Optimal design of morphological filters based on adaptive immune algorithm

Gray-scale morphological filters are a class of important nonlinear filters in the research field of image processing, but whose effects are greatly dependent on the size and shape of its structuring element. Since the selection of structuring element rests with the experience of designers in the classical design of morphological filters, it is difficult to ensure that the selected structuring element is the best one. Immune algorithm has a good property in selecting the optimal parameter and can overcome such drawbacks as immature convergence, poor local searching ability, etc, featured by genetic algorithm. In this paper, we proposed a novel method for designing morphological filters based on adaptive immune algorithm. By the adaptive searching ability of immune algorithm, the best morphological filters with optimal structuring element can be obtained. Compared with the classical approach, our method is more efficient and powerful.