摘要
在移动机器人的同步定位与建图(SLAM)研究中,使用单一的Kinect传感器或激光传感器时,出现建图精度低、信息不完整、回环检测易出错、可靠性差等问题,为此提出一种方案,将这两种传感器进行融合,利用融合之后的数据来创建环境地图。该方法首先将Kinect采集的深度图转换成伪激光数据后,与激光传感器的数据进行融合,来构建环境的局部栅格地图,并利用融合后的数据进行激光回环检测。当检测到有回环时,再利用Kinect的RGB图的丰富的视觉信息进行视觉回环检测。通过双回环检测来过滤掉错误的回环信息,提高全局地图的精度和完整性。最后在机器人操作系统上进行试验,结果表明该方案能显著提高建图的精度和完整性。
In the research of mobile robot simultaneous localization and mapping(SLAM),there are some problems such as low accuracy of mapping,incomplete information,error loop closure and poor reliability when using a single Kinect sensor or laser sensor.A scheme is proposed to fuse these two sensors,and the environment map is constructed by means of the fusion data.In this method,the depth map acquired by means of the Kinect is converted into the pseudo laser data,which is fused with the data of the laser sensor,to create a local raster map of the environment and conduct the laser loopback detection with the fused data.When a loopback is detected,the visual loop detection will be performed by using the visual information of the Kinect′s RGB image.The double loop closure detection is used to filter out the wrong loop closure and improve the accuracy and integrity of the global map.The experiments on the robot operating system show that the method can significantly improve the accuracy and integrity of the map.
作者
陈超
徐军
张伟伟
CHEN Chao;XU Jun;ZHANG Weiwei(College of Mechanical Engineering,Jiangsu University of Science and Technology,Zhenjiang 212003,China)
出处
《现代电子技术》
北大核心
2020年第14期164-169,共6页
Modern Electronics Technique
基金
国家自然科学基金资助项目(51705217)。
关键词
同步定位建图
移动机器人
多传感器融合
环境地图
回环检测
系统试验
simultaneous localization and mapping
mobile robot
multi⁃sensor fusion
environment map
loopback detection
system testing
