KR0148226B1 - Code method and apparatus of quantizer - Google Patents

Abstract

The present invention relates to a method and apparatus for encoding a quantization level, and more particularly, to encoding and transmitting variable length data instead of fixed length data when transmitting a value of a quantization level in a video signal compression method. In variable length encoding of the quantization level, the difference from the previous quantization level is obtained, and the difference length encoding is performed to reduce the amount of data generation. Furthermore, when setting the variable length encoding table, if the difference between the two quantization levels is zero, by not assigning any codeword, the data generation amount can be further reduced, thereby increasing the data compression efficiency.

Description

Quantization level encoding method and apparatus

1 is a block diagram showing an encoding apparatus for compressing general video data.

2 is a conceptual diagram showing a processing unit of image data.

3 is a block diagram showing an encoding device of a quantization level according to the present invention.

* Explanation of symbols for main parts of the drawings

31: variable length encoder 35: memory

A1, A2: Adder

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an encoding apparatus for compressing data amount by encoding image data. In particular, the present invention relates to DPCM (Differential Pulse Code Modulation) of quantization levels between adjacent macroblocks, and variable length encoding to reduce the number of coding bits of the quantization level. The present invention relates to a method and an apparatus for encoding a quantization level for further improving the compression rate of data.

In general, a representative coding system for digital transmission and reception of a video signal is a DCT (Discrete Cosine Transform) encoding system using dynamic compensation. The encoding system uses a very high correlation between picture data to obtain a high data compression ratio by encoding a difference signal between frame data. Coding systems using such compensation are widely used in digital TVs, digital HDTVs, video phones, and digital VCRs.

FIG. 1 is a block diagram of a general encoding apparatus for compressing image data, in which input digital image data Vin is divided into small blocks of a predetermined size. Then, the motion estimation unit 16 detects the motion vector (MV) by predicting the motion from the portion of the image data of the previous screen stored in the frame memory 15 corresponding to the currently input block data with respect to the input block data. Motion Estimation is performed. The motion compensator 17 performs motion compensation by extracting the block of the previous screen closest to the current block by using the motion vector MV obtained from the motion estimation unit 16. Then, the encoding process processes only the currently input block (intra mode) or processes the difference signal between the block that is synchronously compensated with the current input block (inter mode). This encoding process is performed by the DCT unit 11, the quantization unit 12, and the variable length encoding unit 18, and the inverse quantization unit 13 and the inverse to store the image data of the previous frame in the frame memory 15. There is a DCT section 14.

In addition, the buffer 19 is controlled to maintain a constant bit rate of the encoded and transmitted signal. Slice the quantization level to adjust the amount of data generated by the quantization unit 12 according to the cumulative state of the buffer 19 in order to prevent overflow or underflow of the buffer 19. Occurs in (Slice) units. Recently, the omnidirectional analysis unit 20 is provided to detect the image characteristics of the image data in units of macroblocks according to the energy distribution characteristics of the block data supplied from the DCT unit 11, and from the buffer 19 according to the image characteristics. The supplied quantization level SQL is varied to generate the quantization level MQL in macroblock units.

FIG. 2 shows one frame of video data encoded by the encoding apparatus of FIG. As shown, one frame is composed of a plurality of slices, and this slice is a unit for applying a quantization level (SQL) determined by the state of the buffer 19. Also, a slice is divided into macroblocks of a predetermined size unit for encoding, and the macroblock is also composed of a plurality of blocks. This macroblock is a unit for applying the quantization level (MQL) determined by the image characteristic of the image data.

Since the quantization level that controls the amount of data generation is mainly between 0 and 31, at least 5 bits (2 ⁵ = 32) are required to encode one quantization level. Furthermore, when the quantization level (MQL) occurs in macroblock units, a large amount of bits are required to encode the quantization levels. Therefore, a need for more efficient encoding is required for encoding and transmitting the quantization levels.

Accordingly, an object of the present invention is to provide a method of reducing the amount of data generated by reducing the number of bits required for encoding a quantization level by obtaining a difference between quantization levels of adjacent macroblocks and variable length encoding the difference.

Another object of the present invention is to provide an apparatus for implementing the variable length coding method of the quantization level described above.

The quantization level encoding method according to the present invention for achieving the above object, in the encoding method for transmitting the quantization level generated in the image data encoding system, the quantization is generated corresponding to the image data of a certain size unit Receiving a level, calculating a difference value between two adjacent quantization levels, and calculating the difference based on a variable length code table defined in relation to the difference value of the quantization level and the variable length codeword. Variable length encoding the value.

An apparatus for encoding a quantization level for achieving another object of the present invention is an apparatus for encoding a quantization level in an image compression encoding system, comprising: a quantization level of a current macroblock to be input and a quantization level of a previous macroblock to be fed back A first adder for calculating a difference value between the variable length encoder and a variable length encoder for outputting a code word corresponding to the difference value of the quantization level calculated by the first adder based on a predefined variable length table; A second adder for restoring the quantization level of the current macroblock by adding the difference value of the quantization level calculated by the first adder to the quantization level of the previous macroblock, and a quantization level of the current macroblock restored in the second adder Storing and storing the quantized level of the restored previous macroblock into the first adder and the second adder, respectively. It includes a memory for inputting.

