Parallelization of fractal image compression over CUDA

Abstract

Fractal Image Compression has been a promising scheme to allow for substantially high compression ratios for image compression. Several algorithms have been proposed based on this scheme. But the enormous amount of computing and therefore the long runtime needed to encode the image makes these algorithms impractical for commercial purposes especially on personal computers. Recently most of the personal computers & laptops are being shipped in with dedicated graphic processing units (GPU). In this paper we present an approach to parallelize the Fractal Encoding Scheme over GPU using CUDA. This makes fractal image compression feasible for personal computers. © 2013 Springer Science+Business Media New York.