Performance analysis of impulsive noisy image restoration filters

An adaptive simplified model Kalman filter (ASMKF) reported recently has been shown to be more effective in suppressing the impulsive noise, especially when the signal-to-noise ratio is low. However, the typical image model parameters used in the filter may not be optimal for all the images. Hence, in this paper, an RLS algorithm is formulated to estimate the unknown image model parameters. Then, the ASMKF and RLS estimator are coupled to estimate jointly the image model parameters and the restored image pixels. Performance of the proposed approach is analyzed in comparison with the standard median filter, truncation filter, and the cascade truncation filter through implementation results on impulsive noisy color image restoration.