BINARY DECOMPOSITIONS FOR HIGH-ORDER ENTROPY CODING OF GRAYSCALE IMAGES

High-order entropy coding (HOEC) has the potential to provide higher c ompression ratios than the usually used zero-order entropy coding (ZOE C) approaches. However, serious implementation difficulties severely l imit the practical value of HOEC for grayscale image compression. We e xamine the bit-plane decomposition (BPD) representation as a simple al ternative that bypasses some of the implementation difficulties of HOE C. We show, however, that BPD introduces undesired coding overhead whe n used to represent grayscale images. We therefore propose a new binar y image representation called magnitude-based binary decomposition (MB BD) which avoids any coding overhead when used to represent grayscale images. Thus, MBBD both bypasses the implementation difficulties of HO EC and does not have the drawbacks of the BPD. We present numerical ex periments that verify the theoretical analysis of the BPD and MBBD rep resentations. In addition, our experiments demonstrate that MBBD-HOEC yields better results than ZOEC for lossy image compression and is als o very effective for progressive image transmission.