计算机工程 ›› 2019, Vol. 45 ›› Issue (12): 232-236.doi: 10.19678/j.issn.1000-3428.0053108

• 图形图像处理 • 上一篇    下一篇

基于稀疏化双向二维主成分分析的人脸识别

张裕平, 龚晓峰, 雒瑞森   

  1. 四川大学 电气工程学院, 成都 610065
  • 收稿日期:2018-11-09 修回日期:2019-01-10 发布日期:2019-01-23
  • 作者简介:张裕平(1994-),女,硕士研究生,主研方向为图像处理、机器学习;龚晓峰,教授、博士;雒瑞森,讲师、博士。
  • 基金项目:
    中国博士后科学基金(2017M612958)。

Face Recognition Based on Sparse Two-Direction Two-Dimensional Principle Component Analysis

ZHANG Yuping, GONG Xiaofeng, LUO Ruisen   

  1. College of Electrical Engineering, Sichuan University, Chengdu 610065, China
  • Received:2018-11-09 Revised:2019-01-10 Published:2019-01-23

摘要: 双向二维主成分分析((2D)2PCA)易受异常值影响,鲁棒性差,且所提取的特征向量是非稀疏的。针对上述不足,提出基于L1范数的稀疏双向二维主成分分析方法(2D)2PCA-L1S。在(2D)2PCA目标函数中加入L1范数约束,以提高算法的抗干扰能力,同时引入弹性网约束,通过Lasso与Ridge惩罚函数实现稀疏性。在Feret和Yale数据库中进行基于最近邻的人脸分类、人脸重构和基于粒子群优化SVM参数的人脸识别实验,结果表明,相较于2DPCA、(2D)2PCA、(2D)2PCA-L1等主成分分析方法,该方法能准确提取人脸主要信息,人脸识别和人脸重构效果较好。

关键词: 双向二维主成分分析, 稀疏化, 粒子群优化, 支持向量机, 人脸识别

Abstract: Two-Direction Two-Dimensional Principle Component Analysis((2D)2PCA) is an improved method of Principle Component Analysis(PCA) in the two-dimensional space.However,just like PCA,the (2D)2PCA is susceptible to abnormal values,its robustness is weak and the extracted feature vectors are non-sparse.So this paper proposes a sparse (2D)2PCA method based on L1 norm,the (2D)2PCA-L1S,to tackle these problems.First,this paper adds the L1 norm constraint to the (2D)2PCA objective function to improve the anti-interference ability.Then,this paper introduces the elastic network constraint into the objective function,so the sparsity is realized by the Lasso and Ridge penalty functions.The following experiments are carried out on the Feret and Yale datasets:the face classification and face reconstruction based on nearest neighbors,and the face recognition based on Particle Swarm Optimization(PSO)-SVM Parameters.The results show that compared with other PCA method,such as the 2DPCA,(2D)2PCA and (2D)2PCA-L1,the proposed method can accurately extract the main face information,and it has better effects on face recognition and face reconstruction.

Key words: Two-Direction Two-Dimensional Principle Component Analysis((2D)2PCA), sparsity, Particle Swarm Optimization(PSO), Support Vector Machine(SVM), face recognition

中图分类号: