly, the present invention relates to a variable length coding method suitable for real-time processing by variable length coding the transformed cosine transformed coefficient values using a delimiter. .

As is well known in the art, the transmission of discrete image signals can maintain better image quality than analog signals. When a video signal consisting of a series of image frames is represented in digital form, a significant amount of data must be transmitted, especially for high quality television. However, since the usable frequency range of the conventional transmission channel is limited, in order to transmit a large amount of digital data, it is necessary to compress the transmitted data and reduce the amount thereof. As such, among the various compression methods for compressing data, a hybrid coding method combining a stochastic coding method and a temporal and spatial compression method is known to be the most efficient.

Most hybrid coding techniques use motion compensated DCPM (differential pulse code modulation), two-dimensional DCT (discrete cosine transform), quantization of DCT coefficients, variable length coding (VLC), and the like. The motion compensation DPCM determines a motion of an object between a current frame and a previous frame, and predicts a current frame according to the motion of the object to generate a difference signal representing a difference between the current frame and a predicted value. This can be done for example by Staffan Ericsson's Fixed and Adaptive Predictors for Hybrid Predictive / Transform Coding, IEEE Transactions on Communication, COM-33, NO.12 (1985, December), or A motion Compensated Interframe Coding by Ninomiy and Ohtsuka. Scheme for Television Pictures, IEEE Transactions on Communication, COM-30, NO.1 (1982, January).

In general, two-dimensional DCT utilizes or eliminates spatial redundancy between image data, such as digital image data blocks, e. Convert 8 blocks to DCT conversion factor. This technique is described in Chen and Pratt's Scene Adaptive Coder, IEEE Transactions on Communication, COM-32, NO.3 (March 1984). The DCT conversion coefficient is processed through a quantizer, a zigzag scan, a VLC, etc., thereby effectively reducing the amount of data to be transmitted.

More specifically, the motion compensation DPCM predicts the current frame from the previous frame according to the motion of the object estimated between the current frame and the previous frame. The estimated motion may be represented by a two-dimensional motion vector representing the displacement between the previous frame and the current frame.

On the other hand, in the variable length coding scheme, a discrete cosine transformed DCT transform coefficient value is encoded with a variable length, that is, a frequency having a large frequency according to a frequency of occurrence of each code is encoded with a short length code, and a small frequency of occurrence is a long length. The Huffman code is widely known as a typical variable length code.

In general, in implementing a variable length encoding apparatus, it is considered that a particular consideration is real-time processing of data (DCT transformation coefficient) that can minimize hardware complexity and delay time for ease of implementation. Therefore, in view of the above, the present invention relates to the improvement of a variable length coding scheme suitable for simplification of structure and real-time processing of data.

SUMMARY OF THE INVENTION An object of the present invention is to provide a variable length encoding method using a delimiter capable of simplifying hardware for encoding, real-time processing of data, and realizing decoding of codes at high speed by using a delimiter in encoding. There is.

In order to achieve the above object, in the present invention, the quantized DCT transform coefficients are coded using short codes according to the frequency of occurrence of each code using short codes, and shorter codes with shorter frequency are generated. A coding method in a variable length coding apparatus for encoding with a code of claim 1, wherein the variable length coding apparatus comprises: a delimiter table including delimiters which are header information of respective codes, and codes in which the respective codes are arranged in the order of their probability of occurrence. A method comprising: a first step of detecting a probability of occurrence for input codewords, accumulating them in sequence, and calculating a cumulative sum by a power of two; Each time the cumulative value is calculated, the cumulative sum is calculated and the threshold for the total occurrence probability is accumulated until the cumulative sum of the probability of occurrence for the newly generated codewords is equal to or greater than the threshold of the total occurrence probability. A second step of repeating the sum calculating step; A third step of selecting a corresponding delimiter from the delimiter table when the result of the comparison in the second step, the calculated cumulative sum or the new calculated cumulative sum is equal to or greater than a threshold for the total occurrence probability; A fourth step of allocating a fixed length code in the code table to correspond to the number of codewords when the delimiter for the cumulative sum is selected in the third step; And a fifth step of repeatedly performing the first to fourth steps until the assignment of the fixed length code to all the codewords is completed.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of a typical video encoding system suitable for applying a variable length encoding method using a delimiter according to the present invention. As shown in the figure, a typical image encoding system includes a first frame memory 100, a subtractor 110, an image encoding block 120, a variable length encoding block 130, a buffer 140, and an image decoding block. 150, an adder 160, a second frame memory 170, and a predictive frame generation block 180.

Referring to FIG. 1, the input current frame signal is stored in the first frame memory 100 and simultaneously provided to the subtractor 110 and the current frame prediction block 180 through the next line L11.

First, the subtractor 110 performs motion compensation on a moving object provided from the current frame prediction block 180 described later through the line L19 from the current frame signal provided in the first frame memory 100 through the line L11. The predicted current frame signal is subtracted, and the resulting data, i.e., the error signal representing the differential pixel value, is either one of discrete cosine transform (DCT) and one of the quantization methods well known in the art via the image coding block 120. By using, it is encoded into a series of quantized transform coefficients. At this time, the step size of the quantization of the error signal in the video signal coding block 120 is adjusted based on the data layer degree QP from the buffer 140 on the output side.

Then, the quantized DCT transform coefficients are sent to the entropy coding block 130 and the image decoding block 150, respectively. Here, the quantized DCT transform coefficients provided to the variable-length coding block 130 are encoded with a short length code having a large frequency according to the frequency of occurrence of each code using a Huffman code or the like, and having a low frequency of occurrence. It is encoded with a long length code.

More specifically, according to the present invention, the variable length coding block 130 converts the quantized DCT transform coefficients into variable length codes using a delimiter, and thus converts the quantized DCT transform coefficients into codes using a delimiter. This will be described later in detail with reference to FIG. 2 showing the detailed block configuration thereof.

Meanwhile, the quantized transform coefficients provided from the image coding block 120 to the image decoding block 150 are converted into a reconstructed frame signal through inverse quantization and inverse discrete cosine transform, and then provided to the adder 160. The adder 160 adds the reconstructed frame signal from the image decoding block 150 and the predicted current frame signal provided from the current frame prediction block 180 described later through line L19 to reconstruct the previous frame signal. The previous frame signal reconstructed as described above is stored in the second frame memory 170. Accordingly, the immediately previous frame signal for every frame encoded through this path is continuously produced, and the updated previous frame signal will be provided to the current frame prediction block 180 described later for motion estimation and compensation.

FIG. 2 is a schematic block diagram of a variable length coding block of FIG. 1 shown for application to a variable length coding method using a separator according to the present invention. As shown in the figure, the variable length coding block 130 includes a variable length encoder 1310, a delimiter table 1330, and a sign table 1350.

First, according to the present invention, in a variable length encoding method using a delimiter, the variable length coded image data (bit stream) has a fixed identifier corresponding to header information and a code length following the delimiter. It is different from the conventional variable length coding method in that the length is determined. That is, the variable length coding technique using the delimiter according to the present invention may be a combination of the conventional fixed length coding and the variable length coding method. For example, as shown in FIG. As indicated, 0, 10, 110, 1110, ----) and subsequent sign values are each fixed-length by definition. Therefore, unlike the conventional method in which hardware increases as the number of symbols increases, and the values to be stored increase, which leads to complicated encoding and decoding, the present invention can be implemented not only by simple hardware but also by the delay time of data processing. By minimizing this, an improvement in processing speed can be expected.

In other words, the variable length coded image data according to the present invention has predefined values following the header (delimiter) of each code. This is done by decoding (decoding) at. Therefore, the processing speed of the data is greatly improved, and the hardware is simplified, so that the number of tables is reduced, and as a result, the delay time that occurs during decoding can be reduced.

In other words, in the conventional variable length coding method, the code can be changed into a variable code based on the number of symbols to be transmitted and the probability of occurrence. The transmission code is determined based on the generation probability and the tree structure of the symbols to be transmitted. Can be. For example, in the conventional method, as shown in FIG. 4A, as an example, symbol 1 having the highest transmission probability is assigned to symbol 1, and symbol 9 having the lowest probability of occurrence of burning is coded 0000 110. In the present invention, however, the code is defined as a method of distinguishing each code based on the delimiter 0 (or 1).

As an example, as shown in FIG. 4 (b), the present invention defines the number of bits following the delimiter 0, 10, 110, 1110, ---, that is, the sign value 0 or 0 when the delimiter is 0. If 1 is followed and the delimiter is 10, the sign value is not followed. If the delimiter is 110, the code value is 2 bits, and if the delimiter is 1110, the code value is 3 bits. That is, in the present invention, codes corresponding to transmission symbols can be defined in this manner.

