KR940000680B1 - Picture signal decoding method by using runfield - Google Patents

Abstract

The method reduces a quantity of data by improving an efficiency of AC coefficient (ACC) and reducing an average length of coded language. The system comprises a CPU (10), a ROM (20), a scan section (30) and a RAM (40). The method involves 10 steps of; step for initializing, step for checking all the ACC, step for reading the run field, step for checking a MSB, step for decoding a coded language, step for outputting an ACC, step for checking a decoding process of nonzero coefficient, step for checking an end of decoding process, step for returnning to the 4th step, and step for outputting the ACC and increasing the frequency one step.

Description

Decoding method using run field when receiving and playing video signal

1 is a conventional block diagram.

2 is an inverse zigzag scanning explanatory diagram.

3 is a block diagram according to the present invention.

4 is a flow chart according to the present invention.

5 is a codeword read flow diagram of a nonzero coefficient in FIG.

6 is a codeword decoding flowchart of FIG.

* Explanation of symbols for main parts of the drawings

10: CPU 20: ROM

30: Inverse Zigzag Scan Part 40: RAM

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a decoding method using runfield when receiving and playing video signals, and more particularly, to a decoding method capable of reducing the amount of data.

In order to reproduce the video signal that is reduced-coded at the time of reception, Huffman decoding is performed according to the length of the codeword included in the codeword and the value of the Discrete Cosine Transform (DCT) coefficient. A system for reconstructing an analog video signal of digital data converted into a time domain using an inverse DCT should consider RLC (Run-Length Coding) during Huffman decoding. That is, in the encoding process, the added run length is excessively included and displayed on each bit of the variable length run field. Therefore, the amount is reduced only by dividing and decoding the zero coefficient and the nonzero coefficient according to the information of each bit. Image data can be decoded.

FIG. 1 is a block diagram according to a conventional RLC method, and FIG. 2 is an explanatory view of the inverse zigzag scanning unit 4 in the first diagram.

In the conventional RLC scheme, the codeword introduced through the reception / reproduction system is decoded by the Huffman decoder 1 to obtain a run length and a category. Here, run length indicates how many DCT AC coefficients (0 of 64 DC coefficients are DC coefficients and 1 to 63 AC coefficients) are consecutively 0, and the category is 0, followed by a non-zero AC coefficient. The categories are classified into 0, 1, 2, ... according to the values of the coefficients 0, ± 1, ± 2 to ± 3, ...). The run length counter unit 2 sends as many zeros as the number corresponding to the run length value to the inverse zigzag scan unit 4. The coefficient decoding unit 3 determines a predetermined value based on additional bits and signs applied through the reception / reproducing system within the range corresponding to the value of the category, and determines the predetermined value by the inverse zigzag scan unit 4. Output

Therefore, the inverse zigzag scan unit 4 restores the AC coefficient, which has been zigzag scanned, to the original position in order to increase the run length during encoding.

Meanwhile, the decoding process as described above is defined in the recommendation JPEG regarding data compression.

The RLC scheme as described above has a problem in that many codewords corresponding to 0 to 15 run lengths are required for each category. Therefore, a table defining a codeword corresponding to each pair of run lengths and categories is large, and the length of the average codeword is long accordingly. As a result, the decryption logic is complicated.

Accordingly, an object of the present invention is a decoding method for the encoding method of Patent No. 90-13673, filed by the applicant of the present application, in the decoding method using the run field when receiving and playing the video signal, the run length for the image data DCT AC coefficient It is possible to provide a decoding method that can reduce the amount of data by adopting a separate run field having a separate length and decoding the non-zero AC coefficient by using its own information to increase its efficiency and reduce the average length of codewords.

Hereinafter, the present invention will be described with reference to the accompanying drawings.

FIG. 2 is a block diagram according to the present invention, and includes a CPU (Central Processing Unit) 10 for controlling the system as a whole, and a ROM 20 for storing a table for executing and decoding the CPU 10. And an inverse zigzag scan unit 30 for controlling the AC coefficient decoded by the control of the CPU 10 to be inverse zigzag scanned and output to the RAM 40, and storing the DCT AC coefficient under predetermined control. It consists of the RAM 40.

