A study on FDNN applying the hybrid fuzzy membership function and the genetic algorithm
Oh Sung Byun, Soo Hyung Cho, Chun Hwa Seo, Sung-Ryong Moon · 2003
We apply the hybrid method fuzzy membership function in order to obtain the result that is likely to the original image, and the generic algorithm in order to find the optimal image to the FDNN. If some of the data is input, it is selected as a local winner to find a basis image of the largest similarity. We realize the hierarchical FDNN obtaining the last output value selection to a global winner among a local winner. In this paper the noise is removed from an image using FDNN to which is applied both the hybrid fuzzy membership function and the genetic algorithm, also the superiority of the proposed algorithm to the conventional FDNN is found. As a result of the comparison by the MSE for each image, we show the superiority of the FDNN to which is applied both the hybrid fuzzy membership function and the genetic algorithm.