摘要
将基于主成分分析的特征脸人脸识别方法进行改进,以提高人脸识别率。首先利用主成分分析法提取人脸图像的特征脸,然后经图像重构得到二阶特征脸,最后将两种特征脸组合,构造组合特征,用三阶近邻法进行识别。在ORL人脸数据库上的试验结果表明,组合特征脸法用于人脸识别有较高的可行性和较好的稳定性,且在识别率上优于特征脸方法,准确率达到93.8%。
In order to increase face recognition rate,we improve the eigenface method of principal component analysis. Firstly principal component analysis method is used to extract the eigenface of initial images, and then we get the second-order eigenface images by rebuilding images.Lastly, we combine the two kinds of eigenface vectors of everyone into a longer vector, use third-order neighbor classifier to classify and identify the test images. To verify the efficient of the method, we test this method on ORL face database and experiment result shows that eigenface combination method for face recognition have high feasibility and good stability, and the recognition rate is better than eigenface method, the accuracy rate reach to 93.8%.
出处
《商洛学院学报》
2009年第4期40-43,共4页
Journal of Shangluo University
基金
陕西省教育厅专项科研基金项目(05JK255)
关键词
主成分分析
特征脸
二阶特征脸
人脸识别
principal component analysis
eigenface
second-order eigenface
face recognition
