Study On Fractal Image Compression

Fractal image coding is a very promising compression technique. It has been paid great attention since its presence because of its novelty, high compression rate, resolution independence and rapid decoding. Firstly this paper introduced the significance and method of image compression, the origin and the actuality of fractal image compression. Fractal image compression bases on IFS(Iterated Function System) theory, so the IFS theory is presented afterwards, we also introduced the steps of image coding using IFS theory.At high compression rate all block-based image compression methods have a common defect, that is, tile effect. There is no exception for fractal image coding. So, in the chapter 4, we proposed a simple but effective method to reduce the tile effect. Firstly, we decode the image at lower resolution. Then, we employ the bilinear interpolation technique to reconstruct the image at the same resolution. Compared with other methods, this method does not have to alter the encoding steps.Another disadvantage of fractal image compression is long time-consumption. According to this, we proposed three fast algorithms in chapter 5. Among lots of fast algorithms, most algorithms base on the block-classified technique, but the effect is limited. So we used the most novel matching method-feature vector method to propose two fast algorithms: one is relative gradient-based algorithm, the other is cross trace-based algorithm. These two algorithms converted the range-domain block- matching problem respectively to the nearest neighbors search problem in the sense of relative gradient and cross trace. The search pool decreased greatly, so the time-consumption is much more less than before. Finally this paper also proposed a new algorithm based on the mean deviation. The mean deviation can be considered as a feature when the two blocks were matched. The simulations showed the three algorithms are effective.