摘要
提出了一种基于主成分分析的人脸特征定位及识别的方法 .利用剩余误差能量作为准则进行人脸特征定位 ,并对人脸图像提取主成分特征进行识别 ,经实验表明该方法能处理人脸图像变化较大的情况 ,具有计算量较小 ,定位精度高 ,识别率高的优点 .
A new PCA based facial feature location and face recognition algorithm is investigated.A facial feature location algorithm is developed according to the residual reconstruct error energy.Extract the principal component features for face recognition.Experimental results show that this method can locate facial features accurately and gain high recognition rate even when the face database has large variability.
出处
《云南大学学报(自然科学版)》
CAS
CSCD
2001年第3期181-186,共6页
Journal of Yunnan University(Natural Sciences Edition)