English

应用科学学报 ›› 2006, Vol. 24 ›› Issue (6): 555-559.

• 论文 • 上一篇    下一篇

一种改进的对齐算法及其在面部特征定位中的应用

冀贞海1,2, 邹采荣2, 赵力2   

  1. 1. 东南大学学习科学研究中心, 江苏南京 210096;
    2. 东南大学无线电工程系, 江苏南京 210096
  • 收稿日期:2005-10-18 修回日期:2005-12-02 出版日期:2006-11-30 发布日期:2006-11-30
  • 作者简介:冀贞海,博士生,研究方向:图像处理和模式识别,E-mail:jizhenhai@seu.edu.cn;邹采荣,教授,博导,研究方向:多维数字信号处理,E-mail:cairong@seu.edu.cn;赵力,教授,博导,研究方向:语音信号处理和情感信息处理,E-mail:zhaoli@seu.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(60503023)

An Improved Alignment Method and Its Application to the Localization of Facial Features

JI Zhen-hai1,2, ZOU Cai-rong2, ZHAO Li2   

  1. 1. Research Center of Learning Science, Southeast University, Nanjing 210096, China;
    2. Department of Radio Engineering, Southeast University, Nanjing 210096, China
  • Received:2005-10-18 Revised:2005-12-02 Online:2006-11-30 Published:2006-11-30

PDF

90

摘要： 在主动形状模型基础上,借助于核的技巧,对其前期所需的广义对齐算法进行了研究改进,使之能够在后续的主分量分析中更迅速有效地捕捉不同的面部形状变化,进而达到快速定位面部关键特征点的目的.实验结果表明,改进的算法在主分量分析之后能够更好地表达不同脸形变化的模式,而且在定位面部关键特征点上精度更高,收敛速度也有明显改善.

关键词: 主动形状模型, 广义对齐算法, 人脸识别, 面部特征定位

Abstract: This paper presents a new method to align the facial feature points, which is developed from the generalized procrustes analysis, and improved by adding nonlinear information of facial points and using the active shape model and the kernel trick.Experiments show that the improved method has better performances in obtaining more detailed modes of face shape and better accuracy in searching unknown target image points.Meanwhile less time is spent for the search.

Key words: active shape model, face recognition, face alignment, generalized procrustes analysis

中图分类号: