An Algorithm for Impulsive Noise Removal in Color images

We propose an effective algorithm for impulsive noise removal in color images. The proposed method is a two-phase scheme. In the first phase, an impulse detector is designed to identify color pixels which are likely to be contaminated by noise; in the second phase, the image is restored by using a detail-preserving regularization method that applies only to those selected noise candidates. Extensive computer simulations indicate that our proposed algorithm provides significant improvement over many of the existing filtering algorithms in suppressing impulsive noise in color images.