基于独立分量分析和径向基网络的人脸识别方法

Face recognition based on independent component analysis and RBF network

人脸识别是当前人工智能和模式识别的研究热点。基于小波分解和定点独立分量分析,提出了一种新的准正面人脸径向基网络识别算法。二维的小波分解具有对表情变化不敏感的特点,可以很好地压缩和表征人脸图像的特征,定点独立分量分析是一种基于高阶统计信息提取特征的方法,克服了一般ICA收敛慢的缺点;径向基网络作为分类器具有很高的推广性能,有利于大容量样本的分类。在对人脸库ORL和YEL的识别实验中,该算法的识别率达到98%以上,与传统算法相比,识别速度和识别率都明显提高。

Face recognition is an active research area in the artificial intelligence.A new face recognition algorithm using the RBF network is proposed based on wavelet analysis and fixed point ICA. Since wavelet analysis is insensitive to changes in expression,it can effectively express the principal features of the face image by compressing data and corresponding wavelet coefficients.The eigenface of face image can be obtained by the fixed point ICA which is faster than the standard ICA. The RBF network with high generalization is a good classifier,especially for larger samples.Experiment results on ORL and YEL face show that the proposed algorithm,which achieves recognition accuracy of above 98% is more effective and faster than the traditional method.