KR0138368B1 - Encoder/decoder for digital writing - Google Patents

Abstract

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for storing digital video data in a digital manner, and in particular, to converting compressed Megg (Moving Picture Expert Group, hereinafter referred to as MPEG) digital data into a data stream for recording on a video tape, The conversion coder / decoder converts the data reproduced from the data into displayable data. The conversion coder includes an ECC remover, a first memory, a demultiplexer, a variable decoder, a second memory, an interpolator, a suppler, A multiplexer, channel ECC inserter is provided, and the conversion decoder includes a channel ECC remover, a demultiplexer, a decoder, a deshuffler, and a trick decoder.

Description

Conversion coder / decoder with supposing and interpolation for digital recording

1 is a structural diagram showing recording of a bitstream compressed by the MPEG recommendation to tape without using a conversion coder / decoder,

2 is a schematic diagram of a VCR shown for explaining the conversion coder / decoder according to the present invention.

3A-3C are schematic diagrams showing standard pictures recommended by CCIR 601,

4 is a conceptual diagram showing I, B, and P images by MPEG,

FIG. 5 is a schematic diagram showing slices constituting the normalized image of FIG. 3a,

FIG. 6 is a schematic diagram showing a macroblock constituting the slice of FIG.

FIG. 7 is a schematic diagram showing blocks constituting the macroblock of FIG.

8 is a block diagram showing a conversion coder according to the present invention;

9 is a block diagram showing a conversion decoder according to the present invention;

10 is a conceptual diagram showing interpolation according to the present invention,

11 is a schematic diagram showing that the interpolated image is divided into macroblocks and numbered sequentially.

FIG. 12 is a schematic diagram showing the interpolation of FIG. 11 interpolated with the present invention.

13 is a schematic diagram showing B pictures and P pictures interpolated by the present invention,

Fig. 14 is a structural diagram showing a tape recorded in a digital manner by the present invention.

The present invention relates to an apparatus for storing digital video data in a digital manner, and in particular, to convert a compressed moving picture expert group (hereinafter referred to as MPEG) digital data into a recording data stream for recording on a video tape, A conversion coder / decoder for inputting data reproduced from a video tape and converting the data into displayable data.

Digital processing of video signals is required, from video teleconferencing and video telephony to high-definition television (HDTV). Since such digital video data is vast data of several tens of megabits, compression is essential for real time processing. The international standardization of image information compression and reconstruction is urgently required, and a group of experts such as JPEG (Ig- ing Photographics Expert Group), MPEG-I, MPEG-II, etc. have been issued. Among these, MPEG's recommendation for moving images uses interframe shift compensation interpolation to remove data redundancy in the time domain and discrete cosine transform (DCT) to remove redundancy in the spatial domain. . The DCT transformed data forms a compressed video signal through quantization and variable length coding (VLC). In other words, an MPEG-compressed image is composed of an intra frame (hereinafter referred to as an I picture), a predicted picture (hereinafter referred to as a P picture), and a bidirectional predicted picture.

The I picture is a pointer for random access and a reference picture that minimizes the accumulation of errors, so the data is relatively large, and the P picture is a picture predicted based on the past picture (I, P). Since the B picture is a predicted picture by the motion compensation interpolation technique, and contains only the relative movement information of the reference picture, the data amount is relatively small to provide a high compression ratio. These P pictures and I pictures are variable in size as Inter frame data.

FIG. 1 shows a case where a bitstream compressed by the MPEG recommendation is recorded on a tape without using a conversion coder / decoder. 'I' represents an I image as described above, and 'B' represents B. An image is shown, 'P' represents a P image, and an arrow represents a running direction of the tape.

In Fig. 1, the MPEG-compressed data is inputted as '... I, B, B, P, B, B .....' so that the I picture is stored in four tracks, and the B picture is It is stored in two tracks, and the P picture is stored in three tracks. In other words, as described above, since the data of the I picture is large and the data of the B picture is small, the tracks occupied by one image (or frame, also called image) may be different when the data input by the MPEG method is stored on the tape as it is. do. Therefore, it is not possible to digitally record a bit stream including intra frame and interframe encoding with a tape and video head running at a constant speed. In other words, playback is possible only when one frame occupies a certain track. Intra-frame and inter-frame have different lengths of bit streams, and thus the recorded tracks are different. Therefore, there is a problem that it is difficult to record the data extruded by the MPEG method on the tape as it is.

