计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (14): 166-168.

• 数据库、信号与信息处理 • 上一篇    下一篇

一种基于BP网络的人体肤色模型

陈 震,王甦菁,周春光,曹欢欢   

  1. 吉林大学 计算机科学与技术学院,长春 130012
  • 收稿日期:2007-08-27 修回日期:2007-11-05 出版日期:2008-05-11 发布日期:2008-05-11
  • 通讯作者: 陈 震

Human skin color model using BP networks

CHEN Zhen,WANG Su-jing,ZHOU Chun-guang,CAO Huan-huan   

  1. College of Computer Science and Technology,Jilin University,Changchun 130012,China
  • Received:2007-08-27 Revised:2007-11-05 Online:2008-05-11 Published:2008-05-11
  • Contact: CHEN Zhen

摘要: 人脸识别已在出入安全检查、视频编码、视频监测与跟踪,基于内容的检索等方面都有着极其重要的应用。皮肤检测是人脸识别和跟踪的关键步骤。其中基于颜色的皮肤检测具有简单、快速、直观,不受物体形状变化、视点改变等影响的优点。提出了一种使用BP人工神经网络建立的人体肤色模型。该模型在不同的颜色的光照下有很好的鲁棒性。

关键词: 人脸识别, 皮肤检测, 神经网络

Abstract: Facial recognition plays an important role in applications such as entrance security,video coding,video surveillance and tracking,and image retrieval based on content.However,skin detection is a key step of facial recognition and tracking.The merits of skin detection based color are simpleness,efficiency and perceptual intuition etc.This paper presents a new human skin color model using BP networks.For various illuminations,it is robust.

Key words: facial recognition, skin detection, neural network