This Phase II project addresses the problems in the quality of images transmitted in a computer environment. A common problem is the "tile effect" which is introduced by compression of the signal for transmission and then decompression for viewing. The principal investigator is developing a new algorithm, called the JPEG/RBC algorithm, which is based on a two source decomposition derived from a noncausal model for image data and fits closely within the framework of the JPEG standard. A preliminary version of the algorithm was shown in Phase I to reduce this problem as well as the compressed data rate when compared to commonly used algorithms. Potential applications include photo ID systems, medical imaging, intelligence and prepress.