Robust reduced update Kalman filtering

A robust version of the two-dimensional reduced-update Kalman filter (RUKF) is presented. The robust RUKF (RRUKF) can be used for restoring images corrupted by noise whose distribution deviates from the Gaussian distribution. It was observed experimentally that when the image is corrupted by 'salt and pepper' type noise, the RRUKF, which is an extension of the RUKF, produces significantly better results than the RUKF, as expected.<
>