兴趣区域地图提取与ROS的多机器人SLAM地图合成研究

Map extraction of interest region and ROS multi-robot SLAM mapmerge

针对单个机器人在大场景下建图与定位效果不佳,以及地图合成不实时的问题,本文提出一种兴趣区域地图提取与ROS的多机器人SLAM地图融合方法。该方法由单个机器人构建局部地图,对地图进行兴趣区域提取后,再上传到给中央处理机器,由中央处理机器利用霍夫变换地图融合算法,进行地图融合处理。验证结果表明,本系统实现了多机器人协作实时建图与融合,降低了计算的繁杂度,提高了建图效率与准确性。

Aiming at the problem of poor mapping and positioning effect of single robot in large scenes and unreal-time map synthesis,a map fusion method of multi-robot SLAM based on region of interest map extraction and ROS is studied.Firstly,a local map is constructed by a single robot,and then the interest region of the map is extracted to reduce the calculation of fusion,and then uploaded to the central processing machine.The central processing machine uses the Hough transform map fusion algorithm to process map fusion.The experimental results show that the system realizes the real-time mapping and fusion of multi-robot collaboration,reduces the complexity of calculation,and improves the efficiency and accuracy of mapping.