An efficient approach for denoising of noises

This paper addresses the problem of image denoising which is still a valid challenge for developing algorithms to reduce the noise that is inherent in all digital images. This paper presents efficient approach for denoising of noises in image. Several algorithms for noise reduction are well established in the domain of image processing. The aim of the paper is to characterize the salt-pepper, Gaussian and Poisson noise & mixture of two or three type of noises or any type of noise present in the image is denoised by proposed algorithm. Filtering methods i.e. MF (Median filter), AWF (adaptive Wiener filter) & AMWF (adaptive median Wiener filter) are used for denoising. Subsequently a threshold based denoising algorithm has been developed using hard thresholding technique, then various filtering and thresholding techniques are applied simultaneously. Performance is measured with respect of PSNR (Peak Signal to Noise Ratio), MSE (Mean Square Error) and visual perception. Proposed method gives excellent agreement of results for any type of noises.