We propose the vector rank M-type K-nearest neighbor (VRMKNN) filters to noise suppression and detail preservation in color image restoration. The proposed VRMKNN filters are based on the combined RM-estimators with different influence functions, and the KNN algorithm. Therefore, we designed the adaptive multichannel non parametric VRMKNN (AMN-VRMKNN) that uses an adaptive non parametric approach to determine the functional form of the density probability of noise into the sliding filtering window to improve the performance of the VRMKNN filters. Numerous simulations illustrate that the proposed filters exhibit robust and adaptive capability in multichannel imaging applications. Finally, we present the implementation of proposed filters on the DSP TMS320C6711 demonstrating that they can potentially provide a real-time solution to quality video transmission.

Gallegos-Funes, F. J. et al. (2006). DSP Implementation of RM-filters for Impulsive Noise Suppression in Color Images. Científica: La Revista Mexicana de Ingeniería Electromecánica, 10(1), pp. 23-37.