A parallel implementation of a fractal image compression algorithm

Abstract

Presented in this paper are experimental results for a parallel implementation of a fractal image compression algorithm. The fractal image compression process exhibits a natural parallelism and the specific parallel implementation is shown to be beneficial for the execution speed of the algorithm. Furthermore, computational speed has been enhanced through an algorithmic modification employing a quadtree recomposition approach in which quadtree subblocks are iteratively combined into larger ranges for fractal encoding. Furthermore, classification schemes which serve to reduce the O(n/sup 4/) algorithmic complexity are explored.