基于人脸识别与光流追踪的移动机器人视觉导航方法

针对移动机器人视觉导航中跟踪目标丢失的问题,提出了基于人脸识别与稀疏光流算法(KLT)结合的移动机器人视觉导航方法(FR-KLT视觉导航方法)。采用OpenCV库中的Haar特征提取人脸识别算法实时检测识别目标人脸,通过Harris角点检测获取目标人体特征点,对目标人体进行精准定位;KLT光流追踪法测算目标移动趋势,并预测目标下一刻大致位置。目标人体位置变动时移动机器人对目标进行实时追踪导航。通过Pioneer-LX机器人在真实环境下试验,验证了该方法准确识别并跟踪目标的实时性和有效性。

Visual Navigation Method of Mobile Robot Based on Face Recognition and Optical Flow Tracking

Aiming at the problem of tracking target lost in vision navigation of mobile robots,this paper presents a visual navigation method(FR-KLT visual navigation method) of mobile robots based on face recognition algorithm and Kanade-Lucas-Tomasi(KLT) sparse optical flow algorithm. Face detection and recognition algorithm based on Haar feature extraction in OpenCV database is used to detect and recognize the target face in real-time,the feature points of human body are obtained by Harris corner detection,and the target human body is accurately located. KLT optical flow tracing method is used to measure the human body movement trend and to determine the approximate position of the target at the next moment according to the obtained feature points. When the position of human body changes,the mobile robot can track and navigate the target in real time. The real-time performance and effectiveness of the method are verified by the experiment of the Pioneer-LX robot in real environment.