Research On Parallel Processing Algorithm And System Of Remote Sensing Satellite Image Decompression Based On JPEG-LS

Remote sensing satellite image data decompression system is an important part of satellite ground pre-processing system,which is mainly responsible for the decompression task of the image data,which is the data base of the subsequent ground processing steps.Nowadays,with the development of remote sensing satellite technology,the number of satellite launch is increasing year by day,the processing of remote sensing satellite image data has presented characteristics such as large amounts of data and high real-time requirements,which puts forward higher requirements for processing speed and expansibility of remote sensing satellite image data decompression system.In this paper,we combine the application scene of the research group on the decompression processing system,based on the analysis of the development of remote sensing satellite image data decompression system,the parallel processing of satellite image data decompression based on Jpeg-ls compression algorithm is studied,and how to realize the multi-star scalability of the system is studied.Firstly,this paper analyzes the present situation of the remote sensing satellite image data decompression system,and introduces the related technologies used in the decompression system in this paper.Then we analyze the key technical points in the compression system and the corresponding solutions.In order to realize the efficient combination of the system flow,a system structure based on pipelining is proposed.A computational load balance and a two-level parallel decompression algorithm based on SMP cluster is proposed to realize the fast processing of the system.In order to achieve the simple and scalable system,this paper proposes a method to solve the scalability of the system with the combination of external configuration file,general data structure and software design pattern.Then,the design and implementation of the decompression system are expounded,the requirement index is determined by the requirement analysis,the system design is completed,and the key algorithm modules in the system are designed and implemented in detail.Then we used the resource No.01 data based on the JPEG-LS compression algorithm to perform experiments on the processing performance,parallel acceleration ratio,and data correctness.The results show that the real-time and accurate processing of resource third 01 star dual-channel data is realized on five blades with 2 Intel Xeon E5620.At the same time,the system scalability was experimented with the multispectral data of GF02,and the data was correctly processed by configuration file configuration and small amount of code,without changing the original code.From the experimental results,it can be seen that the design and implementation of the remote sensing satellite image data parallel decompression system based on JPEG-LS compression algorithm has good processing performance and scalability.It is a positive exploration to the engineering practice of remote sensing satellite image Data decompression system,which has good practical reference value.