International Journal of Computer Applications
This paper presents a new weighted switching median filter is proposed to restore images corrupted by noise. The neighborhood of a pixel is classified into low corrupted, medium corrupted and highly corrupted regions. Based on this classification the weights associated with the median filter employed for denosing is determined. In addition, switching mechanism allows the use of different weights for different regions based on localized statistics. The proposed algorithm can be applicable to digital images with most of the fine details of the image are mostly preserved. The simulation analysis would shows that the proposed method is good for noise reduction and image restoration.