Improvement of Face Recognition Algorithm Based on Neural Network

Face recognition is constructed based on facial feature extraction and classification, the facial feature extraction is taken in the face region according to different parts of the characteristic, it is prone to confuse the issue, this paper proposes a face recognition algorithm based on neural network. In order to reduce the interference of background noise. Reduce the post feature point location and recognition of the complexity of the binarization image denoising method for face image denoising, image noise reduction of output for feature extraction, extract the face value of the peak and valley of two-dimensional features, can get the edge face regions around the border, BP neural network classifier method is used for information facial features batch read, differences and classify facial features are constructed to achieve accurate face recognition. The simulation results show that the calculation face recognition method can accurately extract facial features, the accuracy of face recognition is better, the anti-interference ability is stronger, and the operation speed is higher, it can effectively obtain fast and efficient face recognition.