Image compression is the application of data compression on digital images. Image compression can be lossy compression or lossless compression. Raw data from an image sensor obviously contains information about a scene, but it is not intrinsically recognizable to the human eye. In signal processing, a lapped transform is a type of linear discrete block transformation where the basis functions of the transformation overlap the block boundaries. We propose a new shape-adaptive (SA) transform ,shape-adaptive reversible integer lapped transform (SA-RLT) method. Region-of interest (ROI) coding algorithms have also been proposed, since observers are always more concerned about some special areas, i.e., ROI than the background (BG) area. Based on SA-RLT and object-based set partitioned embedded block coder (OBSPECK), a new ROI compression scheme can be designed for 2D images.