Image Segmentation Using Thresholding by Local Fuzzy Entropy-Based Competitive Fuzzy Edge Detection

In this paper, we present a new thresholding approach by local fuzzy entropy based competitive fuzzy edge detection for image segmentation which assign appropriate threshold effectively and reduces the affects of noise in edge detection and segmentation. In this algorithm first, edges detected by fuzzy logic and competitive rules, then there would be improvement in quality obtained edges by fuzzy entropy. The end by the information of received edges suitable threshold fined for image segmentation and then we will segment the images properly. The in novation, of this paper is the improvement in the edges of image in competitive fuzzy edge detection which it would be usable in the image segmentation. The results show that the quality of segmentation which is based on the suggested approach for the white Gaussian noise images is better than local entropy algorithm.