Support Vector Machine for Fast Fractal Image Compression Base on Structure Similarity

Fractal image compression is promising both theoretically and practically. The encoding speed of the traditional full search method is a key factor rendering the fractal image compression unsuitable for real-time application. The primary objective of this paper is to investigate the comprehensive coverage of the principles and techniques of fractal image compression, and describes the implementation of a pre-processing strategy that can reduce the full searching domain blocks by training the Support Vector Machine which could recognized the self-similar pattern feature to enhance the domain block searching efficiency. In this paper, the novel image quality index (Structure Similarity, SSIM) and block property classifier based on SVM employed for the fractal image compression is investigated. Experimental results show that the scheme speeds up the encoder 15 times faster and the visual effect is better in comparison to the full search method.