Denoising: A dual domain method

Abstract

Image has an important role to play in our daily life. It has its applications from simple documentation to complicated surveillance and medical applications. In the area of image processing, denoising is one among the most studied areas. Many a times the captured image will be degraded. This can happen at the time of acquisition and/or transmission. Noise is one such degrading agent. The presence of noise will affect the performance of the applications like segmentation, recognition, object detection and medical as well as general applications. Hence denoising is a prerequisite in these applications. The proposed method utilizes both transform and spatial domains. Shrinkage technique is applied in wavelet domain and in spatial domain, non-local means is used. Simulation is conducted on standard test images. The tabulated results shows that, the proposed method performs comparatively better.