摘要
通过划分基于同时定位与地图构建(SLAM)技术构建的栅格地图,将地图中的室内区域识别出来,实现机器人导航目标位置的灵活标记。首先对SLAM技术构建的栅格地图进行了二值化,使用直线段检测器(LSD)识别地图中的所有线段,采用分组和拟合方法获得室内环境边界的线段;然后利用边界与边界交点将地图划分为多个矩形块;最后对矩形块进行合并、聚合等处理,从而在栅格地图上划分出室内区域。标记后的室内区域与语音功能相结合扩展了机器人导航的输入方式。测试结果表明:该方法将不同室内环境和存在家具条件下生成的栅格地图划分出室内区域时具有较高的召回率。
The method recognizes the indoor area in the map by dividing grid map based on simultaneous localization and mapping(SLAM)technology,and realizes the flexible marking of the robot navigation target position.Firstly,the grid map constructed by SLAM technology is binarized.Line segment detector(LSD)is used to identify all the line segments in the map.Grouping and fitting methods are used to obtain the line segments of the indoor environment boundary.The map is divided into multiple rectangular blocks,and finally the rectangular blocks are merged and aggregated to divide the indoor area on the grid map.The combination of the marked indoor area and the voice function expands the input mode of robot navigation.The results show that this method has a higher recall rate when dividing the grid maps generated in different indoor environments and in the presence of furniture into indoor areas.
作者
杨丰拓
杨杰
YANG Fengtuo;YANG Jie(School of Electromechanic Engineering,Qingdao University,Qingdao 266071,China)
出处
《传感器与微系统》
CSCD
2020年第10期142-144,148,共4页
Transducer and Microsystem Technologies
