Local Restoration of Images Distorted by Rician Noise

The problem of restoring an image distorted by the Rice model is considered.The restoration algorithm is based on estimating the parameters of the Rice distribution from the sample and correcting distorted pixels accordingly.The solution to the problem is based on the concept of using a restricted local neighbourhood (RLN) of the recovered pixel, which is known in the literature.It is proposed to test the statistical hypothesis of conformity to the Rice distribution for each selected RLN.A method for selecting RLNs with sufficiently large signal-to-noise ratios is proposed.The results of the experimental study are given.