FPGA Implemention For Large Array Remote Sensing Image Compression System

With the development of sensor technology,the frame frequency and resolution of spaceborne remote sensing image has increased higher,which generates a large amount of datum.The limited bandwidth of the channel can't meet the requirement of image transmission.Therefore,the mass data generated by high-resolution images must be compressed to accommodate the transmission demand.This makes image compression technology become one of the key technologies for spaceborne data processing.Taking the large array remote sensing image as the research object,this dissertation mainly studies the implementation of JPEG2000 compression algorithm.Firstly,according to the trait of image and compression algorithms,an image compression project based on the combination of JPEG2000 algorithm and FPGA platform is proposed.Then the solution to the large array remote sensing image compression system is also given.Secondly,the optimal mode of block size is studied in detail.First,the compression performance of the JPEG2000 algorithm is verified to meet the requirements of the system.Then by analyzing the compression ratio,peak signal to noise ratio(PSNR)and memory requirements,the block size of the remote sensing image with a resolution of 4800×3200 is set as 512×512.Finally,the theory and implementation of wavelet algorithm are studied emphatically.The wavelet transform of the image with a size of 64 × 64 is accomplished on the ISE platform,and the device model is xc5vfx100 t.Compared with the result of Matlab,the average of absolute value of gray difference is 1.4,which has little influence on the visual effect.Meanwhile,the hardware resource consumption can also adapt to processing the image with a size of 512×512.