L.-W. Chang, W.-N. Lie, and R. Chiang (Eds.): PSIVT 2006, LNCS 4319, pp. 878 – 887, 2006. © Springer-Verlag Berlin Heidelberg 2006 Wavelet-Based Image Compression with Polygon-Shaped Region of Interest Yao-Tien Chen 1 , Din-Chang Tseng 1 , and Pao-Chi Chang 2 1 Department of Computer Science and Information Engineering, National Central University, Chung-li, Taiwan {ytchen, tsengdc}@ip.csie.ncu.edu.tw 2 Department of Communication Engineering, National Central University, Chung-li, Taiwan pcchang@ce.ncu.edu.tw Abstract. A wavelet-based lossy-to-lossless image compression technique with polygon-shaped ROI function is proposed. Firstly, split and mergence algorithms are proposed to separate concave ROIs into smaller convex ROIs. Secondly, row-order scan and an adaptive arithmetic coding are used to encode the pixels in ROIs. Thirdly, a lifting integer wavelet transform is used to decompose the original image in which the pixels in the ROIs have been replaced by zeros. Fourthly, a wavelet-based compression scheme with adaptive prediction method (WCAP) is used to obtain predicted coefficients for difference encoding. Finally, the adaptive arithmetic coding is also adopted to encode the differences between the original and corresponding predicted coefficients. The proposed method only needs less shape information to record the shape of ROI and provides a lossy-to-lossless coding function; thus the approach is suitable for achieving the variety of ROI requirements including polygon-shaped ROI and multiple ROIs. Experimental results show that the proposed lossy-to-lossless coding with ROI function reduces bit rate as comparing with the MAXSHIFT method in JPEG2000; moreover, when the image without ROI is compressed by the proposed lossless coding, the proposed approach can also achieve a high compression ratio. Keywords: Image compression, region of interest (ROI), lossy-to-lossless coding, ROI coding, difference encoding. 1 Introduction Image compression is used to reduce the image data size as small as possible under a tolerance limit of errors. In general, the techniques of image compression can be classified into two major categories: loss and lossless. Lossy compression requires not only less storage space, but also less transmission time or bandwidth, while lossless compression can completely reconstruct the original data. In addition to offering high- quality compression, an effective approach to image compression should further incorporate value-adding functions, such as ROI coding and lossy-to-lossless coding. A ROI refers to a special region in an image that is of particular interest or imperative importance to the user who can free to identify the ROI based on ones needs. In