Paper Title
An Enhancement Of Fuzzy-Based Image Denoising For Gaussian Noise

In the process of image acquisition and transmission, it is hard to avoid the interference of different types of noises. Image denoising is important because various kinds of images are used as source of information for interpretation and analysis in many applications. Among the various types of noise, impulse noise and Gaussian noise are the most common. Most of the existing fuzzy based filters are keen on impulse noise because Gaussian noise is difficult to distinguish local variation due to noise and due to image structure such as edge, line or curve. In this paper, fuzzy based Gaussian filter is proposed to distinguish between noises and edges. In this system, fuzzy derivatives values for eight directions within a kernel are computed to detect the target pixel is whether noise or edge using a thresholding value. For the noise pixel, the fuzzy based filtering is operated to remove Gaussian noise by using these fuzzy derivatives values, positive and negative truthiness. This filtering process is iteratively according to the remaining of the noise level. Finally, the performance of the proposed filtering technique is compared with the Non-local means filter by calculating the PSNR value. Index Terms- fuzzy based filtering, membership function, fuzzy derivatives, fuzzy sets, derivative neighborhoods