中文摘要：

针对机器人的目标人跟踪问题,提出一种利用目标人腿部信息自适应跟踪的方法。该方法利用激光雷达作为环境感知传感器,实时获取环境的二维激光扫描数据;然后采用设计的激光相邻点聚类算法对激光扫描数据进行分割和聚类;再利用人腿圆弧状特征设计的类圆弧人腿形状识别算法从分割的数据中识别腿部数据,获得目标人腿部相对于机器人的位置信息;最后利用Kalman滤波算法对目标人的位置和速度进行跟踪,使机器人能够平稳地跟踪目标人运动。该算法在iRobot机器人平台上进行实验,实验结果验证了算法的有效性。

英文摘要：

For the target person tracking problem of the robot, in the article we propose an adaptive tracking method which uses leg information of the target person. The method employs laser radar as the environmental perception sensor to timely acquire 2D laser scanning data of the environment ; Then, it segment and cluster the laser scanning data by using the designed adjacent laser points clustering algorithm ; Next, it recognises the leg information from the segmented data by using arc-like human leg recognition algorithm designed according to the arc-shaped feature of human leg to get the position information of the target person leg relative to the robot; Finally, the Kalman filtering algorithm is used to track the position and speed of the target person to make the robot smoothly track the motion of the target person. This algorithm has been experimented on an iRobot robot platform, and the results verify its effectiveness.