计算机应用 ›› 2005, Vol. 25 ›› Issue (12): 2855-2857.

• 图形图像处理 • 上一篇    下一篇

类肤色背景中提取人脸轮廓的方法研究

刘睿1;王晓东2   

  1. 1.西安邮电学院电子与信息工程系; 2.西安交通大学生命科学与技术学院
  • 收稿日期:1900-01-01 修回日期:1900-01-01 发布日期:2005-12-01 出版日期:2005-12-01

Studies on face contour extraction from approximate skin color backgrounds

LIU Rui<sup>1</sup>,WANG Xiao-dong<sup>2</sup>   

  1. 1.Department of Electronic and Information Engineering,Xi’an University of Post and Telecommunications,Xi’an Shaanxi 710061,China;2.School of Life Science and Technology,Xi’an Jiaotong University,Xi’an Shaanxi 710049,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2005-12-01 Published:2005-12-01

摘要: 提出了一个基于肤色并融合多种信息的人脸轮廓提取方法。首先在TSL色彩空间求取肤色概率图,选取种子点,然后利用多源信息进行区域生长,提取出人脸轮廓;为克服区域生长计算量大的缺点,采用了变分辨率图像金字塔策略。经实例验证,该算法能够快速准确地从类肤色背景中较好地提取出人脸轮廓,且具有较高的抗噪性和应用适应性。

关键词: 人脸轮廓提取, 肤色概率图, 信息融合, 图像金字塔处理

Abstract: A new face contour extraction method was proposed by useing skin color and multiple information fusion.The method computed skin probability image based on TSL color space,then selected seed point,finally used multiple information to carry out region growing,extracted face contour.The method employed image pyramid process to avoid heavy computing burdens caused by region growth.Experiment results indicated that the proposed algorithm can extract face contour from approximate skin color backgrounds efficiently and accurately,and bears noise endurance and application adaptability.

Key words: face contour extraction, skin probability image, information fusion, image pyramid processing

中图分类号: