A Multi-DSP Based Remote Sensing Image Compression System

Remote sensing image data is huge and increasing unceasingly, which challenges real-time processing. In accordance with this feature, the paper presents a kind of hardware architecture solution based on 4-DSP blocks, designs and realizes a high performance remote sensing image compression model machine. This paper gives out a compression algorithm which takes advantage of the correlation between near neighhbour, far neighbour and father-children coefficients to predict. The algorithm is better than SPIHT and JPEG2000, and is transported into the model machine successfully. At last, the paper gives out the experiment results and the improving directions