Generalized morphological operators for noise reduction

Based on a pair of structure elements which have the same size and the different shape, a novel type of generalized morphological operators is presented for the noise reduction. The operators can suppress noisy structures which are larger than structure elements while preserving edges and details in the image, and they inherit most of the properties of the classic morphological operators except the extensibility and anti-extensibility. Furthermore, the presented operators are less active compared with the classical morphology operators. The experimental results show that the generalized morphological operators can suppress noise efficiently while preserving the details in the image with higher peak signal-to-noise ratio and smaller root mean square error than many improved morphological operators.