摘要
为了解决不同光照条件下的肤色分割问题,提出了一种新颖并且行之有效的解决方法。该方法先应用颜色恒常性算法对图像光照进行估计,根据光照估计的结果,在YCbCr颜色空间采用不同分布的高斯肤色模型,并利用动态阈值对图像进行分割。试验结果证明了算法的可行性和优越性。
In order to extract skin color under varying illumination,this paper introduces a novel and effective method.Firstly color constancy algorithm is used to estimate the illumination of the image,then different Gauss skin models are built according to different illuminations,and automatic-thresholding technique is used for skin-color segmentation.The experiment results show that it is feasible and rather successful.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第17期82-84,88,共4页
Computer Engineering and Applications
基金
天津市自然科学基金(the Natural Science Foundation of Tianjin City of China under Grant No.05YFJMJC09200)
关键词
颜色恒常性
高斯模型
肤色分割
color constancy
Gauss model
skin segmentation
