论文部分内容阅读
为了解决机器人同时定位,地图构建与目标跟踪(SLAMOT)过程中的多源,异构传感器空间一致性观测问题,提出了基于信息融合的摄像机与激光测距传感器联合标定优化方法。完成基于误差传播公式的激光扫描点图像平面投影不确定范围判定,并利用协方差交集算法实现基于运动物体检验方法和基于Camshift方法的图像坐标系下目标状态融合。在此基础上,利用目标图像平面投影方向误差构造目标函数,通过非线性优化方法实现摄像机与激光测距仪标定参数优化。实验验证了设计方法能有效提高目标跟踪以及多传感器参数标定的准确性。相关成果能够为基于多传感器信息融合的机器人同时定位,地图构建与目标跟踪滤波方法研究提供观测值支持。
In order to solve the problem of observing the spatial consistency of multi-source and heterogeneous sensors in the simultaneous localization, mapping and target tracking (SLAMOT) of robots, a joint calibration and optimization method based on information fusion between camera and laser ranging sensor is proposed. The determination of the indefinite range of plane projection of laser scanning point image based on error propagation formula is completed, and the target state fusion based on Camshift method and image coordinate system is realized by the method of intersection of covariance. On this basis, the target function is constructed by using the direction error of the projection of the target image plane, and the calibration parameters of the camera and laser rangefinder are optimized by the nonlinear optimization method. Experiments show that the design method can effectively improve the accuracy of target tracking and multi-sensor parameter calibration. The results can provide support for the simultaneous localization of robots based on multi-sensor information fusion, map construction and target tracking filtering methods.