华东师范大学学报(自然科学版)

• 计算机科学 • 上一篇    下一篇

基于LBP和Gabor混合特征的近红外人脸识别

赵骥, 童卫青   

  1. 华东师范大学 计算机系, 上海200241
  • 收稿日期:2015-06-12 出版日期:2016-07-25 发布日期:2016-09-29
  • 通讯作者: 童卫青,男,副教授,硕士生导师,研究方向为图像处理、模式识别、机器学习. Email: wqtong@cs.ecnu.edu.cn.

Face recognition using near infrared images based on LBP and Gabor hybrid feature

ZHAO Ji,TONG Wei-qing   

  1. Department of Computer Science and Technology, East China Normal University, Shanghai 200241, China
  • Received:2015-06-12 Online:2016-07-25 Published:2016-09-29

摘要:

综合局部二值模式(Local Binary Patterns,LBP)和Gabor函数特征的优点,并结合余弦相似度和主元分析(Principal Component Analysis,PCA)方法,提出特征混合的人脸识别算法并研究4种不同的LBP和Gabor特征混合的方法.在871人的近红外人脸图像库上的实验表明,基于LBP和Gabor混合特征的人脸识别方法能获得较高的人脸识别正确率和较低的误识率.

关键词: 人脸识别, LBP, Gabor, 支持向量机

Abstract:

To take advantages of local binary patterns (LBP) and Gabor, we proposed an algorithm for face recognition based on hybrid features, using cosine similarity and principal component analysis (PCA). Four kinds of LBP and Gabor hybrid feature methods were studied and experimental results on 871 people’s nearinfrared face illustrated that the proposed method attained higher correct rate and lower false accept rate (FAR).

Key words: face recognition, LBP, Gabor, support vector machine (SVM)