4 is a flowchart according to the present invention, the first process of initializing the value indicating the frequency (N) of the counter (C) indicating the number of non-zero coefficients, and the non-zero after performing the first process A second step of reading a codeword of a coefficient to determine whether all AC coefficients are zero; a third step of reading a run field when the value is not zero in a second re-addition process; and a value of the most significant bit of the run field is 1; A fourth process of determining whether to be authorized; and a fifth process of outputting a restored non-zero AC coefficient by increasing the frequency N by decoding a non-zero coefficient if it is 1 in the fourth process, and In the fourth process, if the most significant bit of the run field is 0, the sixth process of outputting the AC coefficient value at the corresponding frequency value to 0 and increasing the frequency N by one; A seventh process of determining whether the decoding of the coefficients is finished; When the decoding is not finished in the seventh process or after performing the sixth process, the eighth process of determining whether the AC coefficient is decoded, and if the decoding is not finished in the eighth process corresponds to the next frequency of the run field A ninth process of moving the bit to the highest level and returning to the fourth process; and if the decoding of the non-zero AC coefficient is completed in the seventh process, a tenth process of outputting an AC coefficient for the remaining frequencies to 0 and increasing the frequency by one And an eleventh step of checking whether all AC coefficients for the remaining frequencies are output as 0 in the tenth step.

FIG. 5 is a detailed control flow diagram for the second process (Step A1) of FIG. 4, wherein the first step of reading codewords of nonzero coefficients and the codewords read in the first step of non-zero coefficients are shown. A second step of determining whether it is an end-of-non zero (EONZ) indicating an end; a third step of storing a codeword read if the non-EONZ in the second step is not restored to the AC coefficient; After performing the third step, a fourth step of counting the number of codewords of nonzero coefficients is performed.

FIG. 6 is a detailed control flow diagram of codeword decoding (step A5) of the fifth process in FIG. 4, which is classified into categories through a read step of reading a codeword stored in advance and a codeword read from the read step. A calculation step of calculating a; and an additional read step of reading an additional word transmitted to a receiving / reproducing system to restore a value of an AC coefficient within a range corresponding to the category after performing the calculating step; And a counting step of counting the remaining number of non-zero AC coefficients after performing the sign reading step.

Hereinafter, an embodiment of the operation based on the above-described configuration will be described in detail with reference to FIGS. 3 to 6.

First, the video signal input to the receiving / reproducing system is a reduced coded signal. Therefore, to reproduce this, Huffman decoding is performed according to the length of the codeword included in the reduced coded codeword and the value of the DCT coefficient, and then the coefficients in the frequency domain are converted into the time domain using the inverse DCT.

Therefore, it should be understood that the decoding scheme in the present invention is processed in the reverse process to the coding scheme of Patent No. 90-13673 filed by the applicant of the present application. For example, if DCT is performed on a sub-block unit of 8x8 pixels on the transmitting side, 63 AC coefficients except for the DC coefficient of one of the DCT coefficients by the inverse zigzag scanning unit 20 on the receiving side. Will be stored).

Here, the CPU 10 outputs a DCT AC coefficient by performing a control flow as shown in FIG. 4 according to a program in the ROM 20 and a table for decoding.

In FIG. 4, the CPU 10 initializes a counter (referred to herein as “C”) indicating the number of non-zero coefficients in the AC coefficient. Here, it can be seen that step (A0) corresponds to the first process. After performing the first process, in step (A1), a codeword of a nonzero (nonzero) coefficient is read, and in step (A2), all ACs determine whether the coefficient is zero. The steps (A1-A2) correspond to the second process and the detailed control flow of the step (A1) corresponds to the second process and the step (B1 → B2) of FIG. 5 corresponds to the second process and the details of the step (A1). It should be understood that the control flow consists of the steps of FIG. 5 (B1 → B4). The run field is for indicating whether or not each received bit is an AC coefficient corresponding to its frequency, and the count c is for counting a difference of nonzero coefficients among 63 AC coefficients, for example. It should be noted that the frequency value N is a coefficient value for indicating how many AC coefficients among the 63 AC coefficients are decoded.

Thus, the run field represents the maximum number of 63 bits without sign. After performing step (A3), the CPU 10 determines whether the value of the most significant bit (MSB) of the run field is 1 in step (A4). That is, step (A4) corresponds to the fourth process and determines whether the AC coefficient of the corresponding frequency value N is 0 or 1 according to the value of the most significant bit of the run field. In step (A4), if "1" is non-zero, in step (A5), the codeword of the non-zero coefficient is decoded. This is achieved by performing step (C1? C5) of FIG. When the step A5 is completed, the restored AC coefficient other than zero is output in step A6, and the frequency value is increased by one.

