JPH07170407A - Image data compression system - Google Patents

Abstract

PURPOSE:To provide an image data compression system applying high speed expansion processing to compressed data. CONSTITUTION:The system is provided with a discrete cosine transformation (DCT) processing section 5, a quantization processing section 6, an inverse quantization processing section 9, a coding processing section 7, and a correction processing section 13 correcting a signal after DC transformation by a limiter. Then the DCT processing section 5 applies DC transformation to an input picture signal for each block, a quantization processing section 6 quantizes a transformation coefficient being a DC transformation output, an inverse quantization processing section 9 applies inverse quantization to a quantized output signal, the correction processing section 13 corrects a transformation coefficient of a DC transformation output to the inverse quantization output signal so that the level of the output signal does not exceeds a threshold level for the inverse quantization processing section 9, the quantization processing section 6 quantizes the transformation coefficient of the DC transformation and the coding processing section 7 codes the quantization output signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a discrete cosine (DCT) transform coding system which is one of highly efficient compression coding systems, and more particularly, to process coded compressed data at a high speed during decompression. The present invention relates to an image data compression system.

[0002]

2. Description of the Related Art Conventionally, in this type of image data compression system, when a signal is reconstructed by inverse quantization after being quantized once, the reconstructed value contains a quantization error, so that the original input level is There is a possibility that it will be exceeded. Therefore, in order to prevent the restored level from exceeding the threshold value of the input signal during dequantization processing, the signal that exceeds the threshold value is the closest signal within the range that does not exceed the threshold value. Has been replaced with.

A conventional example in which limiter processing is performed before quantization processing during image data compression is disclosed in Japanese Patent Laid-Open No. 188951/1992.
No. In this case, before the coded data of the transform coding is quantized, the coded data is limited so that the coded data exceeding the quantizable level is not quantized. Quantization noise due to overflow is suppressed.

[0004]

In the above-mentioned conventional image data compression system, it is necessary to check the signals by the number of the dequantized coefficients when decompressing, and there is a problem that a lot of processing time is required. It was

It is an object of the present invention to provide an image data compression system capable of decompressing compression data at high speed.

[0006]

SUMMARY OF THE INVENTION The present invention provides an image data compression system for approximating image data by a predetermined transform code to obtain coded data, and quantizing the coded data to obtain compressed data. The method is characterized in that the inverse quantization process is performed once during the quantization process and the inversely quantized output signal is corrected so as not to exceed the threshold value in the inverse quantization process.

The present invention also provides a discrete cosine transform processing section, a quantization processing section, an inverse quantization processing section, an encoding processing section, and a correction processing section for correcting the signal after the discrete cosine transform processing by a limiter. And for each block of the input image signal, performs a discrete cosine transform in the discrete cosine transform processing unit, quantizes the transform coefficient of the discrete cosine transform output in the quantization processing unit, in the inverse quantization processing unit. Dequantize the quantized output signal, correct the transform coefficient of the discrete cosine transform output so that the dequantized output signal in the correction processing unit has a level that does not exceed the threshold value in the dequantized process, and again Quantize the transform coefficient of the discrete cosine transform output in the quantization processing unit,
In addition, the quantized output signal is encoded in the encoding processing unit.

[0008]

According to the present invention, the input image signal is subjected to DCT conversion in the DCT conversion processing unit for each block, the conversion coefficient of the CDT conversion output is quantized in the quantization processing unit, and the inverse quantization processing unit is operated. The quantized output signal is inversely quantized, and the correction processing unit sets the inverse quantized output signal to a DC level that does not exceed the threshold value in the inverse quantization process.
The transform coefficient of the T transform output is corrected, the transform coefficient of the DCT transform output is quantized again in the quantization processing unit, and the quantized output signal is encoded in the encoding processing unit. Therefore, in the inverse quantization process at the time of decoding, It is possible to speed up the processing by not checking the signal.

[0009]

The present invention will be described in detail with reference to the drawings.

After performing DCT transformation on the image data, quantization and coding processing is performed to compress the image data, and the compressed data is subjected to decoding, inverse quantization and inverse DCT transformation, and the compressed data is expanded. In the processing system, when compressing image data, the transform coefficient of the DCT transform output is corrected so that the threshold value is not exceeded during dequantization processing, and the signal is not checked during dequantization processing. Means will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. The central processing unit (CPU) 1 has a control RAM 3
Is used as a work area to execute the program stored in the control ROM 2.

Original image data to be compressed is stored in the file device 4, and the CPU 1 accesses the file device 4 to read the data and perform a compression process, and the compressed image data is written in the file device 4. When displaying the compressed data, the CPU 1 uses the file device 4
Is accessed, the compressed image data is read, decompression processing is performed, and the decompressed image data is written in the frame buffer memory 11 to display an image on the display device 12.

