Restoration of degraded Images using Denoised model with DSWT Fusion and Enhancement Technique

Abstract

Image restoration is a craftsmanship to enhance the nature of image via evaluating the measure of noises and blur engaged in the image. Notwithstanding the noteworthy research led on this theme, the improvement of proficient denoising strategies is as yet a convincing test. The significant deficiency is that while upgrade, the brightness of the image falls apart in a considerable amount. The image fusion methods perform well spatially but usually introduce spectral distortion. Which means that the variation of hue before and after the fusion process has appeared? There is color distortion when the fusion is appeared in the color images. There are human representation and target assessment criteria related issues when the fusion of two Images happened. The Hue, Saturation and the Intensity of the color images effected due to fusion. Image denoising is a basic prerequisite of image processing as the images contain firmly arranged music and edge discontinuities. Enhancement is finished by Spatial filtering technique known as Histogram Equalization. We have done comparison with our proposed method in which we hybrid the Median Filter with Bayes shrink Wavelet thresholding technique for Denoising and enhancing the images as to preserve brightness more, results in better visualization. Results are estimated by parameters such as PSNR, MSE, CoC and Elapsed Time which demonstrates our hybrid technique has best outcomes from different techniques that are such as Median filter, wiener filter, Wavelet thresholding, Bayes shrink Method etc.