基于点线相合的机器人增量式地图构建

Incremental mapping based on dot-line congruence for robot

提出基于测量数据点和已构建地图线的最佳相合性增量式构建未知环境地图的方法,将机器人地图构建过程分解为局部地图构建、机器人位姿估计和地图合并3个循环步骤,在局部地图构建中,采用哈夫变换拟合、同线性判断和最小二乘拟合相结合的方法从测量数据点中拟合得到局部线段集合,在位姿估计时,首先利用点线匹配寻找测量数据和已构建地图之间的匹配关系,然后通过去除不当匹配和引入加权矩阵来减小测量误差和已构建地图中的不确定性对位姿估计的影响,最后利用加权最小二乘法估计机器人的位姿,使得匹配部分达到最佳相合,同时提出虚拟线和虚拟点的方法解决由伪相合条件所引起的错误位姿估计问题,实验结果证明了算法的有效性和鲁棒性,适于构建室内环境地图。

Based on the best congruence between dot data in the current measurement and line segments in the previously-built map, this paper proposes an incremental mapping approach for the robot in unknown environment. Each iteration of this approach consists of 3 stages： local map building, robot-pose estimating, and map integrating. A combining method of Hough transform, coincided-line detecting and least squares curve-fitting is presented and used to fit the line segments from measurement in local map building. In pose estimating, the rough correspondence between the measurement and the half-baked map is obtained firstly by dot-line matching. Then removing improper match and defining weighted matrix are implemented to refine the correspondence and to reduce the errors of both measurement and map. Finally, the estimated pose is figured out by weighted least squares with the best congruence. The pseudocongruence problem in pose estimating is also discussed and solved by adding virtual lines and dots in this paper. Experimental results with real data are presented, which demonstrate that the approach is effective and robust for indoor environment mapping.