Study Of Fractal Image Compression Coding Based On Modern Optimization Algorithm

Fractal image compression coding is focused attention upon by researchers for it's good quality of high compression ration, resolution irrelevancy and fast decoding. But it can't be put into practice for the matching search of domain and range is too difficult and the coding time is too long.Modern optimization theory mainly focuses on the optional sorting, arranging, grouping and s electing of separated events. The o ptional or hypo-optional s olution o f such problem could be obtained.The Jacquin algorithm of fractal image coding is modified according to the modern optimization theory. The aim is to reduce the coding time.At first the problem of the matching search of domain and range in the Jacquin algorithm of fractal image coding has been analyzed and the result of it belonging to the NP_HARD has been found. Then modern optimization theory has been applied into the matching search of domain and range in fractal image coding. The local-search, taboo-search, and ordinary genetic algorithms are applied into the matching search of domain and range respectively, but the results are all not very perfect. The taboo-search algorithm and the ordinary genetic algorithm are combined in this paper and the new algorithm is named taboo-genetic algorithm. The new algorithm is applied into the matching search of domain and range in fractal image coding. The result shows that the coding time is decreased greatly when the compression ration and the peak signal and noise ration do not decline. The validity of the algorithm has been tested by programming experiment with VC++. The compare experiment has been tested with five different types of JPEG typical testing images.At last the new taboo-genetic algorithm has been combined with a self-adapt fractalimage coding algorithm. The feasibility and efficiency of the new algorithm is proved again.