Still image denoising based on discrete wavelet transform

Abstract

The growth of media communication industry and demand of high quality of visual information in modern age has open an interest to researcher to develop varies method of image denoising based on DWT. The visual information transmitted in form of image is naturally corrupted by Gaussian noise which is classical problem in image processing. This additive random noise can be removed using wavelet denoising technique due to the ability to capture the energy of a signal in few energy transform values. In this paper, an investigation has been made on suitability wavelet thresholding and translation invariant methods of image denoising to remove noise using orthogonal wavelet basis. The performance of the image denoising is shown in terms of PSNR and visual performance. The result shown translation invariant gave better PSNR and visual performance than wavelet transform method.