KR0162201B1 - Dc dpcm system for image data - Google Patents

Abstract

The present invention relates to a system for DPCM of DC components of image data by data distance units of a predetermined size in a system for dividing and encoding each frame of image data into a plurality of windows. DPCM, the first block of each window slice performs DPCM and DPCM of the previous slice, and periodically encodes MMB composed of a plurality of window slices in intra mode periodically or when the power is turned on or when switching channels. send. In this case, the first block data of each MMB encodes a difference from '128', which is an average DC value of block data. The DC-DPCM system of image data can be usefully used for encoding DC component data in a system for encoding image data by lowering a processing frequency band in a multi-window method.

Description

DC component-differential pulse code modulation system of image data

1 is a block diagram showing an example of a conventional multi-window encoder.

2 is a block diagram showing a specific example of a window encoder in the apparatus of FIG.

3 is a conceptual diagram showing a DC-DPCM method of image data according to the present invention.

4 is a block diagram showing a DC-DPCM apparatus for image data according to the present invention.

5 is an operation timing diagram for the apparatus of FIG.

* Explanation of symbols for main parts of the drawings

12: split screen 14: multiplexer

42, 44, 46: flip-flop 48: selection

WM ₁ , WM ₂ ‥‥‥ WM _n : Windows memory

WE ₁ , WE ₂ ‥‥‥ WE _n : Window encoder

The present invention relates to a system for encoding video data in a device for transmitting and receiving video signals. In particular, the present invention relates to a system for encoding video data by dividing a screen of video data into a plurality of windows and converting the data processing frequency into a low frequency to encode the video data. The present invention relates to a DC-DPCM system of image data for DPCM (Differential Pulse Code Modulation).

In general, when transmitting and receiving video signals such as a TV, the transmitting side encodes the image data, and decodes the image data received on the receiving side to restore the original signal. As a video data processing method used in encoding, a DPCM method that obtains a high data compression rate by encoding a difference signal between frames using a high correlation between screens of video data. This coding method is being standardized and developed by the Moving Picture Expert Group (MPEG), which is currently being standardized by ISO / CCITT, and is also used in the digital HD-TV method developed in the United States. It can also be used in communication conference systems and digital VTRs.

The above-described encoding method of image data has an inter mode and an intra mode as an image data processing mode. The inter mode detects a motion vector based on the amount of motion on the previous screen and the video data block to be encoded and encodes and transmits a difference signal between the block data compensated by the motion vector and the current block data. In addition, the intra mode encodes the image data using the degree of correlation of the image data in the screen to be currently encoded. At this time, DC-DPCM is performed to obtain and encode a difference of DC components between adjacent blocks with respect to the DC components of the DCT and the quantized coefficients. In this DC-DPCM method, encoding efficiency of image data can be improved by using correlation between DC components.

The above-described inter-mode and intra-mode are determined in units of data blocks of a predetermined size, and a mode signal of mass inter mode or intra mode for each data block is transmitted to a decoder for use in decoding the coded block data. In the conventional video data coding scheme developed in MPEG, the inter mode and the intra mode are determined in macroblock units consisting of a plurality of blocks, and DC-DPCM is performed on the macroblock data only in the intra mode. In addition, in order to prevent a data error from occurring when the power is turned on or the channel is changed, an entire frame is encoded and transmitted in the intra mode, and the frame is periodically encoded in the intra mode. When intra-mode encoding is performed at a predetermined period, an entire macro frame is encoded in an intra mode over a plurality of frames, without being encoded in the intra mode. Each frame of the image data encoded in the encoding method of the image data is divided into a plurality of windows to encode the image data for each window, and the encoded image data is similarly decoded for each window.

1 shows an example of a conventional multi-window encoder. The apparatus of FIG. 1 includes a screen splitter 12 for dividing image data V _H input in units of frames into a plurality of windows, and a plurality of window memories WM ₁ and WM for storing image data of each divided window. _{2 ,,} ‥‥‥ WM _n ), a plurality of window encoders (WE ₁ , WE ₂ ‥‥‥, WE _n ) for encoding the image data stored in each window memory, and the image data encoded by each window encoder It is composed of a multiplexer 14 to select and output (V _OUT ).