Accordingly, an object of the present invention is to provide a transform coder using shuffling and interpolation to occupy a constant track per frame in order to solve the above problems.

Another object of the present invention is to provide a conversion decoder for reproducing and decoding the recorded data coded with a conversion coder using shuffling and interpolation to display the data in order to solve the above problems.

In order to achieve the above object, the conversion code of the present invention is a digital VCR for inputting and recording / reproducing video data compressed by MPEG method having an I picture, a B picture, and a P picture,

An error correction code eliminator for removing the error correction code by inputting the image data; A demultiplexer which decodes the image data from which the error correction code is removed and outputs only pure video data; A first memory for storing image data from which the error correction code has been removed; A variable length decoder for variable length decoding by inputting only the pure video data; A second memory for inputting the output of the variable length decoder to store only the I image data; An interpolator for inputting the next I picture data from the variable length decoder and interposing and inputting the previous I picture data stored in the second memory to generate new I picture data; A shuffler to shuffle the output of the interpolator; A multiplexer for adding the output of the shuffler to image data stored in the first memory; And a channel error correction code inserter for inputting an output of the multiplexer to add a channel error correction code for error prevention during recording / playback.

The data can be recorded by interpolating the data of the I picture, the B picture, and the P picture so as to occupy the same number of tracks.

In order to achieve the above object, the transform decoder of the present invention is a digital VCR for inputting image data compressed by MPEG method having I picture, B picture, and P picture to record / reproduce the detier coded by the conversion coder.

A channel error correction code eliminator for inputting reproduced video data to remove the channel error correction code; A demultiplexer inputting an output of the channel error correction code remover and outputting the output signal according to a mode signal; A decoder configured to input, decode, and output an output of the demultiplexer in a normal mode; A de-shuffler that restores the data shuffled from the conversion coder by inputting the output of the demultiplexer in a high speed mode; A trick decoder which decodes and outputs high-speed data by inputting the output of the deshuffler; And a multiplexer for inputting an output of the decoder and an output of the trick decoder to output one according to a mode signal.

And decoding the data coded by the transform coder.

Next, the apparatus of the present invention will be described in detail with reference to the accompanying drawings.

2 is a schematic diagram of a VCR shown for explaining the conversion coder / decoder according to the present invention. In Fig. 2, the terminal 1 receives digital video data from a video camera and is coded and compressed by the MPEG encoder / decoder 3. The data compressed by the MPEG method is inputted to the change coder / decoder 5 into I, B, and P images, and is a recordable data stream, that is, a recording data stream in which interframe data is extended to occupy a predetermined track per frame. Is output and recorded on a tape (not shown) in the recording and reproducing system 6. The signal recorded on the tape (not shown) is reproduced in the recording and reproducing system 6 again, decoded by the conversion coder / decoder 5 and displayed on the monitor (not shown) through the terminal 2. The digital video cassette recorder (DVCR) 4 here comprises a conversion coder / decoder 5 and a recording and reproducing system 6.

3a to 3c are schematic diagrams showing the sizes of standard images recommended by CCIRR 601, and FIG. 3a shows 720 x 480 pixels as the luminance (Y) signal, and FIG. 3b shows 360 x 480 pixels as the color difference (Cr) signal. 3C shows 360 x 480 pixels as a color difference (Cb) signal.

4 is a conceptual diagram showing a B picture, an I picture, and a P picture by MPEG. The I, B, B, P, B, B, I, B, B, P, B, B, .. A part of successive images such as ... is listed. At this time, from the I picture to the next I picture is called a Group Of Picture (hereinafter referred to as a GOP). Six pictures, such as I, B, B, P, B, and B, form one GOP. In general, a GOP formed of N pictures has one intra frame and N-1 interframes, and implements a moving picture repeated in units of N frames. In reality, the GOP is composed of 15 pictures, such as B, B, I, B, B, P, B, B, P, B, B, P, P, B, B, and I. Here, as the number of GOP constituent pictures increases, the number of B pictures increases and the compression ratio of data increases, but the correlation between data decreases. Therefore, implement appropriate GOP according to the purpose of the system.

