多彩色空间相关分析的人脸识别算法

A fast color face recognition algorithm

人脸识别是当前人工智能和模式识别的研究热点,得到了广泛的关注。基于对不同色彩空间数据的分析,论文提出了多彩色空间典型相关分析的人脸识别方法。文中对2维的Contourlet变换特性进行了分析和讨论,利用Contourlet的多尺度,方向性和各向异性等特点,提出了一种基于Contourlet变换的彩色人脸识别算法。算法对原图进行Contourlet分解,对分解得到的低频和高频图像进行cca分析。典型相关分析是一种有效的分析方法,其实际应用十分广泛。低频系数反映图像的轮廓信息,高频系数反映图像的细节信息,使用cca充分利用不同频率的信息,使不同色彩空间的不同分辨率图形的相关性达到最大,得到投影系数,最后,采用决策级最近邻分类器完成人脸识别。在对彩色人脸数据库AR的识别实验中,该算法识别率达到98%以上,与传统算法相比,该算法不仅既有良好的识别结果,而且具有很快的运算速度。

Face recognition is an active research area in the artificial intelligence,which has aroused great concern.Multiple color space canonical correlation analysis is proposed based on different color spaces analysis.This paper analyses and discusses the characteristics of Contourlet transform,and,by using the contourlet＇s advantage of multiscale,directionality and anisotropy,proposes a novel color face recognition algorithm.First,the source images are transformed into contourlet domain to get the LP and HP images.Then,canonical correlation analysis（CCA） is used to recognize the face.CCA is an efficient projection operator,which can analyze different color spaces to get the biggest correlation.Finally,nearest neighbor classifier is selected to perform face recognition.Experimental results on color AR face database show that the proposed algorithm,which achieves recognition accuracy of above 98%,is more effective and faster than the traditional method.