On the efficiency of a fast technique of impulsive noise removal in color digital images

In the paper the detection efficiency of a fast algorithm of impulsive noise reduction in color images is investigated. The analyzed algorithm is based on the sum of dissimilarity measures between the processed pixel and its neighbors. The detection part of the algorithm is mainly examined using binary classification, in contrast to other works, which focus on overall impulsive noise detection and suppression efficiency. The most efficient pixel dissimilarity measures are depicted among various distance measures and tuning parameters for the analyzed algorithm, adjusted to the impulsive noise intensity, are recommended.