A Fast Human Face Detection Algorithm Based on Complezion Model

摘要：

In order to improve face detection accuracy rata and speed,the algorithm based on complexion model is proposed. The algorithm first pay lighting compensation to the collection image which counteract the color bias in the image. And do nonlinear conversion to the RGB color space,Divide the non-skin color and the skin color regional binary image according the nonlinear conversion results, established complexion model,the initial screening of the general location of face,then do expansion and corrosion to the image, remove some of the noise may be regional,final borders based on the four border connectivity accurate algorithm will face regional separate. The algorithm is implemented in Visual C++6.0,the results show that the algorithm is appliance rapid human face detect and the location,which has higher detection correct rate and higher detection speed.

关键词： face detection YCbCr space skin color model region segmentation four border connectivity algorithm

作者: Kai SONG Ying-ying LI

作者单位: Info.Science and Engineering College,ShenYang Ligong University,Shenyang 110168,China

会议类型: 国际会议

会议名称: 2009 International Conference on Information,Electronic and Computer Science(2009 国际信息、电子与计算机工程学术会议)

会议地点: 青岛

会议语种:英文

页码: 607-610

在线出版日期: 2009-11-21（万方平台首次上网日期，不代表论文的发表时间）