南京理工大学学报（自然科学版）2017，Vol.41Issue(1)：74-79,6.DOI:10.14177/j.cnki.32-1397n.2017.41.01.010
曲波变换和独立分量分析的人脸识别
Face recognition based on curvelet transform and independentcomponent analysis
摘要
Abstract
In order to obtain high recognition rate and accelerate the speed of face recognition,a face recognition algorithm is proposed by combining curvelet transform and independent component analysis.Firstly,face images are processed by curvelet transform and get curvelet coefficients in scale and direction.The obtained curvelet coefficients are weighted and fused and then independent component analysis is used to select the features which have important contributions,reducing the feature dimension to accelerate the recognition speed of face classifier.Finally,least square support vector machine is used to establish face recognition classifier,and the classical face database is used to test the performance of face recognition.The experimental results show that the average recognition rate of the proposed algorithm is more than 95%,and the average recognition time can meet the requirements of face recognition.关键词
人脸识别/特征提取/曲波变换/独立分量分析/小波变换Key words
face recognition/feature extraction/curvelet transform/independent component analysis/wavelet transform分类
信息技术与安全科学引用本文复制引用
张琳梅,张雪峰..曲波变换和独立分量分析的人脸识别[J].南京理工大学学报（自然科学版）,2017,41(1):74-79,6.基金项目
河南省教育厅科学技术研究重点项目(16A520091) （16A520091）
