Research On Face Recognition Method Based On Neural Networks

Biometric identification technology utilizes the body’s inherent in some of the biometric, it has high security, reliability, and availability features, more and more people’s attention. Automatic face recognition biometric technology as an important part, in recent decades has been widespread concern and research, has become a computer vision, pattern recognition research focus areas. Face recognition in the field of public safety, document authentication, access control systems, video surveillance, etc. have a wide range of applicationsThis is the human face algorithm feature extraction and face recognition based on a database on the following key elements:(1)Read a lot of literature on face recognition technology were reviewed.(2)Principal component analysis K-L transform to achieve facial feature extraction, using principal component analysis to achieve the purpose of dimensionality reduction based on Through experimental results verify that this method is a fast and stable method for facial expressions, gestures and other changes with good tolerance.(3)In this paper, the image acquisition system design face recognition, face detection and segmentation, feature extraction and selection and recognition of four modules, and each module and functional analysis of the data processing mode. presented research on improved BP neural network algorithm for face recognition. BP algorithm uses an improved classification of the classifier. Face and minimum distance classification by comparing the same characteristics, the algorithm has obtained good results in the recognition rate and stability.