Wavelet Based Image Denoising and Adaptive Threshold Optimization

Improving quality of noisy image is very important and it is one of the major area of research. Wavelet thresholding technique are very important for Denoising of images as compared to the classical methods. However it is difficult to suppress additive noise since it corrupts almost all pixels present in an image. The arithmetic mean filters are used to suppress AWGN but it introduces a problem of blurring. The objective of denoising is to suppress the noise quite efficiently while preserving the edges and other important features as much as possible. The filter-performances are compared in terms of Peak Signal to Noise Ratio (PSNR), Mean Squared Error (MSE) and Structure Similarity (SSIM). Large values of PSNR and small values of MSE indicate less noise. Simulated noise images are used to evaluate the denoising performance of proposed algorithm along with another wavelet-based denoising algorithm. Experimental result shows that the proposed denoising method outperforms standard wavelet denoising techniques in terms of the PSNR and the preservation of edge information.