复杂背景下视频序列中的人脸定位算法

Face Location Algorithm in Video Sequences with Complex Background

视频序列中的人脸定位是利用视频图像进行人脸识别的关键技术。为提高视频序列中人脸定位的准确性,通过分析彩色视频序列中的人脸特点,提出一种采用背景去除、肤色区域探测、水平亮度投影和垂直梯度运算、特征点拟合等人脸模板特征的方法,对视频序列中的人脸进行定位。实验结果表明,该方法定位的人脸能够满足实际应用,已利用本方法开发了人脸门禁系统。

Face location is a key technique of face recognition using images from video sequences. In order to improve the accuracy of face location in video sequences, the paper presents a method for locating face in video sequences through analysis of face features in video sequence. Background removing, skin region detecting, horizontal brightness projection and face template with gradient calculating, feature component fitting are adopted for this method. The experiments demonstrate that this method can meet the requirements of actual application. In addition, face recognition-based access control system has been developed using this method.