说明：  针对彩色图像提出了一种基于肤色和模板的人脸检测方法, 由肤色分割、模板匹配和人工神经网验证3 部 分组成. 首先使用HS I 空间的肤色统计模型分割出可能包含人脸的区域, 然后使用平均脸模板匹配和人工神经 网验证的方法在这些区域中搜索人脸. 该方法将彩色图像的肤色信息和灰度图像的模板匹配及人工神经网分类 模型综合起来, 既极大地提高了速度, 又具有较强的鲁棒性. 实验结果表明, 该算法是快速而有效的.
(For color images is presented based on color and template face detection methods, from the color segmentation, template matching and artificial neural network validation three parts. The first to use HS I split the statistical model of color space which might contain human face region, and then use the average face template matching and artificial neural network authentication methods in these areas in search human face. This method of color image color information and gray-scale image of the template matching and artificial neural network classification model together, not only greatly enhanced the speed, but also has strong robustness. The experimental results show that the algorithm is fast and effective.)