基于Haar-like T和LBP特征的人脸识别方法  

Face Recognition Method Based on Haar-like T and LBP Features

作　　者：胡宇晨 李秋生[1,2] HU Yuchen;LI Qiusheng(Research Center of Intelligent Control Engineering Technology,Gannan Normal University,Ganzhou 341000,China;School of Physics and Electronic Information,Gannan Normal University,Ganzhou 341000,China)

机构地区：[1]赣南师范大学智能控制工程技术研究中心,赣州341000 [2]赣南师范大学物理与电子信息学院,赣州341000

出　　处：《自动化与仪表》2023年第10期52-56,61,共6页Automation & Instrumentation

基　　金：江西省研究生创新专项资金资助项目(YC2022-s939)。

摘　　要：人脸识别技术是一种可靠的身份识别技术,是当今研究的热点课题之一。该文将Haar-like T和LBP特征相结合的算法进行人脸识别的研究。首先将Haar-like T方法与Haar-like方法进行人脸检测率的对比,再分别用该文算法与CNN、LBP等其他算法相比较。实验结果表明,该算法与基于Haar-Like T特征的人脸识别方法和一些主流方法相比较,增加了改进LBP方法对于图像的处理,使得该方法对光照具有较强的鲁棒性,对光照不敏感,同时对人脸的识别时间更短,在Yale和ORL人脸库中有着较好的识别效果,具备良好的可行性。Face recognition technology is a reliable identification technology and one of the hot topics in current research.This article combines Haar-like T and LBP features for facial recognition research.Firstly,compare the facial detection rates of Haar-like T method and Haar-like method,and then compare the proposed algorithm with other algorithms such as CNN and LBP.The experimental results show that compared with facial recognition methods based on Haar-like T features and some mainstream methods,this algorithm adds an improved LBP method for image processing,making it highly robust to lighting,insensitive to lighting,and has a shorter recognition time for faces.It has good recognition performance in Yale and ORL facial databases and has good feasibility.

关 键 词：人脸识别 卷积神经网络 Haar-like T LBP 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]