To search, Click below search items.


All Published Papers Search Service


Novel Image Denoising with Method Noise Thresholding using Statistical Nearest Neighbor Filter


Panchaxri, Basavaraj N Jagadale, and Vijayalaxmi Hegde


Vol. 22  No. 4  pp. 551-558


The goal of this research is to offer a novel denoising approach for generating a denoised image(s) having fewer artifacts and improved efficiency at higher noise levels. Statistical Nearest Neighbor with Method Noise Thresholding (SNNMNT) is a new filter that improves the quality of the final image. Aside from direct filtering, the noisy image and pre-filtered image would be used to generate method noise in this work. This method employs Neighshrinksure estimation to make wavelet filtering. These computed values are then added to the prefiltered image to produce the desired resulting image. The noisy image is denoised without losing the original image details for precise analysis and extraction of image features. The benchmark images denoised with standard deviation (?=10) using bior6.8 wavelet when filtered using earlier filters such as Gaussian Bilateral Filter with Method Noise Thresholding and Statistical Nearest Neighbor show improvement resultant image quality in terms of PSNR and ISSN as compared to the proposed filtering technique. The proposed filter produces higher PSNR and ISSN values (PSNR=34.49 and SSIM=09997). This functional filter proposed in this work provides an improvement in image quality parameters of the image when compared with the earlier methods. The pictorial analysis was also carried out in the present work.


Method Noise, Wavelet Thresholding, Wiener Filter, and Statistical Nearest Neighbor.