摘要
该文提出了一种基于广义对称变换的从具有复杂背景的彩色图像中进行人脸检测的方法 ,它首先利用人类肤色在色度空间分布的稳定性 ,检测出图像中的皮肤区域 ,通过先验知识进行甄别 ,选出候选人脸区域。然后根据在人脸图像中人的双眼具有的高对称性 ,利用广义对称变换来求得具有高对称性的点即为双眼的位置 ,最后采用通用人眼图像进行模板匹配来验证。实验证明 ,该方法具有良好的效果。
In this paper, an approach to face detection, based on skin color mode l and generalized symmetry transform is introduced. With the stability of human skin color's distribution in the chromatic space, the skin color areas are segme nted. Since the two eyes on the face are highly symmetrical, the generalized sym metry transform is then employed to locate the eyes on the skin color region. F inally, the average eye images are used as template to match the eye location. T he result of the experiment proved the efficiency of the algorithm.
出处
《计算机仿真》
CSCD
2003年第2期117-119,64,共4页
Computer Simulation
