An improved image denoising using wavelet transform

Image is one of the most important part of multi-media that is used in several areas from simple photography to medical and satellite imaging. In each field its usage and requirements are very different. So, an image required to be clean and free from artifacts to convey better information. But, acquisition is always associated with some sort of degradation that may be due to atmospheric conditions, camera sensors and/or lighting conditions. In this paper we are considering the degradation only due to noise and in specific additive Gaussian noise. Here, we are proposing to use a dual step approach for denoising. In the first step it uses stationary wavelet based denoising and in continuation to second step, a spatial domain method, Non-local means, is used to remove the artifacts. The simulation is done on both real and synthetic images and it shows an improvement over existing methods. Image is one of the most important part of multi- media that is used in several areas from simple photography to medical and satellite imaging. In each field its usage and requirements are very different. So, an image required to be clean and free from artifacts to convey better information. But, acquisition is always associated with some sort of degradation that may be due to atmospheric conditions, camera sensors and/or lighting conditions. In this paper we are considering the degradation only due to noise and in specific additive Gaussian noise. Here, we are proposing to use a dual step approach for denoising. In the first step it uses stationary wavelet based denoising and in continuation to second step, a spatial domain method, Non-local means, is used to remove the artifacts. The simulation is done on both real and synthetic images and it shows an improvement over existing methods.