PCA人脸识别算法的优化

Improvement of Human Face Recognition Algorithm Based on PCA

生物识别技术的快速发展,使得人脸识别技术成为研究热点。主成分分析算法是人脸识别技术中被使用最多的算法之一,它因识别速度快,识别率高被广泛认可。针对传统的PCA算法由于外界的干扰因素会影响它的识别率。基于提高PCA算法的识别率和抗干扰特性,通过对原始图片的去噪声,直方图均衡化,归一化三种预处理方法,同时结合Adaboost算法,能一定程度地提高识别的成功率和算法的抗干扰性。实验结果表明优化后的PCA算法的识别率相对提高了10%,识别速率比原始算法提高了30%。

With the rapid development of biometrics,human face recognition technology has become a research hotspot.PCA algorithm is one of the most adopted algorithms for human face recognition,and has gained wide recognition for its high identification speed and accuracy.Conversely,with external interference factors,traditional PCA algorithm is affected in its identification efficiency.The results of the present experiment indicate that identification accuracy and anti-jamming of algorithm are improved through combining with Adaboost algorithm and adopting three pretreatment methods in original image processing：de-noise processing,histogram equalization and histogram normalization.The experiment result shows that the identification of improved PCA has increased 10% than traditional PCA algorithm,and identification speed has increased30%.