基于ROS与三维点云图像的室内物体精准定位

Precise location of indoor objects based on ROS and point cloud images

考虑到ROSSLAM构建的地图只能描述环境的二维信息,三维点云图像只能描述物体独立的三维信息等特点,本文融合了ROSSLAM的Gmapping算法构建的室内二维地图与物体的三维点云图像信息,提出了一种复合坐标定位系统。首先对不同室内进行分类,进行一维坐标的标定,其次通过对Gmapping算法构建好的地图等进行二三维坐标标定,再结合空间信息构成外部坐标系φ,最后通过对采集到的物体三维点云坐标进行仿射运算获得物体基于外部坐标三维坐标,结合一维坐标,对物体进行复合四维坐标定位。整个定位实验数据表明,物体室内的位置平均测量误差只有4.2cm,其定位精度比起常见的超声波与红外线定位系统提高6.7%,比基于蓝牙角度测量的定位系统定位精度提高20%,比超宽带定位系统提高72%。物体定位误差小,定位精准。

Considering that the map constructed by ROS SLAM can only describe the two-dimensional information of the environment, the three-dimensional point cloud image can only describe the independent three-dimensional information of the object,this paper combines the information of indoor two-dimensional map and object′s three-dimensional point cloud image constructed by ROS SLAM′s Gmapping algorithm, and proposes a composite coordinate positioning system. Firstly, different indoors are classified and one-dimensional coordinates are calibrated. Secondly, two-dimensional and three-dimensional coordinates are calibrated by the map constructed by Gmapping algorithm, and then external coordinates are formed by combining spatial information. Finally, three-dimensional coordinates of the collected object are obtained by affine operation based on external coordinates. Combined with one-dimensional coordinates, the object is combined with four dimensional coordinates. The whole positioning experiment data show that the average error of indoor position measurement is only 4.2 cm, and the positioning accuracy is 6.7% higher than that of the common ultrasonic and infrared positioning systems, 20% higher than that of the positioning system based on Bluetooth angle measurement, and 72% higher than that of the ultra-wideband positioning system. The object location error is small and the location is precise.