论文标题
机器人使用情境图和构建建筑计划进行定位
Robot Localization using Situational Graphs and Building Architectural Plans
论文作者
论文摘要
建筑行业的机器人可以使用高精度数据捕获来通过不断监视工作进度来降低成本。准确的数据捕获需要在环境中精确的移动机器人定位。在本文中，我们介绍了关于机器人本地化的新颖工作，该工作以墙壁和房间的形式提取了从建筑计划中提取几何，语义以及拓扑信息，并在环境中导航之前创建了情境图（Sgraphs）的拓扑和度量语义层。当机器人在施工环境中导航时，它使用机器人探光仪和从3D激光雷达测量的平面壁的形式使用感官观察，以估计其依靠粒子过滤器方法的姿势，通过利用先前构建的情境图及其可用的几何形状，语义和拓扑信息。我们在将其与基于传统几何的本地化技术进行比较时，在实际持续的施工站点上捕获的模拟和真实数据集中验证了我们的方法。
Robots in the construction industry can reduce costs through constant monitoring of the work progress, using high precision data capturing. Accurate data capturing requires precise localization of the mobile robot within the environment. In this paper we present our novel work on robot localization which extracts geometric, semantic as well as the topological information from the architectural plans in the form of walls and rooms, and creates the topological and metric-semantic layer of the Situational Graphs (S-Graphs) before navigating in the environment. When the robot navigates in the construction environment, it uses the robot odometry and the sensorial observations in the form of planar walls extracted from the 3D lidar measurements, to estimate its pose relying on a particle filter method, by exploiting the previously built situational graph and its available geometric, semantic and topological information. We validate our approach in both simulated and real datasets captured on actual on-going construction sites presenting state-of-the-art results when comparing it against traditional geometry based localization techniques.