Such an encoding apparatus can be used for encoding HD-TV signals. The encoding apparatus divides input image data to be encoded into a plurality of window regions and encodes image data at a low processing frequency. As such, an example of a window encoder for encoding each window data is shown in FIG. 2.

In FIG. 2, block data D _IN having a predetermined size is sequentially input, and the block data are converted into DCT coefficients in the first DCT unit 21 as blocks in an intra mode, and the DC-DPCM unit 22 is converted into DCT coefficients. The DC value is DPCMd and then quantized in the first quantization unit 23 according to a predetermined quantization level. Also, the first adder ADD1 calculates the input block data D _IN , the motion compensated block data, and the difference component. The image data of this difference component is converted into a DCT coefficient by the second DCT unit 24 and then delayed by the delay unit 25 by a predetermined time to match the output data of the DC-DPCM unit 22 in time. The output data of the delay unit 25 is quantized by the second quantization unit 26 according to a predetermined quantization level. The intra mode block data and the inter mode block data output from the first quantization unit 23 and the second quantization unit 26 are respectively supplied to the mode selection unit 27. The mode selector 27 selects and outputs block data of intra mode or inter mode according to the mode signal. The output data of the mode selector 27 is supplied to the transmission channel via the variable length encoder 28 and the buffer 29.

On the other hand, in order to obtain the block data compensated for motion in the DPCM process, there is a decoding unit of the inverse quantization unit 30, the inverse DCT unit 31 and the DC restoration unit 32, the restored block data is stored in the frame memory 33 Stored. The motion prediction unit 34 detects a motion vector by predicting a motion from a portion of the data of the previous frame stored in the frame memory 33 corresponding to the currently input block data. The motion compensator 35 extracts the block of the previous screen closest to the current block data using this motion vector and supplies the same to the first adder ADD1 described above.

In such an apparatus for encoding image data, an object of the present invention is to provide an encoding method for DPCM DC components of image data in a system for dividing a frame into a plurality of windows and encoding the image data in a lower frequency band. .

Another object of the present invention is to provide a DC-DPCM device of image data for implementing the above-described DC-DPCM method.

The purpose of this method is to divide a frame into a plurality of windows in an image data processing method and to encode a DC component of the image data at a reduced processing frequency, which is different from the first block data of an MMB composed of a predetermined number of window slices. A first step of encoding a difference between the average DC values of the block data, the block data after the first block data of the MMB, a second step of performing DC-DPCM between two adjacent block data, and a slice divided by the window. The first block of the corresponding window slice includes a third step of performing DPCM and DPCM of the previous window slice, and a fourth step of regularly encoding the MMB in an intra mode by a DC-DPCM method of image data. Is achieved.

Another object of the present invention is to provide an apparatus for encoding a DC component of image data at a reduced processing frequency by dividing a frame into a plurality of windows in an image data processing apparatus, wherein the quantized image data is a block having a predetermined size. A first input terminal input in units and a second input terminal in which a block synchronization signal having a predetermined pulse is input at a DC component position of each block data in synchronization with the block data, and the block data applied to the first input terminal; And a first latching means for latching a DC component of the block data by receiving a block synchronous signal applied to the second input terminal, a block synchronous signal applied to the output data of the first latching means and the second input terminal. Means for delaying the output data of the first latching means for one block period; a window slab corresponding to a slice divided by the window; And a third input terminal to which a window slice synchronization signal having a predetermined pulse is input at a DC component position of the predetermined block data of each window slice in synchronization with the switch, and the block data applied to the first input terminal and the third input terminal. A second latching means for latching a DC component of a predetermined block data of the window slice by receiving a window slice synchronization signal applied to the window slice; a first terminal to which an output data of the delay means is applied; and an output of the second latching means. A second terminal to which data is applied, a third terminal to which a predetermined DC value corresponding to an average DC level of block data input to the first input terminal is applied, and among the first terminal, second terminal, or third terminal, Selection means having a selection terminal to which a selection signal is applied for selecting one and outputting data inputted to the selection terminal and the output of the first latching means; Itaconic when supplied with the output data of the selecting means, is achieved by the image data DC-DPCM apparatus characterized in that comprising a calculating means for calculating a difference between two data.

