Noiseless (or reversible) methods are preferred over noisy (or irreversible) methods for the compression of grayscale images in situations where the quality of the images is critical and/or where postprocessing of the images is involved. Two significant application areas for such methods are in the storage and transmission of medical images and industrial radiographic images used for nondestructive evaluation (NDE). The main objective of the proposed research is to improve existing and develop new methods for the noiseless compression of grayscale images. Sophisticated source modeling techniques and the powerful arithmetic coding technique are used for this purpose. It is expected that some of the new methods to be developed will perform 30% to 40% better than the existing methods. Extensive statistical analyses of images and digital computer simulation experiments will be the primary tools in implementing the project. In addition to providing new, improved methods for the compression of grayscale images, the research project will also help shed some light on certain important aspects of the noiseless data compression area as a whole.