On the other hand, it can be seen that step (A5-A6) corresponds to the fifth process. If the most significant bit of the run field is 0 in step (A4), the process proceeds to step (A7) to output the AC coefficient value at the corresponding frequency value (N) to 0 and increases the frequency value (N) by one. Let's do it. After performing step (A6), in step (A8), the CPU 10 checks whether the decoding of the non-zero AC coefficient is completed by the value of the counter (C), which corresponds to the seventh process. do. If the decoding is not completed in step (A8), it is checked by the frequency value (N) whether all AC coefficients are decoded in step (A9), which corresponds to the eighth process. If the decoding of all the AC coefficients is not completed in step (A9), step (A10) is performed, and the CPU 10 sets the bit corresponding to the next frequency value of the run field to the highest level (MSB). After shifting, the process returns to step (A4). This corresponds to the ninth process. When the decoding of the non-zero AC coefficient in step (A8) is completed (A11), the AC coefficient for the remaining frequencies is output as 0 and the value N is increased by one, which corresponds to step 10. . After performing step (A11), it is checked whether all AC coefficients for the remaining frequencies are output as 0 in step (A12), and if all outputs are completed, otherwise, the method returns to step (A11). It can be seen that this corresponds to the eleventh process.

As described above, in this embodiment, since the run length of the 63 AC coefficients for one sub-block is synthesized in the run field of up to 63 bits, it is possible to know whether or not each coefficient is 0 by checking the run field according to the frequency. It also knows how many nonzero coefficients it has, so you only need to have the corresponding bits of the run field for the last nonzero coefficient. Therefore, the run field needs only as many bits as necessary.

Meanwhile, the present invention can be used for digital CTR, compact disc interactive (CDI), digital camera, electronic game machine, etc., which is a medium for recording / reproducing and recording / reproducing image data, and can transmit / receive image data by reducing coding and decoding. Applications include color fax machines, video phones, video conferencing systems, electronic mailboxes, digital TV broadcast equipment and tuners.

As described above, according to the present invention, in the decoding method of reducing the amount of data by the RLC method when the video signal is received and reproduced, the run length for the DCT AC coefficient of the video data is decoded by using a separate length field. It can be seen that non-zero AC coefficient is a method of decoding only its own information. Therefore, by decoding the non-zero AC coefficient irrespective of the run length, the coding table can be reduced, the average length of the decoded word can be reduced, and the sign and decoding can be simplified. In addition, the run length information can be greatly reduced by decoding the run length by combining the variable length run field.

Claims (3)

In the decoding method using the run field when receiving and reproducing a video signal, a first process of initializing a counter indicating the number of nonzero coefficients and a frequency representing an AC coefficient, and a nonzero operation after the first process is performed. A second step of reading a codeword of a coefficient and determining whether all AC coefficients are zero; a third step of reading a run field when the second step is not 0; and a value of the most significant bit of the run field is 1; A fourth process of determining whether to approve, a fifth process of outputting a non-zero restored AC coefficient after decoding a non-zero coefficient if it is 1 in the fourth process, and increasing the frequency by one; In step 6, if the most significant bit of the run field is 0, the 6th process of outputting the AC coefficient value at the corresponding frequency value as 0 and increasing the frequency by one, and after the fifth process is performed, decoding of the non-zero coefficient is completed. To A seventh process of judging, a process of judging whether the decoding has not been completed in the process of the seventh process or whether the decoding of all the AC coefficients is performed after the process of the process of the sixth process, and The ninth step of returning to the fourth step by moving the bit corresponding to the next frequency of the field to the highest level; and if the decoding of the non-zero AC coefficient is completed in the seventh step, the AC coefficient of the remaining frequencies is output as 0 and the frequency is output. The decoding method using the run field when receiving and playing the video signal, characterized in that the step 10 to increase the one. The end-of-non of claim 1, wherein the second process reads a codeword of a non-zero coefficient and the codeword read in the first stage represents an end of a non-zero coefficient. A second step of determining whether or not zero), a third step of storing the codeword read in order to restore the AC code when the codeword is not EONZ in the second step, and a nonzero coefficient after performing the third step. And a fourth step of counting the number of codewords. The method of claim 1, wherein the fifth process includes: a step of reading a code word stored in advance; a step of calculating a category through a Huffman decoding table with the code word read in the read step; A supplementary read step of reading an adjunct transmitted to a receiving / reproducing system to restore the value of the AC coefficient within a range corresponding to and a sine read step of determining a sign of the AC coefficient by reading a sine after performing the adjunct reading step; And a counting step of counting the remaining number of non-zero AC coefficients after the sign read step is performed.