Novel multiscale filtering method based on mathematical morphology

Mathematical morphology has been applied in many fields. A morphological filter as a nonlinear technique that is making great progress is developed for image noise elimination. Considering the details losing in image filtering, in this paper, using genetic algorithms in the design of morphologic filters and presents a novel adaptive multiscale filtering method based on mathematical morphology (GAMMF). First, the method of multiscale morphological filter used in this paper is introduced. Then, the method adds the multiscale parameters optimization to the conventional multiscale morphological opening and closing filtering. The parameters of the multiscale have great effects on the performance of the whole filtering. So, the parameters are optimized by aid of the genetic optimization algorithms.