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

3 is a conceptual diagram illustrating a method of DPCM DC components of image data in a coding system of image data divided by a multi-window method. In the example of FIG. 3, one frame is composed of m slices (SLC ₁ , SLC _{2 ...} SLC _m ), which is divided into n vertical windows (WD ₁ , WD _{2 ...} WDn). . Therefore, each slice is divided into n blocks by n vertical windows. As described above, when slices divided in each window are referred to as window slices, each window is composed of m window slices WS ₁₁ and WS _{12 ...} WS _1m . The divided image data is processed in units of a macro block group (MMB) consisting of a plurality of window slices. In the embodiment of FIG. 3, one macroblock group (MMB) consists of four window slices, and the first MMB consists of the first, second, third and fourth window slices WS ₁₁ , WS ₁₂ , WS ₁₃ , and WS. ₁₄ ). In the embodiment of the present invention, if the input data is 8 bits, the DC value of this input data is represented by 0 to 256 levels, and the center level '128' is determined as the average DC value. When DC-DPCM is performed on image data having such a data processing unit, the first block data of each MMB is processed in intra mode to calculate a DC value, and the calculated DC value corresponds to the average level of block data. It transmits the difference value after limiting '128'. Subtracting the average level '128' is to reduce the amount of DC data transmitted. Then, as shown in FIG. 3, each block data is DPCM in zigzag order to calculate a DC value of each block. When the DPCM for the first window slice WS ₁₁ is completed, DPCM is performed on each block data of the second window slice WS ₁₂ to calculate a DC value of each block. At this time, the DC value of the first block of the second window slice performs DPCM with the DC value of the first block of the previous window slice. In addition, a forced intra mode encoding process is performed for a macroblock of a predetermined size at a predetermined period. In this case, the first block data of the macro block in which the forced intra mode processing is performed is '128' and DC-DPCM. Will be done. As described above, when image data is encoded in MMB units, each window encoder should transmit a quantization level for each window slice.

4 shows a DC-DPCM unit according to the present invention, and FIG. 5 shows an operation timing diagram for each unit in the apparatus of FIG. The first flip-flop 42 is supplied with a DCT coefficient D _IN as shown in FIG. 5 (b) in accordance with the clock CLK as shown in FIG. 5 (a) as a data input terminal, and DC as a clock terminal as shown in FIG. In order to latch the value, a block DC latch signal LAT _B as shown in FIG. 5 (c) is supplied, and the DC value of each block is latched and output from the DCT coefficient inputted in units of blocks. The second flip-flop 44 is supplied with the DC value output from the first flip-flop 42 as the data input terminal, and the block DC latch signal LAT _B described above is supplied as the clock terminal, and is provided by one block period. Output the delayed block DC value. That is, the block DC value output from the first flip-flop 42 corresponds to the DC value of the current block, and the block DC value output from the second flip-flop 44 corresponds to the DC value of the previous block. The third flip-flop 46 inputs the above-described DCT coefficients in block units to the data input terminal, and the window terminal DC latch signal shown in FIG. 5 (d) to latch the DC value of each window slice into the block terminal. LAT _WS ), latching the DC value of the first block in each window slice. The selector 48 receives the output data of the second flip-flop 44 to the first input terminal IN ₁ , and supplies the output data of the third flip-flop 46 to the second input terminal IN ₂ . '128' is input to the third input terminal IN ₃ . In addition, a predetermined selection signal SEL supplied from a controller (not shown) is applied to the selection terminal SEL of the selection unit 48. The selector 48 selects data input to the first, second or third input terminal in accordance with this selection signal and outputs it through the output terminal. The adder ADD subtracts the output data of the selector 48 from the output data of the first flip-flop 42 and outputs the difference value. When the output data of the second flip-flop 44 input to the first input terminal IN ₁ of the selection unit 48 is selected and output by the selection signal, the adder ADD is determined from the DC value of the current block data. By subtracting and outputting the DC value of the previous block data, the DC value DC-DPCM is obtained between the block data. This time, when the output data of the third flip-flop 46 inputted to the second input terminal IN ₂ of the selection unit 48 is selected and output by the selection signal, the adder ADD is the first in the window slice. It outputs a DC-DPCM DC value for the block. In addition, when an output data of '128' input to the third input terminal IN ₃ of the selection unit 48 is selected and output by the selection signal, the adder ADD is set to '128' from the DC value of the current block data. Subtract 'and print the result. This case corresponds to the DC-DPCM when the power is turned on or the channel is switched and when the periodic intra mode data processing is performed.

