DENOISING OF IMAGES WITH MULTIPLICATIVE NOISE CORRUPTION (TueAmOR2)
Author(s) :
Zhiling Long (Mississippi State University, United States)
Nicolas Younan (Mississippi State University, United States)
Abstract : Multiplicative noise is signal dependent and is difficult to be removed without impairing image details. It causes difficulties for many real world imaging applications. Previously, a hypothesis test based wavelet denoising algorithm had been proposed with promising results. In this paper, the algorithm has been further studied by fitting it into the framework of contourlet transform, an emerging two-dimensional technique for image processing and analysis. The developed contourlet denoising algorithm has been evaluated with standard test images, yielding successful results. It has also been demonstrated that the proposed algorithm outperformed the original wavelet based approach.
Menu