Restoration-Multiple Removing of Noise under Water Images using Multidirectional Filtering Techniques

2 Abstract: Image restoration one part is the Denoising which plays important tasks in image processing. Despite the significant research conducted on this topic, the development of efficient denoising methods is still a compelling challenge. Image denoising is an essential requirement of image processing. The images contain strongly oriented harmonics and edge discontinuities. Wavelets, which are localized and multiscaled, do better denoising in single dimension using multiple local thresholding technique. Filter based denoising and reconstruction exhibit higher quality recovery of edges and curvilinear features. This thresholding scheme denoises images embedded in Speckle noise. The experiment shows denoising using Filters such as Wiener, Median, Wavelet Transform , Bayes Shrink and our proposed technique as (median and bayes Shrink wavelet) to outperforms in terms of PSNR(peak signal-to-noise ratio) , MSE (mean square Error), Elapsed time and Coc (Coefficient of Correlation ), but also in better visual appearance of the resulting images. In this thesis, we will study and investigate the application of using best filters to remove noise using our proposed method Median with Bayes shrink wavelet with soft thresholding for denoising techniques to remove multiple noises from under water images. In this, Gaussian, Poisson, Salt & pepper, Speckle is used for restoration. Our Technique, works best for all types of noises but speckle is better restored as denoised by wavelet based only technique.