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

3 is a block diagram illustrating an apparatus for encoding a quantization level according to an embodiment of the present invention. The encoding apparatus of the quantization level shown in FIG. 3 includes a first adder A1 that calculates a difference between the quantization level MQL of the input macroblock and the predetermined feedback data. An output terminal of the first adder A1 is provided with a variable length code table, and a variable length coder 31 for encoding the difference value calculated from the first adder A1 according to the variable length code table is connected. The apparatus of FIG. 3 also includes a second adder A2 which adds the output value of the first adder A1 and the feedback data to restore the quantization level of the macroblock. A memory 35 is connected to an output terminal of the second adder A2 to store the quantization level of the macroblock obtained from the second adder A2, and to feed back the inputs to the first adder A1 and the second adder A2. It is constructed.

First, the variable length encoder 31 receiving the macroblock unit quantization level (MQL) input to the quantizer (not shown) performs variable length encoding on the quantization level (MQL) according to a preset variable length table to be transmitted. do.

On the other hand, since the value of one quantization level per macroblock has the same value as the quantization level of the adjacent blocks statistically, the difference is likely to appear very small value. Therefore, the quantization level MQL applied to the variable length encoder 31 is also input to the second adder A2 and stored in the memory 35. The first adder A1, which receives the quantization level MQL input through the input terminal, calculates a difference value from the quantization level of the previous macroblock applied from the memory 35. The difference value calculated by the first adder A1 is applied to the variable length encoder 31 and the second adder A2, respectively. The variable length encoder 31 variably encodes the quantization level difference value of the adjacent macroblock calculated from the first adder A1. The second adder A2 adds the difference between the quantization level of the current macroblock calculated from the first adder A1 and the quantization level of the previous macroblock, and the quantization level of the previous macroblock applied from the memory 35. Restore the quantization level of the macroblock. The quantization level of the reconstructed macroblock is stored in the memory 35. The memory 35 applies the quantization level of the previous macroblock stored to the first adder A1 and the second adder A2, respectively, and the quantization level of the restored current macroblock applied from the second adder A2. Save it. That is, first, the input quantization level is encoded and transmitted as it is, and after that, only the difference between two adjacent quantization levels is encoded and transmitted.

If the variable length encoder 31 presets the variable length table as shown in Table 1 below, the variable length encoder variably encodes the quantization level of the generated macroblock as shown in Table 2 below.

The variable length code table of Table 1 is obtained through an optimization process.

As shown in Table 2 above, the amount of data generated in the case of encoding the difference value between the quantization levels (MQL) by the variable length code is compared with the case where each of the quantization levels (MQL) is coded by the fixed length, that is, 5-bit fixed length code. It can be seen that less.

In addition, if the variable length encoder 31 sets the variable length table as shown in the following Table 3, the variable length code is set when the two adjacent quantization levels are the same and the difference value is zero (0). By not assigning, the amount of data can be further reduced.

If the quantization level of the previous macroblock is the same as the quantization level of the current macroblock, it may indicate that there is no change in the quantization level of the macroblock by transmitting side information as in the video signal compression method of the Moving Picture Experts Group (MPEG). If the difference between the quantization levels is 0 during quantization level coding, the code rate is not assigned, thereby reducing the data generation rate and further increasing the data compression rate.

As described above, the method and apparatus for encoding a quantization level according to the present invention can further reduce the amount of information by obtaining the difference between two adjacent quantization levels and performing variable length encoding, as compared with the conventional method of fixed length encoding for transmitting the quantization level. The data compression efficiency can be further improved.

Claims (6)

An apparatus for encoding a quantization level in an image compression encoding system, comprising: a first adder for calculating a difference value between a quantization level of a current macroblock input and a quantization level of a feedback macroblock input; A variable length encoder for outputting a code word corresponding to a difference value of the quantization level calculated by the first adder based on a predetermined variable length code table; A second adder for restoring the quantization level of the current macroblock by adding a difference value of the quantization level calculated by the first adder to the quantization level of the feedback macroblock inputted; And a memory for storing the quantization level of the current macroblock reconstructed by the second adder and feedback the quantization level of the reconstructed previous macroblock to the first adder and the second adder, respectively. Encoding device. The quantization level encoding apparatus of claim 1, wherein the variable length encoder further comprises a variable length encoding table which does not allocate a code word when a difference value of the quantization level is 0. CLAIMS 1. A method for encoding a quantization level generated by an image data encoding system, the encoding method comprising: receiving the generated quantization level corresponding to image data of a predetermined size unit; Calculating a difference value between two adjacent adjacent quantization levels; And variable length encoding the calculated difference value based on the variable length code table defined by the relationship between the difference value of the quantization level and the variable length codeword. The method of claim 3, wherein the calculating of the difference value comprises: calculating a difference value between a current quantization level in macroblock units input and a previous quantization level in macroblock units inputted; Restoring the current quantization level by adding the calculated difference value to a previous quantization level of the feedback macroblock unit; And storing the reconstructed quantization level, and feedbacking the previous quantization level stored therein. The method of claim 3, wherein in the variable length encoding step, when two adjacent quantization levels are equal to each other and the calculated difference between the quantization levels becomes 0, codewords are not assigned based on the variable length code table. Encoding method of quantization level. The apparatus of claim 1, wherein the variable length code table is defined by a relationship between a difference value of a quantization level and a variable length codeword obtained through an optimization process.