5 shows a slice (SLICE) constituting an image as shown in FIG. 3A according to the present invention, where SL represents a slice and one image is composed of 30 slices.

FIG. 6 illustrates a macroblock constituting the slice of FIG. 5, where MB represents a macroblock. In FIG. 6, one slice constitutes 45 macroblocks, and one macroblock has a size of 16 × 16.

FIG. 7 shows a block constituting the macroblock of FIG. 6, one macroblock is composed of four blocks, and one block has a size of 8x8. Thus, there are 1350 macroblocks in one picture.

8 is a block diagram showing a conversion coder according to the present invention, including an ECC remover 12, a demultiplexer (DEMUX) 14, a first memory (MEM1) 16, and a variable length decoder (VLD) ( 18), a second memory (MEM2) interpolator (22), a shuffler (24), a multiplexer (MUX) 26, and a channel ECC inserter (28).

In FIG. 8, the input data compressed by the MPEG method through the input terminal 10 is input to the ECC remover 12. The ECC eliminator 12 removes the added error correction code (ECC) from the input data to prevent the error on the transmission path from the video data transmitting side, and outputs the demultiplexer 14 and the first memory 16 to the first data. . The demultiplexer 14 extracts only pure image data from the data transmitted by the MPEG method and outputs it to the variable length decoder 18, and the variable length decoder 18 decodes the variable length coded data on the MPEG transmitter side. The original data is restored to store only the I image data in the second memory 20. When the next I image data is input from the VLD 18, the interpolator 22 takes the previous I image data stored in the second memory 20, performs interpolation, and stores the interpolated image 22 in the interpolator 22. The I-picture data is stored in the second memory 20 again, and then new interpolated I-picture data (hereinafter referred to as Ii-pictures) are generated and output to the suppler 24. The suppler 24 randomly extracts and relocates the input interpolated I image data (Ii image data) macroblocks (MB) to prevent the image from deteriorating due to an error in the recording / reproducing system. . In other words, shuffling means that image data in a frame is randomly rearranged at the time of recording on tape to prevent deterioration of an image due to unread data in order to prevent intensive degradation of a specific portion of the image due to data that cannot be read at high speed. Dispersion is a technique that makes it difficult to discern with the naked eye. The multiplexer 26 inputs shuffled interpolation I-picture (Ii-picture) data to the input data stored in the first memory 16 to make the size of the intraframe data and the interframe data the same per frame. The number of tracks to be allocated is made constant. The channel ECC inserter 28 adds a channel error correction code (ECC) to the output of the multiplexer 26 to correct an error occurring during recording / reproduction and outputs it to the recording system through the output terminal 30. .

9 is a block diagram showing a conversion decoder according to the present invention, wherein the conversion decoder 40 includes a channel ECC remover 44, a demultiplexer 48, a decoder 50, a decoupler 52, and a trick decoder. 54, a multiplexer 55 is provided and the video data reproduced from a tape (not shown) in the reproduction system 42 is input and processed and output to the display device 56.

In Fig. 9, the playback system 42 reproduces the track data from the tape on which data coded by the conversion coder as shown in Fig. 8 is recorded and outputs it to the channel ECC remover 44. Figs. The channel ECC remover 44 removes the channel ECC added by the channel ECC inserter 28 of FIG. 8 from the reproduced video data and outputs it to the demultiplexer 48. The demultiplexer 48 outputs the reproduced video data to the decoder 50 when the operation mode signal MS of the VCR input to the terminal 46 is in the normal mode, and high speed (also called a trick mode). If the mode is output, it is output to the decoupler 52. Therefore, in the normal mode, the decoder 50 discards the interpolation intraframe data (Ii image data) shuffled by the conversion coder of FIG. 8 of the reproduced video data, and removes the original I image data, B image data, and P image data. It decodes the video data into displayable video data and outputs the same to the display device 56 through the multiplexer 55. On the other hand, in the high speed mode, the data shuffled by the shuffler 24 of FIG. 8 of the reproduced video data is restored to the original position and output to the trick decoder 54, and the trick decoder 54 is connected with the reproduced data. The data restored to the original position by the desuppler 52 is decoded and output to the display device 56 through the multiplexer 55. The multiplexer 55 connects the output of the decoder 50 to the display device 56 in the long mode according to the mode signal MS input to the terminal 46, and displays the output of the trick decoder 54 in the high speed mode. Connect to device 56.

