Image Denoising in Wavelet Domain Based on Thresholding with Applying Wiener Filter

An image is often corrupted with noise throughout procurement, compression, transmission, storage and retrieval processes. These effects are leading to distortion and loss of image information. Image denoising used to eliminate the noise in order to reserve all the fine details in the image while retaining as much as possible the vital signal features. Wavelet denoising aims to remove the noise in the signal while maintaining the features of the signal, regardless of its frequency content. In this work, a new approach is introduced to denoising image that has been affected by Additive White Gaussian Noise (AWGN). The proposed system realized using Wiener filter before and after the wavelet transform. To remove noise from pixels in the wavelet domain, discrete wavelet transform (2D-DWT) is applied. Threshold techniques and Wiener filter have been used for denoising. Then, the 2DIDWT inverse discrete wavelet transform applied to remove noise and complete the denoising technique. Also, in this work, the image is denoised using the connotation of Wiener filtering and denoising method in the wavelet domain with multiresolution at three levels. The performance of the proposed methods has been measured by using the Peak Signal to Noise Ratio (PSNR). Experimental evaluation shows that the results of the proposed methods give an improvement with about 17.5% through the comparison with the results of the related works and the essence of images is improved in terms of noise-reducing better than using a wavelet transform or Wiener filter solo as well as edge preservation.