| Abstract: | This paper introduces a new face recognition method based on fuzzy fusion and KNN classifier for face identification. The proposed method comprises two steps. In the first stage, the membership degrees of the information provided by different attribute images are estimated using fuzzy c-means algorithm and then combined by the fuzzy fusion technique, in order to improve the information quality and to get a more reliable and accurate segmentation results, whereas the fuzzy combination rules are used to combine different attribute images in order to obtain the best human face recognition. In this proposed method, the purpose of segmentation is to isolate the face of the human. In the second stage, the K-nearest Neighbors’ (KNN) algorithm is employed for classification task.
The True Success Rate of the proposed Human face recognition method is evaluated and a comparative study with existing methods is presented. Experimental results using 400 test images of 40 people show the superiority of combining different attribute images for human face recognition. Experimental results show that the algorithm identifies the human face with accuracy of 98.63%.
|