计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (6): 169-170.

• 图形、图像、模式识别 • 上一篇    下一篇

傅里叶变换的多视角步态识别

刘砚秋1,2,王 旭1,王玉梅2,吕 红2   

  1. 1.东北大学 信息学院,沈阳 110004
    2.辽宁大学 信息学院,沈阳 110036
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2012-02-21 发布日期:2012-02-21

Multi-view gait recognition using Fourier transform

LIU Yanqiu1,2, WANG Xu1, WANG Yumei2, LV Hong2   

  1. 1.College of Information, Northeastern University, Shenyang 110004, China
    2.College of Information, Liaoning University, Shenyang 110036, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-02-21 Published:2012-02-21

摘要: 步态识别作为一种全新的生物特征识别技术,通过人走路的姿势实现对个人身份的识别和认证。步态能量图将一个周期的步态组合在一起,增强了各帧的相关性,减少了噪声的干扰。对步态能量图进行傅里叶变换,利用傅里叶变换的低频分量对多个视角的步态进行识别。在CASIA数据库中进行实验,结果表明算法简单快速,取得了较好的识别效果。

Abstract: Gait recognition is a new topic in the biometric technology. Human authenticates or identifies the individuals by the way they walk. Gait energy image combines frames of one gait cycle together to enhance the relevance among them, to reduce the noise interference. The Fourier transform is carried out on the gait energy image. Their low frequency components are adopted to perform multi-view gait identification. The method is applied to CASIA database. The experimental results demonstrate that the proposed method is fast and effective.