Based On The Sex Of The Face Image Recognition Research

As the development of the study of face recognition recent years, there hasbeen an increasing development of face recognition study based on thebiological features. Compared with other biological features, face features aremore natural and more convenient, and they are non-contact. Face features havea wide application prospects in authentications, security monitoring systems,pattern recognitions and human-computer interactions. Face recognition basedon sex characters is one of the two typical pattern recognition problems, it has aimportant application prospects in authentications, human-computer interfaces,video retrievals and robot visions.Some studies have been done on the difficulty to the extract the humansecondary sex characteristics, the low recognition rate and robustness. In orderto remove the influence from noise, illumination and camera angles, Gaussfiltering, gray and geometric normalization have been done to preprocess theimages.This paper introduces a new template for face image cutting.In order toimprove the stability and robustness of the recognition, we also propose agender recognition method based on support vector machine and compared withthe method based on neural network. The experimental results show that theproposed method improves the ability of error-resilient, recognition rate of theclassifier and robustness due to less sensitivity of extraction of the facial featurepoints. The complexity of the proposed method with high recognition efficiencyis lower than the gender recognition method based on neural network.