Image Denoising with Modified Wavelet Feature Restoration

Image denoising is the principle problem of image restoration and many scholars have been devoted to this area and proposed lots of methods. In this paper we propose modified feature restoration algorithm based on threshold and neighbor technique which gives better result for all types of noise. Because of some limits of conventional methods in image denoising, several drawbacks are seen in the conventional methods such as introduction of blur and edges degradation. Those can be removed by using the new technique which is based on the wavelet transforms. The shrinkage algorithms like Universal shrink, visue shrink, bays shrink; have strengths in Gaussian noise removal. Our proposed method gives noise removal for all types of noise, in wavelet domain. It gives a better peak signal to noise ratio as compared to traditional methods