摘要
从人脸图像特征提取和分类器构建两方面分析了人脸识别系统设计的关键点，提出了以主成分分析技术和支持向量机技术相结合构建人脸识别系统的策略，同时在主成分分析技术的理论基础上提出了一种快速PCA算法。通过实验系统在ORL人脸库上的测试结果，分析了该系统的相关参数和特征向量维度的选取对系统识别率的影响，并得到了其最优解。同时，通过实验证明了所提出方法在小训练集下的识别率优于其它一般方法，其识别率比一般的人工神经网络法提高了7％～10％左右。
This paper analyzed the key points of face recognition system design by the extraction of face image feature and build of classifier , and it proposed a plot which built face recognition system by in-tegrating principal component analysis and support vector machine technology .And a fast PCA algorithm had been proposed on the basis of the theory of principal component analysis technology .This paper ana-lyzed the impact of the system parameters and dimensions of the selected feature vector on recognition rate of the experimental system and got the optimal solution on basis of the system test results on ORL face da -tabase .And experimental results have demonstrated that recognition rate of the method this paper proposed is superior to other general methods on the small training set , and its recognition rate increased about 7%to 10%higher than the average artificial neural network .
出处
《湖北民族学院学报（自然科学版）》
CAS
2015年第2期193-196,214,共5页
Journal of Hubei Minzu University(Natural Science Edition)
基金
国家自然科学基金项目(61263030
61463014)
湖北省自然科学基金资助项目(2014CFB612)
湖北民族学院博士启动基金(MY2014B018)
