介绍了在室内未知环境下移动机器人利用激光雷达、电子罗盘和里程计等传感器信息创建特征地图的方法；从激光雷达数据中提取直线特征作为地图的主要环境描述特征,采用构建直线模板的方法对雷达数据进行分簇,通过最小二乘法拟合出相应的直线并对冗余地图线段进行合并,从而得到较精确的特征地图；实验表明该机器人建立的环境特征地图是精确有效的,且与栅格地图相比数据量小,可进一步用于机器人的避障、路径规划等复杂任务.%The method of feature map building for the mobile robot in indoor environment by using multi-sensors information, such as laser rangefinder, electronic compass and odometer, is presented in detail. The line feature extracted from the laser rangefinder is chosen to describe the environment in the map. The laser rangefinder' s data is clustered by creating line templates. The line segments are fitted by Least Square Method and the map' s redundant line segments are merged, then a relative accurate feature map is built in the end. The experiments results demonstrated that the indoor environmental feature maps built for the robot are accurate and effective. Compared with grid map, the quantity of the feature map data is much smaller. The feature maps can be used in the robot' s future complex tasks, such as obstacles avoidance, path planning and so on.
展开▼
