Abstract
The color filters that are used to attenuate noise are usually optimized to perform extremely well when dealing with certain noise distributions. Unfortunately it is often the case that the noise corrupting the image is not known. It is thus beneficial to know a priori the type of noise corrupting the image in order to select the optimal filter. A method of extracting and characterizing the noise within a digital color image using the generalized Gaussian probability density function (pdf) (B.D. Jeffs and W.H. Pun, IEEE Transactions on Image Processing, 4(10), 1451–1456, 1995 and Proceedings of the Int. Conference on Image Processing, 465–468, 1996), is presented. In this paper simulation results are included to demonstrate the effectiveness of the proposed methodology.