2 and 3 show the flow of compression processing and decompression processing for performing the above-mentioned processing, especially decompression processing at high speed.

The compression process will be described with reference to the flowchart of FIG.

When the compression process is started, the original image file is read from the file device 4 (step S201), and blocks are cut out (step S202). In this example, the original image data is processed in blocks of 8 pixels in the vertical direction and 8 pixels in the horizontal direction. Therefore, it is necessary to cut out data for one block of 8 × 8 pixels. Step S203
Then, the DCT conversion processing unit 5 performs DCT conversion on one block of image data, and decomposes the data from low frequency components to high frequency components. In step S204, the quantization processing unit 6
Thus, the quantization processing for reducing the amount of information is performed from the low frequency component to the high frequency component. Then, in step S205, the inverse quantization unit 9 performs the inverse quantization process, and in step S206, it is checked whether the data after the inverse quantization process exceeds the threshold value.

Usually, when a value is restored by dequantization after being quantized once, the restored value includes a quantization error, so that the original input level may be exceeded. If the level after restoration exceeds the threshold value, the correction processing unit 13 corrects the DCT-converted data (step 207) and limits the value so that it falls within the range of the original input value. Quantization processing is performed again.

The data after the quantization processing is performed in step S20.
The data is encoded by the encoding processing unit 7 at 8, and the compression of the data for one block is completed. When the processing of steps S202 to S208 is completed for each block for the image data size and all blocks are completed, the compressed data is written to the file device 4 in step S210, and the compression processing ends.

FIG. 3 shows data correction means. Figure 3
FIG. 3A is a diagram showing a signal after DCT conversion, and when this signal is quantized, it becomes as shown in FIG. In this way, the data may exceed the threshold value by simply quantizing. Therefore, as shown in FIG. 3 (c), the data is corrected for the signal exceeding the threshold so that the data does not exceed the threshold.

Next, the decompression process will be described with reference to the flowchart of FIG.

When the decompression processing is started, the compressed data is read from the file device 4 (step S301), the decoding processing unit 8 is accessed to decode the compressed data (step S303), and the inverse quantization processing unit 9 performs inverse quantization. Conversion processing is performed (step S304). Since the data after dequantization does not generate data that exceeds the threshold value when compressed,
It is not necessary to check the signal of only the coefficient after the dequantization during the data expansion processing. The data after the inverse quantization is accessed by the inverse DCT transform processing unit 10 in step S305 and the inverse D is obtained.
CT conversion is performed and the image data is written in the frame buffer memory 11 (step S306).

[0021]

As described above, according to the image data compression system of the present invention, it is possible to eliminate the need for checking the limiter at the time of inverse conversion regardless of the format of the input image signal. The compressed data can be decompressed at high speed with a small number of steps.

Further, the present invention is not peculiar to a DCT system, and generally enables high-speed decompression processing of compressed data with respect to general image data compression systems having quantization and inverse quantization processing.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a flow chart showing a compression process in the embodiment of FIG.

FIG. 3 is a diagram illustrating a signal correction process according to the present invention.

FIG. 4 is a flowchart showing a decompression process in the embodiment of FIG.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 CPU 2 Control ROM 3 Control RAM 4 File device 5 DCT conversion processing unit 6 Quantization processing unit 7 Encoding processing unit 8 Decoding processing unit 9 Inverse quantization processing unit 10 Inverse DCT conversion processing unit 11 Frame buffer memory 12 Display device 13 Correction processing unit

Claims (2)

[Claims] 1. An image data compression system for approximating image data by a predetermined transform code to obtain coded data, and quantizing the coded data to obtain compressed data. An image data compression system characterized by performing processing and correcting an inversely quantized output signal so that the inversely quantized processing does not exceed a threshold value. 2. A discrete cosine transform processing section, a quantization processing section, an inverse quantization processing section, an encoding processing section, and a correction processing section for correcting the signal after the discrete cosine conversion processing by a limiter. For each block of the input image signal, the discrete cosine transform processing unit performs discrete cosine transform, the quantization processing unit quantizes the transform coefficient of the discrete cosine transform output, and the inverse quantization processing unit quantizes the output signal. Is inversely quantized, and the transform processing unit corrects the transform coefficient of the discrete cosine transform output so that the inverse quantized output signal has a level not exceeding a threshold value in the inverse quantized processing, and the quantized processing unit again. In the image data, the transform coefficient of the discrete cosine transform output is quantized, and the quantized output signal is coded in the coding processing unit. Compression system.