Research and realization of image separation method based on independent component analysis and genetic algorithm
Junan Yang, Liang Tao, Zhenquan Zhuang, Guo Li · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2002
An advanced Independent Component Analysis (ICA) algorithm based on genetic algorithm is proposed with analysis to the ICA method. The proposed algorithm can be used to solve the problem of local optimum that is easily stacked into by normal numerical solution. The image separation simulation shows that the global optimum can be acquired through the proposed algorithm under the circumstance of adequate colony size and genetic generations.