A Weighted Unsymmetrical Trimmed Mean Median Based Impulse Noise Removal For Color Image Enhancement

Abstract

The elimination of colour impulse noise in digital Image is a broadly renowned area of investigation. A typical problem in image analysis is the presence of colour noise sources, which originates at the moment of photograph acquisition or transport. Fixed valued   and Random Valued Impulsive Noise, are two types of chromatic impulsive noise. Image enhancement, satellite imagery, medical imaging, and robotic all require colored image processing. Impulse noise is present in all of these applications.  Since it protects edges, median filter is commonly used to eliminate this noise from an images.  In the research work proposed a weighted unsymmetrical trimmed median  for the elimination of impulse noise at both medium to high density values . Detection of the impulse noises on the basis of maximum and lowest pixel values is the first step in the impulse noise filtration, followed by the elimination of these noises.  weighted unsymmetrical trimmed median computation is used in the second step of noise reduction for case-1, trimmed median for case-2, and WMF for case-3. Using noise-free pixels to de-noise an image achieves improved result and higher visual quality. Also present good result in terms of peak signal to noise ratio, mean square error and other result parameters.