计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (26): 245-248.

• 工程与应用 • 上一篇    

基于小波矩的面部特征匹配定位方法

章品正1,徐琴珍2,王 征1   

  1. 1.东南大学 计算机科学与工程系 影像科学与技术实验室,南京 211189
    2.东南大学 无线电工程系 信号与信息处理实验室,南京 211189
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-09-11 发布日期:2007-09-11
  • 通讯作者: 章品正

Wavelet moment based facial feature matching and detection method

ZHANG Pin-zheng1,XU Qin-zhen2,WANG Zheng1   

  1. 1.Lab of Image Science & Technology,College of Computer Science and Technology Engineering,Southeast University,Nanjing 211189,China
    2.Lab of Signal and Information Processing,Department of Radio Engineering,Southeast University,,Nanjing 211189,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-09-11 Published:2007-09-11
  • Contact: ZHANG Pin-zheng

摘要: 面部特征是实现面部表情分类与刻画面部表情强度的重要信息。提出了结合金字塔分解技术和小波矩的面部特征匹配定位方法。该方法通过小波矩实现图像信息的多尺度表征,而应用金字塔分解在金字塔图像各层之间传递信息并最终实现局部处理与全局处理之间的联系。实验结果表明,提出的方法可以在面部特征发生较大形变时依旧取得较好面部特征匹配定位结果。

关键词: 小波矩, 金字塔分解, 特征点匹配

Abstract: Facial features contain important information for facial expression analysis and its intensity estimation.In this paper,a new facial feature detection method,which based on pyramid decomposition and wavelet moment,is proposed.It provides multi-resolution representation of the original images and combines images’ globe and local information to detect the facial feature.Experimental results show that it can accomplish the facial feature detection even if facial feature’s deformation is large.

Key words: wavelet moment, pyramid decomposition, feature point matching