Design And Implementation Of The Face Recognition System

Face recognition is one of the biology recognition technologies. It is a particular application of pattern recognition in image filed. The researchers from various fields are paying more attention to face recognition for it is one of the most important components of biometrics. Face is easier to disturbed than fingerprint and iris, but the face recognition technology would develop widely. Face recognition has some advantages such as non-encroached, nature and high cost-effective ratio.Based on two-dimension face image database, this paper discusses the pre-processing method feature face extraction and recognition method for two-dimension face recognition. It designs and implements a simple face recognition system based on these methods. This system is composed by image pre-processing, face detection and location, feature face extraction and recognition. In image pre-processing part, a pre-processing image is obtained by light compensation, grayscale change, Gauss smoothness process, binarization. In face detection and location part, it finds human face and the position from inputting picture. Then it cuts apart human face from the background. In feature face extraction and recognition part. It adopts principle components analysis (PCA) recognition method. It reads the face database and obtains a set of feature faces by principle components analysis. These feature faces form the face feature subspace and they are saved in database. In recognition stage, the face image that should be recognized is mapped to face feature subspace. Minimum distance classifier is used to recognize the face image.Face recognition is widely used in the fields of personal identification, gateways to limited access areas, bank ATM and social security, nation security etc.