On-orbit Intelligent Compression Method For Preserving Weak And Small Target Information And Its FPGA Implementation

With the continuous development of satellite remote sensing technology in our country,the downloading of remote sensing images with huge amounts of data has become a difficult problem to be solved.The traditional solution is to use image compression technology to process the image,however,when the satellite-to-ground data transmission bandwidth has not been significantly improved,and the lossless image compression algorithm that can be applied in orbit has no breakthrough in the coding principle,satisfying the throughput rate of big data and the possible loss of small infrared targets has become a contradiction.In order to solve this contradiction,this article proposes an intelligent compression method for preserving weak and small target information under the "FPGA+ASIC" architecture.The on-orbit image compression algorithm is integrated in the ASIC,and the hardware implementation of the infrared small target detection algorithm is optimized in the FPGA,and the adaptive threshold adjustment is designed according to the feedback of the image compression.On-board test results show that the method proposed in this paper can achieve a better balance between image compression ratio and target detection rate,and meet the task requirements of the source of the subject.In the final test,the detection rate is not less than 95% for infrared images with a local signal-to-noise ratio of 3-10.For a largeformat single-frame image of 8192x8192,the theoretical processing time is only 1.25 s.In the process of research,the FPGA design in this article uses multiple clock domains to perform rate matching on the global data pipeline to achieve a balance between processing speed requirements and timing margins;For the hardware implementation of the infrared small target detection algorithm,the algorithm is improved and parallelized acceleration,such as improved multi-scale mean filtering,etc;various designs are carried out in terms of system security,such as the design of self-reset of the camera frame period,the design of Hamming code protection for SDRAM and the improved three-mode redundancy measure XTMR,etc,in order to maximize the long-term stable working ability of the system in the complex space environment.