基于RGB-D相机的室内移动机器人自定位方法

A Self-localization Method Based on RGB-D Camera for Indoor Mobile Robots

精准可靠的自定位是移动机器人实现多机协同、路径规划与控制决策等自主能力的基础;因此,室内服务机器人作为移动机器人的典型代表,要求能够实时进行自定位,并且有效地避开各种静态和动态障碍物;基于此,提出一种基于RGB-D相机的室内移动机器人自定位方法,该方法利用Hough变换对机器人建立的环境地图进行线特征提取,并建立环境模型误差查找表,将非结构化环境中的自定位问题转化为结构化环境中的自定位问题,然后利用匹配优化算法实现自定位;实验结果表明,采用所提出的方法,机器人能够实现室内环境下实时精确的自定位。

Accurate self-localization is the basis to realize mobile robot’s autonomous ability such as path planning,cooperation and control decision.Compared with the mature industrial application of the robot,indoor service robots have some of their own key technologies,such as real-time self-localization,and avoiding a variety of static and dynamic obstacles effectively.Therefore,a self-localization method of indoor mobile robot based on RGB-D camera is proposed.Firstly,the Hough transform is used to extract the line feature of the environment map and establish the environment model error table.Then the self-localization method combined with matching optimization algorithm is used to calculate the robot’s position rapidly and accurately.The experimental results show that the robot can realize real-time and accurate self-localization.