彩色静态图像的人脸检测研究  

Research on face detection in color static images

作　　者：潘翔[1,2] 王万森[1] 黄远鸣[2] 

机构地区：[1]首都师范大学,北京100037 [2]广西经济管理干部学院,广西南宁530007

出　　处：《计算机工程与设计》2009年第11期2785-2787,共3页Computer Engineering and Design

基　　金：北京市教委重点科技项目暨北京市自然科学基金重点项目(KZ200810028016)

摘　　要：讨论了一种彩色静态图像的人脸检测法,先使用光照补偿对图像进行预处理,再采用YCbCr肤色模型检测出图像的肤色部分,然后对肤色部分进行膨胀、腐蚀,去除非人脸区域,最后使用人工神经网络对候选人脸区域进行验证,最终定位出人脸区域。实验证明了该方法对不同光照环境和复杂背景的图片有较好的适应性,提高了检测速度与精度。An algorithm for face detection in color static images is discussed. The algorithm works by first doing lighting compensation on input image, then it detected the skin regions by the YCrCb skin model. Third, it eliminated the regions which are not belonged to the real face region by dilation and erosion. Finally, it used the neural network to evaluate the face candidates and then located the real face region. The experiments proved that the adaptability ofthis algorithm to those images of different lighting environment and complex background is good. It heightens the speed and the accurate degree of the detection.

关 键 词：人脸检测 光照补偿 肤色模型 膨胀腐蚀 神经网络 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]