This paper addresses the problem of image denoising using wavelet thresholding, focusing on the development of a spatially adaptive thresholding method. The authors propose a technique that models each wavelet coefficient as a random variable following a generalized Gaussian distribution (GGD) with an unknown parameter, and uses context modeling to estimate these parameters. This allows for adaptive thresholding, where the threshold is adjusted based on the local characteristics of the image, such as smooth or edge regions. The method is extended to overcomplete wavelet expansions, which outperform orthogonal transforms in terms of denoising performance. Experimental results show that the proposed spatially adaptive wavelet thresholding method significantly improves image quality and reduces mean squared error compared to uniform thresholding and other state-of-the-art denoising methods.This paper addresses the problem of image denoising using wavelet thresholding, focusing on the development of a spatially adaptive thresholding method. The authors propose a technique that models each wavelet coefficient as a random variable following a generalized Gaussian distribution (GGD) with an unknown parameter, and uses context modeling to estimate these parameters. This allows for adaptive thresholding, where the threshold is adjusted based on the local characteristics of the image, such as smooth or edge regions. The method is extended to overcomplete wavelet expansions, which outperform orthogonal transforms in terms of denoising performance. Experimental results show that the proposed spatially adaptive wavelet thresholding method significantly improves image quality and reduces mean squared error compared to uniform thresholding and other state-of-the-art denoising methods.