基于遗传算法的支持向量机人脸识别技术

人脸识别身份验证技术是目前一个非常活跃的研究课题.文章针对人脸识别系统涉及到的人脸特征提取、识别验证等环节,利用K-L变换首先对人脸图像进行特征参数提取,并提出用支持向量机与遗传算法相结合的新型算法进行分类识别,利用遗传算法自动选择最优的核函数参数,将以上方法相结合的新型人脸识别方法的实验结果表明,该方法所得参数确定的SVM具有较优的识别率,其整体性能优良.

Face Recognition Based on Support Vector Machines

Face recognition is a very activity research,In allusion to some problems of face recognition,such as getting the representation basis of face image set and recognition.Karhunen-Loeve transform is employed to get the representation basis of face image set,bring forward support vector machines based on GA are used to classify,GA is used to select excellent kernel function,the results of experiment show Face recognition based on support vector and GA are effective.