A fuzzy central cluster neural classifier

Pattern classification is a critical stage in an image recognition system. In this paper, we incorporate the concepts of fuzzy sets into the recognition system, and develop an efficient algorithm based on fuzzy classification decision criterion to classify features of images with a high degree of success. A fuzzy central cluster neural network classifier is proposed, which takes advantages of the neural network´s high computation rate provided by the massive parallelism, and fuzzy neural network classifiers can be trained by both supervised and unsupervised techniques