Abstract:
Fractal image compression for color image uses the same concept as those of "Jacquian's algorithm". It searches for portion of domain block which has relationship with range block within the color image. It then divides the color image using the quad-tree partition with non-overlap approach and collects data from the domain block. The model of the color image is a factor which has impact on the efficiency of this algorithm. In order to find any benefit or drawback of the fractal's approach, there are comparisons, graphical representation and conclusion based on various models of color image in this paper.The test indicates the color model YUV is suitable for various application. Statistically, this approach has a unique characteristic which will reduce the color radiation of the U and V color channels. It also reduces the difference between domain and range block, this will in turn keeps the changing of the decompression image to a minimum as the other color model approaches. Based on human vision system, it is quite clear that at a higher compression ration, the YUV color model has a better image structure when compared on the RGB, CMY, HSV color models.