Combined Multiple Image Filter for noise removal

We introduce Combined Multiple Image Filter (CMIF) as a new approach for noise removal. The processed output is the Bayesian mean square error estimate of the original image. A set of four filters are selected to improve the quality of noisy image that contains regions with different characteristics. Feature vectors are computed at each pixel in order to predict the performance of the filters in estimating the corresponding pixel intensity in the original image. The final output is the convex combination of combined filters outputs.