FPGA Implementation Of 9/7 Wavelet Transformation For Image Compression

Space images are important carriers for deep-space explorations and near-earth observatories,and they play important roles in the fields of spaceflight and social economic life.Because of the contradiction between the magnanimous data of the space images and the limitation of download-channel-bandwidth,real-time,highly effective image compression algorithm is the inevitable choice to reduce storage volume and transmission bandwidth.In recent years,lots of novel image compression algorithms are proposed,such as JPEG2000,SPIHT and so on.But these algorithms all have their own deficiencies. Therefore,the CCSDS organization proposed an algorithm for effective space image compression.The algorithm has lower complexities and good parallel architecture, and is suitable for the high-speed low-power hardware realization and real-time processing.So it has the good application prospect in the outer space exploration.Based on the characteristic of space image,this paper analyzes the theory of 9/7 integer DWT which is selected by the CCSDS to perform lossless image compression and proposes a FPGA architecture design of 2D-linfting wavelet transformation.The whole architecture is optimized in the cycle overlay of RAM and the pipeline design,and it takes full advantage of the memory bits in FPGA, which lead to row and column processors work in parallel.The whole architecture is simulated with Quartusâ…¡and the results are identical with the Matlab's,so the validity of this architecture is proved.The hardware design of 9/7 integer DWT system is completed and the USB interface is designed at the stage of FPGA verification to decrease the hardware verification time.Experiments show that the system works stably and has ability of real-time processing.