Near-Lossless Image Compression Based on Lossless Regions of Interest(ROI)
基于感兴趣区的图象近无损压缩

Image compression based on Lossless Region of Interest (ROI) means to compress interesting regions in an image without loss, and to compress uninteresting regions with a little loss. Based on this idea a high compression ratio can be obtained and the important information can be preserved losslessly. In this paper, the near lossless image compression is implemented, based on Integer Wavelet Transform(IWT) and embed zero tree coding scheme. In addition, the tree mapping shape coding scheme is proposed. In this paper the IWT and embed zero tree coding scheme are introduced simply, then the tree mapping shape coding algorithm are presented in detail, finally the experimental results are presented and compared with the results of lossless compression. Experimental results represent that the algorithm can obtain much higher compression ratio than lossless compression algorithms, but the improvement of compression ratio is related to the size of ROI and quality of uninteresting regions.