오류 역전파 알고리즘을 이용한 얼굴인식

Face Recognition using Error Back Propagation Algorithm

In this paper, we propose a new method for facial feature extraction using mosaic image and face recognition algorithm, The preprocessing steps of the face image are as follows. 1. binary image, 2. edge detection of binary image, 3. smoothing of contour image, 4. mosaic of smoothing image, 5. feature (eye, nose, mouth) extraction using mosaic image, 6. rotation of the tilted image in accordance with slope information. In the recognition step, we use the back propagation algorithm of the neural networks and in the learning step, to reduce the computation time, we vary learning and momentum rates. Experimental results show 100% recognition rate for twenty men and ten women face images.