计算机工程 ›› 2011, Vol. 37 ›› Issue (7): 228-230.doi: 10.3969/j.issn.1000-3428.2011.07.077

• 人工智能及识别技术 • 上一篇    下一篇

改进的模块2DPCA人脸识别算法

张 岩1,2,武玉强2   

  1. (1. 济宁学院物理与信息工程系,山东 曲阜 273155;2. 曲阜师范大学自动化研究所,山东 曲阜 273165)
  • 出版日期:2011-04-05 发布日期:2011-03-31
  • 作者简介:张 岩(1981-),男,讲师、硕士,主研方向:计算机视觉,模式识别;武玉强,教授、博士、博士生导师
  • 基金项目:
    教育部科学技术研究基金资助重点项目(208074);济宁学院科研基金资助项目(2009KJLX04)

Improved Face Recognition Algorithm of Modular 2DPCA

ZHANG Yan  1,2, WU Yu-qiang  2   

  1. (1. Department of Physics and Information Engineering, Jining University, Qufu 273155, China; 2. Research Institute of Automation, Qufu Normal University, Qufu 273165, China)
  • Online:2011-04-05 Published:2011-03-31

摘要: 提出一种改进的模块2DPCA人脸识别算法,即基于子距离的模块2DPCA人脸识别算法。该算法对图像进行分块,对每一子块独立地利用2DPCA进行处理,求出测试样本子块与训练样本对应子块间的子距离,将所有子距离相加得到测试样本与训练样本的距离,用最近距离分类器分类。在ORL人脸库上的实验结果表明,该算法在识别性能上优于普通的模块2DPCA算法和修正的模块2DPCA算法。

关键词: 二维主成分分析, 子距离, 模块二维主成分分析, 特征提取, 人脸识别

Abstract: An improved modular Two-Dimensional Principal Component Analysis(2DPCA), modular 2DPCA based on sub-distance, is proposed. The original images are divided into sub-images in proposed algorithm. Each kind of sub-images at the same position is disposed by 2DPCA independently, and the sub-distance between the corresponding sub-images of the test sample and the train sample can be given. The distance between the test sample and the train sample can be calculated by adding all these distances between the sub-images together, and the nearest distance classification is used to distinguish each face. Experimental results on ORL face database indicate that the improved modular 2DPCA is obviously superior to that of general modular 2DPCA and amendatory modular 2DPCA.

Key words: Two-Dimensional Principal Component Analysis(2DPCA), sub-distance, modular 2DPCA, feature extraction , face recognition

中图分类号: