Frame-groups based fractal video compression and its parallel implementation in Hadoop cloud computing environment

Abstract

Fractal video compression is based on the self-similarity search between range cubes and domain cubes, so it can achieve a high compression ratio. However, its computational complexity is relatively high that restricts its studies and applications. Further studies show that the compression process exhibits a high natural parallelism as there exist data independence when computing the compression codes. In this paper, we utilize parallel processing techniques to implement the fractal video compression algorithm to reduce the run time. There are two main works in this article: firstly, a parallel fractal video compression algorithm based on frame-groups is proposed. Secondly, we implemented the parallel algorithm in Hadoop cloud computing environment. The experiment results show the parallel algorithm has a high speedup and the distributed parallel computing systems can utilize network resources sufficiently to implement high-performance computing, and provide a good practicability and a promising future in application.