Meanwhile, referring to FIG. 2, the variable length encoder 1310 performs encoding using a delimiter on DCT transform coefficients provided from the image encoding block 120 of FIG. 1 through a line L13. That is, the code table 1330 is arranged in a tree structure in order of increasing probability of codewords. In the variable length encoder 1310, the cumulative sum of occurrence probabilities is continuously accumulated and updated. According to whether or not to exceed the threshold that is half of the cumulative sum of the total probability of occurrence, one delimiter provided in the delimiter table 1350 or one delimiter in multiple codewords Variable length coding is performed by the method of assigning.

That is, the variable length encoder 1310 selects a corresponding delimiter from the delimiter table 1330 for DCT transform coefficients on the line L13, and then assigns a fixed length code according to the number of codewords from the sign table 1350. The encoding is performed, and the encoded bit strings are sent to the buffer 140 of FIG. 1 through a line L14.

Next, a process of performing encoding using a delimiter according to the present invention will be described with reference to FIG. 3.

First, in a state in which codes are arranged in the code table 1350 in the order of the high probability of generating a codeword, the variable length encoder 1310 performs the first codeword (k). The probability of occurrence from 0), i.e., the probability of occurrence for one codeword (step 11), and then by the square multiple (i.e. the number of codewords 1, 2, 4, 8 --- -) Accumulate the probability of occurrence one by one to calculate the cumulative sum (step 13). In this case, in the case of the first codeword, the cumulative sum of the probability of occurrence is the probability of occurrence of the corresponding codeword since there is no previous probability of occurrence.

Next, the variable length encoder 1310 compares the calculated cumulative sum (or, if it is the first codeword, its occurrence probability value) with a threshold value T, which is a value obtained by accumulating and dividing all occurrence probability up to half (step 15). If, as a result of the comparison in this step 15, it is determined that the calculated cumulative value (or the occurrence probability value in the case of the first codeword) is smaller than the threshold value T of the total occurrence probability that can be continuously updated, the process is performed in the above-described step. Returning to (13), a new cumulative sum of the occurrence probability of the second codeword and the first occurrence probability described above is calculated, and the comparison with the threshold value T of the continuously updateable new total occurrence probabilities thus generated is performed through step (15). Will be performed again.

On the other hand, if the comparison result in the step 15 determines that the calculated cumulative sum is equal to or greater than the threshold value T of the total occurrence probabilities that can be continuously updated, the process will proceed to the next step 17. Then, in step 17, the delimiter, which is header information, is sequentially selected from the delimiter table 1330.

Accordingly, in the present invention, one delimiter is selected for one codeword or the same delimiter is selected for a plurality of codewords through the above-described process. As an example, when a probability of occurrence of one codeword is greater than or equal to the threshold T of the total probability of occurrence, one delimiter is selected for the corresponding codeword, and the cumulative sum of the probability of occurrence of four consecutive codewords is equal to the total probability of occurrence. Only one delimiter is selected for the four codewords that fall above the threshold T, i.e. one identical delimiter, as can be seen from the fourth to seventh signs in FIG. 4 (b) as an example. Is selected. In FIG. 4 (b), reference numeral a denotes a separator portion, and reference numeral b denotes a code portion.

Next, when the selection of the delimiter for the corresponding codeword (or codewords) is completed in the above step 17, the process proceeds to the next step 19 to fix the fixed length to the number of codewords for the selected delimiter. A sign (ie a fixed bit) is assigned. Then, after the fixed length code is assigned, the variable length encoder 1310 returns k. It is checked whether K-1, i.e., all codewords have been allocated (step 21). As a result of the check in this step 21, if it is determined that the bit allocation for all codewords is not finished, the process returns to the above-described step 13 and repeats the subsequent process, otherwise all codewords If it is determined that bit allocation for these fields has been completed, the process ends.

For example, if there are 12 symbols and the probability that the k-th symbol occurs is Pk, it can be defined as the following equation.

And each occurrence probability P ₀ 0.3, P ₁ 0.2, P ₂ 0.1, P ₃ 0.08, P ₄ 0.07, P ₅ 0.05, P ₆ 0.05, P ₇ 0.04, P ₈ 0.04, P ₉ 0.03, P ₁₀ 0.02, P ₁₁ 0.01, P ₁₂ Assuming as 0.01, the check in the second step 15 third degree result, the occurrence probability of 0.3 of P ₀ does not exceed the half of 0.5 1 (the threshold of the total occurrence probability can continuously update T), the probability P ₀ The value 0.3 and the value 0.2 of P ₁ are added. Therefore, since the cumulative sum is 0.5 and the check condition in step 15 of FIG. 3 is satisfied, the variable length encoder 1310 selects the delimiter 0 from the delimiter table 1330, and selects P ₀ and P ₁ respectively. Assign 0 and 1 That is, the codeword of P ₀ is 00 and the code word of P ₁ is 01.

