KR20040073095A - A Device for Both Encoding and Decoding MPEG or JPEG Data - Google Patents

Abstract

PURPOSE: An MPEG-JPEG encoding and decoding device is provided to differentiate a data processing path according to an operational mode, and to perform both encoding process and decoding process within one device, thereby removing a need for equipping individual devices. CONSTITUTION: During an encoding process, data is sequentially moved to an I/DCT block(400), a buffer 1(720), a Q block(320), an I/zigzag block(200), and a VLE block(120), and is encoded. During a decoding process, the data is sequentially moved to a VLD block(110), the I/zigzag block(200), a DQ block(310), the buffer 1(720), an I/DCT block(400), a buffer 2(730), and an MC(Motion Compensation) block(500), and is decoded. A multiplexer(600) selects a block which exchanges the data with the buffer 1(720) during the decoding/encoding processes.

Description

MBP / GPP encoding and decoding device {A Device for Both Encoding and Decoding MPEG or JPEG Data}

The present invention relates to an apparatus for decoding and encoding MPEG / JPEG, and more particularly, to an apparatus capable of simultaneously decoding and encoding MPEG or JPEG data.

MPEG stands for Motion Picture Experts Group, and JPEG stands for Motion Picture Experts Group. JPEG stands for Joint Photographic Coding Experts Group. MPEG decoding may be performed by adding a block that performs a simple Motion Compensation (MC) function to hardware for conventional JPEG decoding. The C function refers to a function of compensating for the movement of an object between frames in a video.

When I-picture MPEG encoding is performed, encoding may be performed in a manner almost similar to that of JPEG encoding. I-picture stands for Intra coded Picture and means that one picture is itself a complete picture. In other words, each frame of the video is a complete image. Therefore, if each frame is encoded by JPEG encoding method, the video can be encoded.

FIG. 1 is a block diagram showing the structure of a conventional MPEG / JPEG encoder. The VLE block 1, the ZIGZAG block 2, the Q block 3, the DCT block 4, the buffer 0 (5), and the buffer 2 ( 6).

The DCT block 4 performs a discrete cosine transform on the input image (frame) in blocks of a predetermined size, for example, 8 × 8. The size of the block can be chosen differently depending on the system. The DCT block 4 stores the conversion result in the buffer 2 (6). DCT is one of the orthogonal transform coding schemes, and was adopted by ITU-T in 1988 as the video compression technique of H.261 coding scheme for teleconference telephony. Similar to the fast Fourier transform (FFT), the image signal of the time axis is transformed to the frequency axis, which is called a discrete cosine transform because a discrete cosine function is used as a coefficient for the conversion. When the DCT passes through, the image signal in the spatial domain is converted into a signal in the frequency domain. The DCT block 4 instructs the Q block 3 to perform quantization when the conversion result of one block is stored in the buffer 2 (6).

The Q block 3 performs quantization on the result of the DCT operation. Quantization is a nonlinear operation that approximates an input signal to a finite number of values. The quantization is performed to lower the entropy and adjust the compression rate for the data on which the DCT operation is performed.

The ZIGZAG block 2 stores the data block received from the Q block 3 in the buffer 0 (5), which is an internal buffer, and when the input of one data block is completed, the data block is zigzag-scanned and the result is next. Deliver to stage. Zigzag scanning refers to a method in which scanning is performed by moving from a pixel located at one vertex of a block to a pixel located at a vertex in a diagonal direction while scanning all pixels existing in a line perpendicular to the diagonal line in order. do. The scanning result is sent to the VLE block 1.

The VLE block 1 performs run length encoding on the zigzag scanned block and other information, and performs Huffman encoding. Run length encoding is a compression technique that performs compression using overlapping data and the number of overlaps. Huffman coding is one of variable length coding methods in which a code having a different bit length is used for each value according to a probability of occurrence of a signal value.

FIG. 2 is a block diagram showing the structure of an MPEG / JPEG decoder. The VLD block 10, the IZIGZAG block 20, the DQ block 30, the IDCT block 40, the MC block 50, and the buffer 0 60 are shown in FIG. Buffer 1 (70), and buffer 2 (80).

The VLD block 10 performs Huffman decoding and run length decoding to restore pixels and other information constituting the data block. Pixel information constituting the data block is transmitted to the IZIGZAG block 20. In the IZIGZAG block 20, the pixel transferred from the VLD block 10 is stored in the buffer 0 60, and when one data block is completed, the IZIGZAG block 20 performs zigzag scanning and transfers it to the DQ block 30. In the DQ block 30, quantization is reversely performed on the transmitted data, and the result is stored in the buffer 1 70 in units of blocks. When the DQ block 30 stores the result in the buffer 1 70, the IDCT block 40 requests the inverse discrete cosine transform to be performed. The IDCT block 40 performs inverse discrete cosine transform on the block received through the buffer 1 70 and stores the result in the buffer 2 80. When one block is stored in the buffer 2 80, the IDCT block 40 requests the motion block from the MC block 50. When operation compensation is requested, the MC block 50 performs operation compensation on the block stored in the buffer 2 (80). Motion compensation is used only when decoding MPEG data.

