CN201418137Y - A Lossless Compression Processing System for Spaceborne Images - Google Patents

Abstract

The utility model relates to a lossless compression processing system for satellite images. The processing system includes: a control module, a JPEG-LS prediction module, a RICE coding module and a code stream splicing module. The processing process is: the input image data is grouped by 16 sampling points through the control module; the grouped data enters the preprocessing module Perform decorrelation processing to obtain a set of mutually independent data, and then obtain a compressed code stream through the RICE encoding module and the code stream splicing module. The lossless compression coding efficiency of the utility model is high, and since the core design of the processing system is to perform operations on the original data after grouping, the resynchronization mechanism of decoding can prevent the spread of error codes in the space data transmission, and the power consumption is low, which meets the requirements of spaceborne Image lossless compression requirements.

Description

A Lossless Compression Processing System for Spaceborne Images

technical field

The utility model relates to an image lossless compression processing system, in particular to a space-borne image lossless compression processing system suitable for space applications.

Background technique

In recent years, with the rapid development of my country's aerospace technology, the number and accuracy of on-board sensing and detection equipment have increased greatly compared to before, forming massive data, which has caused certain difficulties for on-board data storage and downlink transmission. Due to the limited communication bandwidth of the spacecraft, it is impossible to increase the capacity of the storage device without limit, so on-orbit image compression has become a necessary part of on-board data processing, and the development of a high-performance on-board image compression system has also become a top priority.

How to optimally implement a high-speed image lossless compression system with strict power consumption, weight, and volume requirements depends on the design scheme. Although the embedded system with general-purpose microprocessor or DSP as the core can complete the compression algorithm relatively conveniently, there is a serious defect - a higher clock frequency (more than 200MHz) is needed to meet the real-time requirements. It will cause a series of electromagnetic compatibility problems and bring a lot of inconvenience in design.

Utility model content

The purpose of this utility model is to provide a lossless compression processing system for spaceborne images in order to overcome the above-mentioned deficiencies in the prior art.

In order to achieve the above purpose, the lossless compression processing system for spaceborne image of the utility model is based on the image compression algorithm recommended by CCSDS121.0-B-1, and adopts the preprocessing module based on JPEG-LS, which effectively improves the compression ratio. The system includes: a preprocessing module based on the JPEG-LS prediction method, a RICE entropy coding module and a code word splicing module.

The preprocessing module receives an n-bit image data DataIn each time through the data line, and the signal NewBlock indicates whether it is the beginning of a new data block Block, and if the NewBlock signal is set, the image data of the data block Block is preprocessed , that is, decorrelation processing, and the preprocessing result DataOut is sent to the RICE entropy coding module through the data line.

The RICE entropy encoding module encodes the preprocessing result DataOut to generate a code word Word with an indeterminate code length Len, and sends the code word Word and the code length Len to the code word splicing module. After the current Block encoding is completed, the EndBlock signal is set .

The code word stitching module receives the code word Word and the code length Len, and outputs the spliced code stream as 8-bit or 16-bit fixed-length code word Byte.

Among them, the input image data is grouped by 16 sampling points through the control module, and the grouped data enters the preprocessor for decorrelation processing to obtain a set of mutually independent data, and then the compressed code stream is obtained by the RICE encoder and the code stream splicing circuit .

Wherein, the preprocessing module includes: a predictor and a mapper.

The predictor subtracts the n-bit input image data from the predicted value DataPre of the corresponding pixel to obtain n+1-bit data DataMap.

The mapper performs mapping transformation on the data DataMap to generate a preprocessing result DataOut, and the preprocessing result DataOut is a prediction residual sequence of an approximate geometric distribution.

Wherein, the RICE entropy encoding module includes: an accumulator, a selector and an encoder.

The accumulator accumulates the preprocessing result DataOut to obtain a sum, and sends the sum to the selector.

The selector performs a table look-up operation on Sum to obtain the coding option K.

The encoder encodes the preprocessing result DataOut according to the encoding option K, outputs the code word Word and the code length Len, and sets EndBlock when the encoding of the data block Block is completed.

The utility model has the advantages of:

1. The on-board image lossless compression processing system conforming to the CCSDS standard of the utility model has high lossless compression encoding efficiency (using the experimental image provided by CCSDS to test, the average lossless compression ratio is 2.0).

2. The core design of the spaceborne image lossless compression processing system of the present utility model is to perform calculations on the original data after grouping, and the resynchronization mechanism of decoding can prevent the spread of errors in space data transmission, and the power consumption is low (≤1watt/Msamples /sec).

3. The spaceborne image lossless compression processing system of the present invention supports frame-based input image formats and strip-based input image formats, and is suitable for near-earth observation and deep space exploration tasks.

