Abstrakt

Performance analysis of Absolute Deviation Filter for Removal of Impulse Noise

G.Bindu, M.Upendra , B.Venkatesh, G.Gowreeswari , K.T.P.S.Kumar

This paper presents the process of removal of impulse noises using different median filters. This contains a vector median filter that includes a new mechanism for the detection of impulses in colour images prior to further processing operations. When compared to other filters, the proposed filter will be tested for images corrupted by two sided fixed impulse noise model, nothing but salt and pepper noise. If the central vector pixel in a kernel is found to be corrupted, it is replaced by the vector median of the kernel, else it is kept unchanged. The vector pixels in a specified window are ranked on the basis of sum of the distances between each vector pixel and the other vector pixels in the window. Since the detection mechanism uses absolute deviation values, the computational complexity is reduced. The results show a better performance of the proposed algorithm in lesser execution time as compared to a number of existing schemes for restoring colour images corrupted by impulse noise.