A Comparative Evaluation of Denoising of Remotely Sensed Images Using Wavelet, Curvelet and Contourlet Transforms

This paper presents an overview of remotely sensed image denoising based on multiresolution analysis. In this paper, the wavelet, curvelet and contourlet transforms are used for denoising of remotely sensed images with additive Gaussian noise. The curvelets and contourlets are two kinds of new multi-scale transforms which can capture the intrinsic geometrical structure of data. At first, we outline the implementation of these multiscale representation systems. The paper aims at the analysis of denoising of image using wavelets, curvelets and contourlets on high resolution multispectral images acquired by the QuickBird and medium resolution Landsat Thematic Mapper satellite systems. We apply these methods to the problem of restoring an image from noisy image and compare the effects of denoising. Two comparative measures are used for evaluation of the performance of the three methods for denoising. One of them is the peak signal to noise ratio and the second is the ability of the denoising scheme to preserve the sharpness of the boundaries. By both of these comparative measures, the curvelet has proved to be better than the other two.