计算机工程 ›› 2012, Vol. 38 ›› Issue (14): 165-166.doi: 10.3969/j.issn.1000-3428.2012.14.049

• 人工智能及识别技术 • 上一篇    下一篇

基于边缘跟踪方法的肤色区域检测

郑纪虎,党德玉   

  1. (东北电力大学信息工程学院,吉林 吉林 132012)
  • 收稿日期:2011-09-14 出版日期:2012-07-20 发布日期:2012-07-20
  • 作者简介:郑纪虎(1986-),男,硕士,主研方向:图像识别;党德玉,教授

Skin Color Region Detection Based on Edge Tracking Method

ZHENG Ji-hu, DANG De-yu   

  1. (School of Information Engineering, Northeast Dianli University, Jilin 132012, China)
  • Received:2011-09-14 Online:2012-07-20 Published:2012-07-20

摘要: 常用的肤色模型只针对肤色的颜色信息,在较复杂的光照情况下检测效果不理想。为此,结合肤色在HSV颜色空间的分布特点以及肤色区域的边缘信息,提出一种应用于肤色区域检测的边缘跟踪方法。实验结果表明,该方法能准确检测图像中的肤色区域,检测结果优于传统的肤色模型,且对侧光、阴影等具有较好的鲁棒性。

关键词: 肤色区域检测, 肤色模型, 颜色空间, 边缘跟踪, 优先搜索方向

Abstract: Common skin models focus on the color information of skin, they cannot therefore obtain an ideal effect in complex light environment. Based on the distribution characteristics of the skin in HSV color space and the skin edge information, this paper presents a new edge tracking method, it can detect skin region in an image accurately, and the recognition result is better than traditional models. It has also better robustness even in slide light or shadow condition.

Key words: skin color region detection, skin color model, color space, edge tracking, first search direction

中图分类号: