Soft switching technique for impulsive noise removal in color images

In this paper a novel, fast soft switching filtering technique intended for impulsive noise removal in color images is presented. The impulsive noise detection algorithm is based on the concept of trimmed cumulated distances, which are assigned to the central pixel of a sliding filtering window. The trimmed sum of distances serves as an indicator of the strength of impulses injected to the image by the impulsive noise process and the restored pixel is obtained as a weighted mean of the processed pixel and the sharpening vector median filter output. Extensive experiments, performed on a set of color test images, proved that the proposed filter is superior to the state-of-the-art impulsive noise reduction algorithms. The new noise filtering design, due to its simple structure and very low computational complexity, can be applied in various applications in which the detail preserving impulsive noise reduction is required.