Face recognition based on multi features extractors

Face recognition is the core application in the biometrie technology area. It is widely used in the advanced application of robotics and computer vision. Raising commercial of face recognition and low enforcement makes a request of it increases in the last decade. In this paper, we present state of the art of in face recognition technologies by focusing on some traditional issues and some techniques to treats these problems. The advantage of suggested algorithm is to solve the popular issues in face recognition such as light conditions and environmental factors that lead to the low-performance. In proposed algorithm, groups of edge detection filters (Sobel, Prewitt, and Roberts) were used to extract edges of the faces in images. First derivative edge detection filters were performed to get best features of data set. moreover, using edge detection process that used by first order filter is to reduce data as much as possible by removing image background. The new method used as feature extractor addition to traditional PCA. Gathering features by using slope method and PCA is to find the optimal faces vectors as the inputs to the classifier (NNMLP neural network). Results have revealed acceptable correct classification. As data test set we used BIO-ID data base in the proposed system.