激光与单目视觉融合的移动机器人运动目标跟踪
Moving target tracking of mobile robots with fusion of laser scanner and monocular camera

针对室内环境下的移动机器人运动目标跟踪问题, 提出一种基于激光与单目视觉传感信息融合的机器人 定位和目标运动估计方法. 首先, 利用激光传感信息实现对目标的检测, 并完成机器人定位与环境建图; 然后, 设计 一种基于单目视觉传感器的目标位置估计算法, 获得目标的距离和角度信息; 为了实现两类传感信息的有效融合, 将激光与单目视觉进行联合标定, 得到二者的相对位姿关系, 基于此, 将激光与单目视觉提取的目标距离和角度通 过具有最优重要性函数和权重的粒子滤波器进行融合, 实现对目标运动状态的准确估计. 实验结果表明该方法具有 良好的跟踪性能.
Moving target tracking with mobile robots in the indoor environments is addressed. A method to localize the robot and estimate the moving target’s states is proposed based on information fusion of a laser scanner and a monocular camera. First, moving target detection, robot self-localization and map building of the environment are simultaneously accomplished using the laser data. Then, a monocular vision-based scheme to detect the target’s location is developed for measuring the range and bearing of the target. To achieve effective fusion between two kinds of sensor information, the laser scanner and the monocular camera are jointly calibrated to obtain the relative coordinate transform between them. Finally, a particle filter (PF) with the optimal importance function and importance weights is designed to fuse the measurements of the target’s range and bearing from the laser scanner and the monocular camera. The experimental results show that the proposed method can achieve satisfactory tracking performances.