在此提出了一个有效的无线远程人脸追踪系统实现方案,将人脸检测技术、人脸识别技术和窄带通信技术有效结合,采用AdaBoost追踪算法获取图像中可分辨的人脸,把各个点摄像终端的数据进行压缩后采用GPRS/CDMA无线传榆发送至处理中心,采用Gabors特征提取算法和SVM分类算法对人脸数据库进行筛选,对实时传输的人脸进行识别,从而达到辅助寻找和追踪特定人的目的.该方案通过实验证明是可行的.%A solution of an effective wireless remote face tracking system is presented. Three technologies such as face detection and face recognition and narrowband modulation are effectively combined in the scheme. AdaBoost algorithm was adopted to extract the distinguishable face in an image. The data from the each point amera terminal is compressed and then sent to the processing center by GPRS/CDMA wireless network. The Gabor feature extraction algorithm and SVM classification algorithm are employed to conduct the screen of the face database, and then carry out the recognition of real-time transmitted face to achieve the auxiliary search and tracking of the special person. The experimental results show the scheme is feasible.
展开▼