Description of drawings

Fig. 1 is the block diagram of the lossless compression algorithm of prior art;

Fig. 2 is the top-level design circuit block diagram of the utility model's spaceborne image lossless compression processing system;

Fig. 3 is the circuit block diagram of the preprocessing module in the lossless compression processing system of the starborne image of the present invention;

Fig. 4 is the circuit block diagram of predictor in the utility model space-borne image lossless compression processing system;

Fig. 5 is a schematic diagram of the pixel positions of the prediction method in the spaceborne image lossless compression processing system of the present invention;

Fig. 6 is a circuit block diagram of the RICE entropy encoding module in the lossless compression processing system for the spaceborne image of the present invention.

Detailed ways

CCSDS announced the lossless compression standard (CCSDS121.0-B-1) suitable for space science data in 1997, and it is recommended to use the RICE algorithm, as shown in Figure 1. The utility model improves the algorithm and realizes the improved algorithm with high-speed hardware.

The image lossless compression processing system provided by the utility model includes: a preprocessing module based on JPEG-LS prediction mode and an adaptive entropy coding module. The preprocessing module includes a predictor and a mapper. The predictor removes the correlation of the data, and then maps them to feature values that are conducive to entropy coding. Adaptive entropy coding for these feature values can get a good compression effect. The entropy coding module is a collection of a series of variable length coders, and a coder with the highest compression ratio is selected to be transmitted together with the identifier. Since each block (J pre-processing samples) can choose a coding mode (basic sequence coding, split sample coding, low entropy coding and uncompressed coding four modes), so the RICE algorithm can adapt to the change of the statistical characteristics of the source.

The utility model adopts a preprocessing module based on JPEG-LS, which effectively improves the compression ratio.

Overall Architecture of Image Lossless Compression Processing System

The algorithm adopted by the lossless compression processing system of the utility model is based on the CCSDS 121.0-B-1 data compression algorithm, and the top-level structure of the compression processing system is shown in Figure 2. It mainly includes: preprocessing module PrePrcessor and entropy coding module Encoder. The preprocessing module receives an n-bit image data each time through the data line DataIn, and the signal NewBlock indicates whether it is the beginning of a new Block. If this signal is set, it starts to preprocess the image data of this Block; the preprocessing result is passed The data line DataOut is sent to the RICE entropy encoder to generate a code word Word with an indefinite code length Len. After the current Block encoding is completed, the EndBlock signal is set; the code word splicing module ByteBuilder receives the code word Word and the code length Len, and the spliced code stream is fixed-length Codeword Byte output.

preprocessing module

The hardware structure design of the preprocessing module, as shown in Figure 3, mainly includes the predictor Predictor (Figure 4) and the mapper Mapper. The present invention adopts JPEG-LS predictor, under the situation of pixel position as shown in Figure 5, JPEG-LS predictor is as follows:

为x₀的预测值。in,

is the predicted value of _x0 .

由多路选择器进行选择，待选值分别为min(x₁，x₃)，max(x₁，x₃)和x₁-x₂+x₃；再通过映射可以得到近似几何分布的预测残差序列。Wherein, the input image data DataIn is n bits, which is subtracted from the predicted value DataPre of the corresponding pixel to obtain n+1 bit data DataMap. Mapper performs mapping transformation on DataMap to generate output data DataOut. forecast result

Selected by the multiplexer, the values to be selected are min(x ₁ , x ₃ ), max(x ₁ , x ₃ ) and x ₁ -x ₂ +x ₃ ; and then the prediction of approximate geometric distribution can be obtained through mapping residual sequence.

Entropy Coding Module

The RICE entropy coding module structure is shown in Figure 6. Among them, the accumulator Accumulator accumulates the output DataOut of the preprocessor PreProcessor, and sends Sum to the selector Selector, and performs a simple table lookup operation in the Selector to obtain the optimal encoding option K; the encoder Encoder according to the optimal encoding Option K encodes DataOut, outputs the codeword Word and the code length Len, and sets EndBlock when the encoding of this data block is completed.

Claims (3)

1. A lossless compression processing system for on-board images, characterized in that the system includes: a preprocessing module for de-correlation processing based on JPEG-LS prediction mode, a code word Word for generating an indeterminate code length Len RICE entropy encoding module and codeword splicing module; The preprocessing module receives an n-bit image data DataIn each time through the data line, and after performing decorrelation processing preprocessing, the preprocessing result DataOut is sent to the RICE entropy encoding module through the data line; Described RICE entropy coding module sends codeword Word and code length Len into codeword splicing module after the above-mentioned preprocessing result DataOut is encoded; The codeword splicing module splices the above-mentioned codeword Word and code length Len to output the codeword Byte with a fixed length. 2. The system for lossless compression processing of spaceborne images according to claim 1, wherein the preprocessing module includes: a predictor and a mapper. 3. The system for lossless compression processing of spaceborne images according to claim 1, wherein the RICE entropy encoding module comprises: an accumulator, a selector and an encoder; The accumulator accumulates the preprocessing result DataOut to obtain a sum, and sends the sum to the selector; The selector performs a table lookup operation on the above sum to obtain the coding option K; The encoder encodes the preprocessing result DataOut according to the encoding option K, and outputs a code word Word and a code length Len.

Images