An FCM-Based Image De-Noising with Spatial Statistics Pilot Study

Image de-noising is an important scheme that makes image visually prominent and obtains enough useful information to produce a clear image. Many applications have been developed for effective noise suppression good quality. This study assumed residual consisted of with edges produced by subtracting the original low-pass-filter-smoothed The Moran statistics were then used measure variation in spatial images we this as feature data input into Fuzzy C-means (FCM) algorithm. Three clusters pre-assumed FCM work: they heavy, medium, less noisy areas. rates each position partially belonged cluster determined using membership function. Each pixel was processing linear combination product three de-noised functions same position. Average filters different windows Gaussian filter priori applied create versions. results showed worked better than non-adaptive smoothing. scheme‘s performance evaluated compared bilateral non-local means (NLM) peak signal ratio (PSNR) structure similarity index (SSIM). pilot study. Further future studies are needed on optimized number smoother versions combination.