The compression effects of the binary tree overlapping method on digital imagery

A new method of digital image compression called binary tree overlapping (BTO) is described. With this method, an image is divided into bit planes that are transformed into a binary tree representation in which initial identical portions of different lines in the bit planes are transmitted as one path in the tree. To increase compression, distortion can be introduced by considering two lines which differ in fewer that a preset number of bit positions to be identical. A time efficient method of implementing BTO based on a communication technique called content-induced transaction overlap is outlined. With this technique, only simple, logical operations are needed and the compression time is proportional to the number of bits in the compressed image. Simulation studies indicate that BTO produces good quality images with a compression ratio of about three. In terms of implementation and compression efficiencies, BTO is comparable to first-order DPCM.< >