摘要
采用2DPCA方法提取人脸图像的特征值,通过RBF神经网络进行训练和识别,提出一种基于2DPCA和RBF神经网络的人脸识别方法,并将此方法应用于ORL人脸库。实验结果表明,该方法不仅具有较好的人脸图像识别能力,而且能明显缩短识别算法的运行时间。
A new approach for human face recognition based on 2DPCA and RBF neutral network is proposed.This method is tested on the standard ORL face database.Experimental results demonstrate that the combined method based on 2DPCA and RBFNN can achieve a better performance in recognition accuracy than other method and this method cost less computational time.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第34期200-203,共4页
Computer Engineering and Applications
基金
山西省自然科学基金(the Natural Science Foundation of Shanxi Province of China under Grant No.2006011028。
关键词
二维主成分分析
RBF神经网络
人脸识别
two-dimensional principal component analysis
RBF neutral network
face recognition
