ROUNDED REDUCED-DIFFERENCE PYRAMID FOR SEMILOSSY IMAGE-CODING

An extension of the reduced-difference pyramid (RDP) algorithm is deri ved and proposed for lossy image compression with upper-bounded maximu m error (semilossy). Quantization has been introduced into the basic R DP in two different ways: approximations are made either before or aft er taking the differences among pairs of adjacent nodes. The resulting peak-error (PE) propagation along the pyramid is theoretically invest igated, and quantization strategies are suggested. Tests on the standa rd gray-scale Lena image show significant and steady rate-distortion i mprovements over the original RDP. As few fixed-point operations are r equired per pixel, on-line compression/decompression is feasible on ge neral-purpose computers.