JPH0437367A - Data compression control method - Google Patents

Abstract

PURPOSE:To devise the method such that run length encoding and Huffman encoding are easily applied to a data after DCT operation in the case of applying DCT operation to the picture data for compression by storing the data being quantized picture data, dividing the data into bit planes from the MSB and the LSB bits and applying run length encoding and Huffman encoding to the data. CONSTITUTION:An input data is quantized by a Discrete Cosine Transform(DCT) computing element 3 for each block, the quantized data is stored in a conversion data memory 4, the stored data is divided into bit planes, data '1', '0' for each plane are encoded by a run length encoder 7, the data subject to run length encoding is encoded by a Huffman encoder 8 to compress the input data. The run length encoding and Huffman encoding for compression are easily implemented by adopting the method employing a binarization data resulting from DCT operation to the picture data.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is used for an image device that handles natural images, and is used to process image data (image information), etc. using CT (Discrete).
Co51ne Transfer@) and calculate this D
CT-calculated data is quantized and run length
The present invention relates to a data compression control method for encoding the encoded data (Length) and encoding the encoded data.

[Conventional example] Conventionally, in this type of data compression control method, for example, as shown by the arrow in FIG. , further move it upward on the drawing and accumulate it at the upper limit of the predetermined block 1. After this, the DC of the predetermined block 1
The T-operated data is quantized using a predetermined threshold value and run-length encoded. Thereafter, the run-length encoded data is Huffman encoded to compress the image data.

By compressing image data in this way, it is possible to speed up the transmission time of image signals, and this is used for natural image transmission and the like.

[Problems to be Solved by the Invention] By the way, in the data compression control method described above, the data value resulting from the DCT operation is not binary, and -way run-length encoding and Huffman encoding are suitable for binary data. , the data subjected to the DCT operation is not suitable for run-length encoding or Huffman encoding.

This invention was made in view of the above problems, and its purpose is to reduce the DC
An object of the present invention is to provide a data compression control method that makes it easy to perform run-length encoding and Huffman encoding of T operations.

[Means for Solving the Problems] In order to achieve the above object, the present invention converts image data into
In a data compression control method that performs CT calculation, quantization, and run-length encoding and Huffman encoding of the quantized data, the quantized data is stored, and the stored data is converted into MSB (Most Significant
Bit) to LSB (Least 51gn1fica
The gist is that the image data is divided into planes of each bit up to ntBit), the data of each plane (1"JjQ") is run-length encoded and Huffman encoded, and the image data is compressed.

[Operation] Since the above method is adopted, for example, the data stored in the original image memory is subjected to DCT calculation and quantized for each block, and the quantized data is stored in the transformed data memory. The data in this conversion data memory is scanned, and the data read out by this scanning is selected into each plane of each bit from MSB to LSB. In this case, for example, as shown in FIG.
Assuming a DCT matrix of X8 pixels, X Hyg
Scanning is performed for each block of Z (10×8×9), and the bit plane of the Z-axis method is selected. Therefore, nine planes are selected in that block, and the data of the selected planes is run-length encoded and Huffman encoded.

In this way, since the data subjected to the DCT operation is converted into data of each plane of bits ("1", 'O'), subsequent encoding by run-length encoding and Huffman encoding becomes easier.

[Example] Hereinafter, an example of the present invention will be described based on FIGS. 1 and 2.

In the data compression control method of the present invention, input data is processed by DCT (Discrete Copy) for each block.
-5ine Transform) operation, quantization, storing this quantized data, dividing this stored data into planes of bits, and converting the data for each plane (“1
”“O”) is run-length encoded and Huffman encoded,
Compressing input data.

For this purpose, for example, as shown in FIG. a conversion data memory 4 for scanning, a scanning section 5 for scanning and reading out the data in the conversion data memory 4, and a scanning unit 5 for scanning and reading out the data in the conversion data memory 4;
A selector unit 6 that divides and selects each bit into planes up to SB, and a run-length encoder 7 that performs run-length encoding on the data (“1” JIQ?+) for each plane of each selected bit. A Huffman encoder 8 that Huffman encodes this run-length encoded data.
and a bit select control section 9 that controls scanning of the data and plane division of each bit. The conversion data memory 4 stores, for example, quantized image data as shown in FIG.
The data for each plane is selected by the selector unit 6.

Further, the converted data memory 4 stores the blocks shown in the figure in a number corresponding to the image data stored in the original image memory 2.

Next, a data compression control method applied to the data compressor having the above configuration will be explained in detail with reference to FIGS. 2 and 3.

First, it is assumed that a video signal is digitally converted and stored in the original image memory 2, and the data in the original image memory 2 is read out. Then, the read data is subjected to DCT calculation and quantization in the DCT calculation section 3 and is stored in the converted data memory 4. Next, scan section 5
The data in the conversion data memory 4 is scanned at
The data stored in the conversion data memory 4 is quantized "1", 'O', and as shown in FIG. 2, for example, x * y?
) is scanned in cubic units (block units). Further, the scanned data is selected by the selector unit 6 for each plane of bits in two axis directions.

For example, in the case of the block shown in FIG. 2, if the first bit is also specified in the Z-axis direction, the selector unit 6 selects the data of the plane of the first bit as shown in FIG. is selected.

Furthermore, since there are 9 bits in the z-axis direction, the selector section 6 selects data of nine planes.

Subsequently, the data of each selected plane is run-length encoded by a run-length encoder 7, and this run-length encoded data is encoded by a Huffman encoder 8.
Since the image data is Huffman encoded in , compressed data of the image data in the optical image memory 2 can be obtained.

In this way, by reading the DCT-calculated data value in bit plane units, the data value can be converted into a binary value (“1”
Since the data can be set to "0"), run-length encoding and Huffman encoding can be easily performed.

[Effects of the Invention] As explained above, according to the data compression control method of the present invention, image data is subjected to a DCT operation and quantized, and the quantized data is run-length encoded and Huffman encoded. , the above quantized data is divided into planes of each bit from MSB to LSB, and the data of each plane C4I IT , u OT
l) is run-length encoded, the run-length encoded data is Huffman encoded, and the above image data is compressed, so that the data values subjected to the above DCT operation can be obtained as binary data. This has the effect of making length encoding and Huffman encoding easier.

[Brief explanation of drawings]

, FIG. 1 shows an embodiment of the present invention, a circuit block diagram to which the data compression control method is applied, FIGS. 2 and 3 are schematic diagrams for explaining the data compression control method, and FIG. FIG. 2 is a schematic diagram illustrating DCT calculation of data. In the figure, 1 is a block, 2 is an optical image memory, 3 is a DCT calculator, 4 is a conversion data memory, 5 is a scan section, 6 is a selector section, 7 is a run length encoder, 8 is a Huffman encoder, 9 is a bit selection control section. Figure 4 Patent applicant: Fujitsu General Co., Ltd. Agent Patent attorney: Takuya Ohara

Claims (1)

[Claims] (1) In a data compression control method in which image data is subjected to DCT calculation and quantized, and the quantized data is run-length encoded and Huffman encoded, the quantized data is stored, and the stored data is converted from MSB to LSB. The data of each plane (“1”,
1. A data compression control method, characterized in that the image data is compressed by run-length encoding and Huffman encoding.