Conventionally, according to the MPEG / JPEG standard, there are devices that perform the above encoding or decoding separately, but there is no device that performs both at the same time, so there is a problem that a separate device must be provided.

An object of the present invention is to provide an apparatus capable of simultaneously performing decoding as well as encoding for MPEG / JPEG.

1 is a block diagram of a conventional MPEG / JPEG encoder.

2 is a block diagram of a conventional MPEG / JPEG decoder.

3 is a block diagram of an apparatus for MPEG / JPEG encoding and decoding according to the present invention.

The MPEG / JPEG encoding and decoding apparatus according to the present invention includes a first block including a variable length encoding unit and a variable length decoding unit, a second block performing zigzag scanning and reverse zigzag scanning, a third block including a quantization unit and an inverse quantization unit. A block, a fourth block for performing DCT transform and an inverse DCT transform, and a fifth block for performing motion compensation.

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

3 is a block diagram showing the configuration of an MPEG / JPEG encoding and decoding apparatus according to the present invention. The VLD block 110, the VLE block 120, the I / ZIGZAG block 200, the DQ block 310, and the Q block ( 320), I / DCT block 400, MC block 500, buffer 0 710, buffer 1 720, buffer 2 730, and MUX 600.

The ZIGZAG block 2 disclosed in FIG. 1 and the IZIGZAG block 20 disclosed in FIG. 2 may be implemented in one block by performing similar functions with only differences in coefficients. The I / ZIGZAG block 200 is one in which they are implemented. Likewise, the DCT block 4 disclosed in FIG. 1 and the IDCT block 40 disclosed in FIG. 2 may also be implemented as a single block as they perform similar functions with differences in coefficients used. The I / DCT block 400 is one of them implemented.

During encoding, encoding is performed while data is sequentially moved to the I / DCT block 400, the buffer 1 720, the Q block 320, the I / ZIGZAG block 200, and the VLE block 120. Refer to the description of FIG. 1 for the function of each block.

The data at the time of decoding is the VLD block 110, the I / ZIGZAG block 200, the DQ block 310, the buffer 1 720, the I / DCT block 400, the buffer 2 730, and the MC block 500. Decoding is performed by moving the sequence in order. Refer to the description of FIG. 2 for the function of each block.

The MUX 600 selects a block for exchanging data with the buffer 1 720 among the DQ block 310 and the Q block 320 during decoding and encoding. Q block 320 is selected for encoding and DQ block 310 is selected for decoding.

By implementing the MPEG / JPEG encoding device and the decoding device in one device as described above, there is no need to have a device that performs each function separately.

Claims (3)

A first block including a variable length encoding unit and a variable length decoding unit; A second block for performing zigzag scanning and reverse zigzag scanning; A third block including a quantization unit and an inverse quantization unit; A fourth block for performing DCT transform and inverse DCT transform; And Fifth block for performing motion compensation Including, In encoding mode, input data is processed sequentially through the fourth block performing DCT conversion, the quantization unit, the second block performing zigzag scanning, and the variable length encoding unit, and in decoding mode, input data is the variable. A length decoding unit, a second block performing inverse zigzag scanning, the inverse quantization unit, and a fourth block performing inverse DCT conversion are sequentially processed, and in the decoding mode for data in MPEG format, the fourth block is processed. MPEG / JPEG encoding and decoding apparatus, characterized in that additionally processing the data output from the block in the fifth block. The method of claim 1, wherein the fourth block is A buffer for storing the converted data in a predetermined unit; And A multiplexer coupled between the buffer and the third block Including In the encoding mode, the fourth block stores the DCT-converted data in the predetermined unit in the buffer, and the quantization unit is configured to provide data from the buffer via the multiplexer in response to a control signal provided in the fourth block. Perform a quantization, and in the decoding mode, the inverse quantizer transfers data to the buffer in the predetermined unit via the multiplexer, and the fourth block is provided to the buffer in response to a control signal transmitted from the inverse quantizer. MPEG / JPEG encoding and decoding apparatus characterized by performing inverse DCT transformation on the stored data. The method of claim 1, wherein the fifth block is And a buffer for storing the data provided in the fourth block in a predetermined unit, and processing the data stored in the buffer in response to the control signal provided in the fourth block.