Parallel implementation of fractal image compression

Abstract

Fractal image encoding is based on the self-similarity search between range and domain blocks. It can be parallelized in three different ways. In this paper, we first formulate the encoding process and then present one of the parallel implementations on KSRI (Kendall Square Research parallel computer). Experimental results include the encoding time as a function of the number of processors and the speedup curve. The speedup of the encoding process is significant when compared to serial computation.