Study of Performance of Different Fractal Image Compression Techniques

For fast processing of image data which is important in knowledge based systems computing, mass storage and easy retrieval image compression becomes a key technique. Fractal image compression is a relatively recent technique incorporated in the compression techniques now-a-days. It is a lossy compression technique in the field of Image compression. This principle encompasses a very wide variety of coding schemes, many of which have been explored in the rapidly growing body of published research. Despite the exploration of many encoding techniques for efficient and fast compression, due to the long computational expense of suitable domain search the encoding phase of this technique is very time consuming. This review addresses the problem of computational complexity and represents a survey of the most significant speed-up techniques in the fractal image coding scheme. The aim of this paper is to compare some of the most significant speed-up techniques such as Classification techniques (namely Fisher scheme and Hurtgen scheme), Genetic algorithm schemes, DCT based techniques and Feature vector based techniques. The five significant speed-up techniques are compared based on the performance metrics such as compression ratio, speed-up and PSNR and a performance comparison is made based on these metrics.