Paper Details: Downloads: 457
Serial Number: P1150913709
Title: Fuzzy Filters for Noise Reduction in Color Images
Authors: Om Prakash Verma and Madasu Hanmandlu and Anil Singh Parihar and Muralidhar Kulkarni
Abstract: This paper focuses on noise removal in color images with a view to develop two fuzzy filters: one for the Gaussian noise and one for the Impulse noise. For dealing with the Impulse noise two sub filters are arranged in cascade. The first sub filter probes for a set of uncorrupted pixels in the neighborhood of a pixel of interest ( i.e the central pixel) and then computes the median of this set. The amount of noise in the central pixel is calculated based on the similarity of the pixel with its neighborhood and also comparison of this similarity value with the median value. The median value is modified depending upon how much of the impulse noise has affected the pixel of interest. The second sub filter makes use of interaction between the color components. Correction term is calculated based on the difference between a pair of colors. For the removal of the Gaussian noise, use of adaptive distances between the color pairs of the central pixel with the neighborhood is made. Weights to the neighboring pixels are assigned according to these adaptive distances needed for the correction term. The results on test color images show the effectiveness of the fuzzy filters in eliminating both noises.
Keywords: Gaussian noise, impulse noise, correction, adaptive distance, fuzzy filters
Journal/Conference: International Journal of Graphics, Vision and Image Processing
Volume: 9
Issue: 5
Submission Date: 3/27/2009 12:00:00 AM
Review Date: 4/16/2009 12:00:00 AM
Publishing Date: 8/5/2009 12:00:00 AM
Article Downloads: 457
Download:

Facebook