计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (21): 197-198.

• 工程与应用 • 上一篇    下一篇

人脸关键特征点自动标定研究

张志刚1,周明全2,耿国华1   

  1. 1.西北大学 信息科学与技术学院,西安 710069
    2.北京师范大学 信息科学与技术学院,北京 100875
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-07-21 发布日期:2007-07-21
  • 通讯作者: 张志刚

Research on automatic localization of facial key features

ZHANG Zhi-gang1,ZHOU Ming-quan2,GENG Guo-hua1   

  1. 1.College of Information Science and Technology,Northwest University,Xi’an 710069,China
    2.College of Information Science and Technology,Beijing Normal University,Beijing 100875,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-21 Published:2007-07-21
  • Contact: ZHANG Zhi-gang

摘要: 人面部特征点的自动标定是计算机辅助颅像重合技术的重要研究内容之一。在人脸分析技术的基础上,提出了一种眉心点和眼角点的自动定位方法。在得到人像双眼位置的前提下,据此确定眉毛和眼睛块;然后利用一种自适应的阈值分割法对特征块进行二值化处理,进而提取出眉心点和眼角点。该方法充分利用了人面部的特征知识,简便易行。实验结果表明算法有效,定位的准确度较高。

关键词: 特征标定, 眼角点, 眉心点, 颅像重合

Abstract: Facial features location is an important research aspect in computer-aided photographic superimposition.Based on technique in human face analyzing,an automatic location method for eyebrows centers and eyes corners is proposed.First,the rectangle regions of eyebrows and eyes are determined according to eyes positions,then,based on an adaptive threshold segmentation strategy,a binarization processing for these regions is conducted to extract eyebrows centers and eyes corners.Making use of gray features knowledge of human faces,the method is simple and easy to be realized.Experimental results prove that this method is efficient and better accuracy can be obtained.

Key words: features localization, eyes corners, eyebrows centers, photographic superimposition