A New Class of Generalized Morphological Filters

Mathematical morphological filter is a class of nonlinear filters. In order to improve the performance of morphological filters, a type of new generalized morphological filters is proposed in this paper. The proposed filters inherit some important properties of the classic morphological filters, such as the translation invariance, incresingness, duality, and so on. Compared with the traditional morphological filters, the proposed filters can preserve more image details (the geometrical features and edges) while remove noise in the image and their speed is faster. Experimental results show that the filtered images which processed by the proposed filters possess high peak signal-to-noise ratio and low root mean square error.