Generalized self-dual morphological filters and the applications in image-denoising

As the traditional self-dual morphological filters(SMF)depend on two mutual-dual morphological filters, which can preserve the image details, but can not suppress noise effectively, the generalized self-dual filters namely the self-dual median filter and the mean filter, which can remove impulse noise and Gassian noise respectively, were proposed based on modified morphological median operator(MMO) and mean operator(MAO) respectively. Experimental results show that the proposed generalized self-dual morphological filters can suppress noises efficiently while maintaining the image brightness, and the filtered image has high peak signal-noise-ratio(PSNR) and low root mean square error(RMSE).