融合MTLBP和2DPCA算法的人脸识别技术研究  

Research on face recognition technology integrating mtlbp and 2DPCA algorithm

作　　者：单祖辉 SHAN Zu-hui(College of intelligence and information engineering,West Yunnan University,Lincang 677000,Yunnan,China)

机构地区：[1]滇西科技师范学院智能与信息工程学院,云南临沧677000

出　　处：《贵阳学院学报（自然科学版）》2022年第2期75-79,共5页Journal of Guiyang University：Natural Sciences

摘　　要：人脸识别的准确率主要取决于人脸特征提取算法对人脸特征的提取效果。传统的人脸特征提取算法具有单一性和不具备针对性的问题。因此,研究结合尺度不变特征变换(Scale-invariant feature transform,SIFT)算法和多阈值LBP算子(Mul-threshold LBP,MTLBP)对人脸局部特征进行提取,利用二维主成分分析法(2 dimensional principal component analysis,2DPCA)提取人脸全局特征,进而构建出人脸识别模型。测试结果表明,人脸识别模型的识别准确率达到了92.6%。因此,研究构建的人脸识别模型能够提高人脸识别的效率和准确率,为各行各业的身份识别与鉴定提供了新的途径,为人们的日常生活提供便利。the accuracy offace recognition mainly depends on the effect of face feature extraction algorithm on face feature extraction.The traditional face feature extraction algorithm has the problems of singleness and lack of pertinence.Therefore,it is studied to extract the local features of the face by combining the scale invariant feature transform(SIFT)algorithm and the mul threshold LBP Operator(mtlbp),The global features of human face are extracted by two-dimensional principal component analysis(2DPCA),and then the face recognition model is constructed.The test results show that the recognition accuracy of the face recognition model reaches 92.6%.Therefore,the face recognition model can improve the efficiency and accuracy of face recognition,provide a new way for identity recognition and identification in all walks of life,and provide convenience for people's daily life.

关 键 词：MTLBP 2DPCA算法 人脸识别技术 SIFT算法 特征提取 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]