Then k Since it is not 12, the process returns to step 13, where the value 0.1 of P ₂ is greater than the threshold T of the total probability of occurrence (half the cumulative sum of the probability of occurrence to date: 1-0.5 divided by 2) 0.25. Since it is small, the values of the probability of occurrence P ₂ , P ₃ , P ₄ , and P ₅ are successively accumulated to satisfy the above check condition so that the power is multiplied by _two . Accordingly, since the cumulative sum of consecutively accumulated P ₂ , P ₃ , P ₄ , and P ₅ is 0.3 and exceeds the threshold 0.25 of the total probability of occurrence, the delimiter 10 is selected from the delimiter table 1330, and codewords 00, 01, 10 and 11 are allocated respectively. That is, P ₂ would be 1000, P ₃ would be 1001, P ₄ would be 1010, and P ₅ would be 1011.

Again, the result of the check in step 21, k Since it is not 12, the process returns to step 13, while the value 0.05 of P ₆ is smaller than the threshold T of the total probability of occurrence 0.1 (1-0.8 divided by 2), while satisfying the above check condition. Accumulate successively the values of P ₆ , P ₇ , P ₈ , and P ₉ so that they are powers of two. Therefore, since the cumulative sum of consecutively accumulated P ₆ , P ₇ , P ₈ , and P ₉ is 0.16 and exceeds the threshold value 0.1 of the total probability of occurrence, the delimiter 110 is selected from the delimiter table 1330, and codewords 00, 01 , 10 and 11 are allocated respectively. That is, P ₆ would be 11000, P ₇ would be 11001, P ₈ would be 11010, and P ₉ would be 11011.

As above, the result of check in step 21, k Since it is not 12, the process returns to step 13, and since the value 0.02 of P ₁₀ is equal to the threshold T of the total probability of occurrence 0.02 (1-0.96 divided by 2), the variable length encoder 1310 Selects a separator 1110 from the separator table 1330. Further, the result of the check in step 21, k Since it is not 12, the process returns to step 13, and since the value 0.01 of P ₁₁ is equal to the threshold T of the total occurrence probability 0.01 (1-0.98 divided by 2), the variable length encoder 1310 Selects the delimiter 11110 from the delimiter table 1330.

On the other hand, the check result in step 21, k Since it is not 12, the process returns to step 13, and since the value 0.01 of P ₁₂ is greater than the threshold value T of the total probability of occurrence, the variable length encoder 1310 selects the separator 111110 from the separator table 1330. . Then, the result of the check in step 21, k By 12, the encoding using the delimiter ends.

Accordingly, through the above-described process, the variable length coded bit strings are provided to the output side buffer 140 of FIG. 1 through the line L14, and finally the decoding of the receiving side through the transport channel (not shown). Will be sent to the system.

As described above, according to the present invention, the variable length encoding is performed by using the delimiter, so that high-speed encoding can be realized under real time without delay, and the hardware of the variable length encoding apparatus can be simplified. Have

Claims (3)

In a variable length encoding apparatus, a quantized DCT transform coefficient is encoded with a short length code having a large frequency according to the frequency of occurrence of each code using a code table, and a long length code with a small frequency of occurrence. The coding method according to claim 1, wherein the variable length coding apparatus comprises a delimiter table including delimiters which are header information of respective codes, and a sign table in which the codes are arranged in order of their probability of occurrence. Detecting a probability of occurrence of the codewords and accumulating them sequentially and calculating a cumulative sum by a power of two; Each time the cumulative value is calculated, the cumulative sum is calculated and the threshold for the total occurrence probability is accumulated until the cumulative sum of the probability of occurrence for the newly generated codewords is equal to or greater than the threshold of the total occurrence probability. A second step of repeating the sum calculating step; A third step of selecting a corresponding delimiter from the delimiter table when the result of the comparison in the second step, the calculated cumulative sum or the new calculated cumulative sum is equal to or greater than a threshold for the total occurrence probability; A fourth step of allocating a fixed length code in the code table to correspond to the number of codewords when the delimiter for the cumulative sum is selected in the third step; And a fifth step of repeatedly performing the first to fourth steps until the assignment of the fixed length code to all the codewords is completed. The variable length encoding method of claim 1, wherein the threshold for the total probability of occurrence is a value obtained by accumulating all occurrence probability values up to now and dividing by two. The variable length encoding method of claim 1 or 2, wherein each separator is selectable from at least one or a plurality of codewords based on a probability of occurrence of each codeword.