The display device 56 expresses the input data as an image so that it can be viewed.

FIG. 10 is a conceptual diagram illustrating interpolation according to the present invention, and extracts only I image data from an input image data stream to generate a new interpolation I image (Ii image) between adjacent I images. In FIG. 10, the interpolated I picture Ii1 is obtained from an I1 picture and an I2 picture, and the interpolated I picture Ii2 is obtained from an I2 picture and an I3 picture. In this case, the interoperation takes a linear intermediate value after summing the DCT coefficients obtained in each macroblock unit.

FIG. 11 shows a part of sequentially numbered interpolated I pictures divided into macroblocks, where MB represents a macroblock. 11, numbers are sequentially assigned to corresponding positions from the upper left to form one image.

FIG. 12 shows a part of the shuffled Ii image of FIG. 11 according to the present invention, and it can be seen that each macro block is randomly located. That is, MB65 is located at the MB1 position in FIG. 11 and MB48 is located at the MB2 position. In this way, since the image of the randomly located macroblocks cannot be seen in mosaic form if they are played back as they are, they must be deshuffled and restored to their original positions as shown in FIG.

Fig. 13 shows the data structures of the B picture and the P picture according to the present invention, where a represents a header, b represents trick data added by the conversion coder, and c represents original slice data. The header consists of an area for SYNC for synchronization of the bitstream, FF / FR for indicating a high speed mode, and other control information. Shuffled data is interpolated and shuffled from an I picture to add a B picture and a P picture to make tracks allocated per frame constant. It is ignored during playback in normal mode, and is ignored at high speed (trick mode or high speed mode). Used to compensate for deterioration of an image. On the other hand, such image data adds an internal code ECC and an external code ECC in order to correct errors that may occur during recording and playback on a tape.

Fig. 14 shows the structure of a tape recorded in the digital manner by the present invention, and it can be seen that both I picture, B picture and P picture are stored in four constant tracks. That is, as compared with FIG. 1, two tracks are added to the B picture, and one track is added to the P picture. In the added track d, interpolation intra frame data shuffled by the present invention is recorded.

As described above, according to the present invention, interframe and interframe data of MPEG system having different data sizes are interpolated to occupy a predetermined track so that recording is possible on a tape. In addition, the newly interpolated data recorded by shuffling is used in the high speed mode at the time of reproduction, thereby preventing the deterioration of the image.

Claims (2)

In a digital VCR that inputs and records / reproduces video data compressed by an MPEG method having an I picture, a B picture, and a P picture, An error correction code eliminator for removing the error correction code by inputting the image data; A demultiplexer which decodes the image data from which the error correction code is removed and outputs only pure video data; A first memory for storing image data from which the error correction code has been removed; A variable length decoder for variable length decoding by inputting only the pure video data; A second memory for inputting the output of the variable length decoder to store only the I image data; An interpolator for inputting the next I-picture data from the variable length decoder and interpolating the previous I-picture data stored in the second memory to generate new I-picture data; A shuffler to shuffle the output of the interpolator; A multiplexer for adding the output of the shuffler to image data stored in the first memory; And A channel error correction code inserter for inputting an output of the multiplexer to add a channel error correction code for error prevention during recording / playback; And converting the data of the I picture, the B picture, and the P picture to a similar length so that recording is possible by interpolating to occupy the same number of tracks. In a digital VCR that inputs video data compressed by MPEG method having I picture, B picture and P picture, and records / reproduces data coded by a conversion coder. A channel error correction code eliminator for inputting reproduced video data to remove the channel error correction code; A demultiplexer for inputting an output of the channel error correction code remover and outputting the output signal according to a mode signal; A decoder configured to input, decode, and output an output of the demultiplexer in a normal mode; A de-shuffler for restoring data shuffled in the conversion coder by inputting the output of the demultiplexer in a high speed mode; A trick decoder that decodes and outputs high-speed data by inputting the output of the deshuffler; And And a multiplexer for inputting the output of the decoder and the output of the trick decoder to output one according to the mode signal. The transform decoder characterized in that for decoding the data coded by the transform coder.