An Efficient Image Denoising Method based on Bilateral filter Model and Neighshrink SURE

Abstract

In all the instances of image acquisition, transmission and storage, the unwanted noise gets into the information content of the image and thereby introduces an unpleasant visual quality to the observer. So the field of image processing has produced a lot of image denoising algorithms and techniques to improve the visual quality of the image. Since noise cannot be reduced to zero practically, the need for faithful and efficient denoising techniques to produce almost noiseless images demands a systematic research work in the field of denoising methods. The denoising process using a bilateral filter even though produces improvement in the image quality, it does not show consistency when the noise level is high and also the peak signal to noise ratio (PSNR) and Image quality Index (IQI) do not show any improvement. This paper proposes an improved algorithm that incorporates the function of bilateral filter model and wavelet thresholding using Neighshrink SURE method. The results show significant improvement in both PSNR and IQI values with respect to the four standard test images under various noise conditions.