As described above, the DC-DPCM system according to the present invention performs DC-DPCM between block data in a slice of each window in an encoding system that divides and processes one frame into a plurality of windows, and performs the first block of each window slice. Performs DPCM and DPCM of the previous slice, and regularly encodes or transmits an MMB composed of a plurality of window slices in an intra mode when periodically turning on the power or switching channels, wherein the first block data of each MMB is transmitted. Encodes a difference from '128' which is an average DC value of the block data. Therefore, the DC-DPCM system according to the present invention can be usefully used for encoding DC component data in a multi-window image data encoding system.

Claims (12)

A method for encoding a DC component of image data at a reduced processing frequency by dividing a frame into a plurality of windows in an image data processing method, the first block data of an MMB consisting of a predetermined number of window slices and an average of other block data. A first step of encoding a difference with a DC value; A second step of performing DC-DPCM between two adjacent block data after the first block data of the MMB; A third step of performing a DPCM and a DPCM of a previous window slice in a first block of the window slice corresponding to the slice divided by the window; And a fourth step of periodically and periodically encoding the MMB in the intra mode. 2. The DC-DPCM method of image data according to claim 1, wherein in the third step, the first block of the window slice performs DC-DPCM with the first block of the previous window slice. The method of claim 1, wherein the fourth step further comprises encoding the MMB in an intra mode even when the power of the image data processing system is turned on. 4. The DC-DPCM method of claim 1, wherein the fourth step further comprises encoding the MMB in an intra mode even when a channel of the image data processing system is converted. An apparatus for encoding a DC component of image data at a reduced processing frequency by dividing a frame into a plurality of windows in an image data processing apparatus, comprising: a first input terminal for inputting the quantized image data in a block unit of a predetermined size; And a second input terminal to which a block synchronous signal having a predetermined pulse is input at a DC component position of each block data in synchronization with the block data, wherein the block data applied to the first input terminal is applied to the second input terminal. First latching means for latching a DC component of the block data by receiving a block synchronization signal; Means for receiving the output data of the first latching means and the block synchronization signal applied to the second input terminal and delaying the output data of the first latching means for one block period; And a third input terminal configured to input a window slice synchronization signal having a predetermined pulse at a DC component position of a predetermined block data of each window slice in synchronization with a window slice corresponding to a slice divided by the window. Second latching means for latching a DC component of a predetermined block data of the window slice by receiving a block data applied to an input terminal and a window slice synchronization signal applied to the third input terminal; A predetermined DC value corresponding to an average DC level of the first terminal to which the output data of the delay means is applied, the second terminal to which the output data of the second latching means is applied, and the block data input to the first input terminal; Selecting means having a third terminal to be applied and a selection terminal to which a selection signal for selecting one of the first terminal, the second terminal or the third terminal and outputting data inputted to the selection terminal is applied; And calculation means for receiving the output data of the first latching means and the output data of the selecting means, and calculating a difference between the two data. 6. The DC-DPCM apparatus of claim 5, wherein the window slice synchronization signal has a predetermined pulse at a DC component position of the first block data. 6. The DC-DPCM apparatus of claim 5, wherein the first latching means comprises a flip-flop. 6. The DC-DPCM apparatus of claim 5, wherein the delay unit comprises a flip flop. 6. The DC-DPCM apparatus of claim 5, wherein the second latching means comprises a flip-flop. 6. The DC-DPCM apparatus of claim 5, wherein the selecting means selects and outputs data input to the first terminal in the case of DC-DPCM between block data. 6. The DC-DPCM apparatus of claim 5, wherein the selecting means selects and outputs data input to the second terminal in the case of DC-DPCM for the window slice. The image data according to claim 5, wherein the selecting means selects and outputs data input to the third terminal when the macroblock having a predetermined size is encoded in the intra mode. DPCM device.