Suppression of Missing Data Artifacts for Deblurring Images Corrupted by Random Valued Noise

For deblurring images corrupted by random valued noise, two-phase methods first select likely-to-be reliables (data that are not corrupted by random valued noise) and then deblur images only with selected data. The selective use of data in twophase methods, however, often causes missing data artifacts. In this paper, to suppress these missing data artifacts, we propose a blurring model based reliableselection technique to select sufficiently many reliables so that all of to-be-recovered pixel values can contribute to selected data, while excluding random value noised data accurately. We also propose a normalization technique to compensate for nonuniform rates in recovering pixel values. Simulation studies show that proposed techniques effectively suppress missing data artifacts and, as a result, improve the performance of two-phase methods. Suppression of Missing Data Artifacts for Deblurring Images 3