摘要
为了在ORB-SLAM运行数据集后得到点云图像数据能够保存,通过添加程序接口,实现保存功能,再使用激光雷达云的bin格式转pcd格式的方法生成pcd格式的点云文件,将多个点云使用基于特征的ICP算法进行拼接,得出了稳定性的拼接点云,为实现未来对点云图像的下一步操作提供了基础,提高多机器人的定位精度和系统建图效果。
In order to save the point cloud image data obtained after the ORB-SLAM runs the data set,the save function is realized by adding a program interface,and then the method of converting the bin format of the lidar cloud to the pcd format is used to generate the point cloud file in the pcd format,and multiple the point cloud is stitched using the feature-based ICP algorithm,and a stable stitched point cloud is obtained,which provides a basis for the next step of the point cloud image operation in the future,and improves the positioning accuracy of multi-robots and the effect of system mapping.
作者
李宗周
刘明芹
LI Zong-zhou;LIU Ming-qin(School of Mechanical Engineering,Jiangsu Ocean University,Lianyungang Jiangsu 222005;China University of Mining and Technology,Xuzhou Jiangsu 221116)
出处
《数字技术与应用》
2021年第8期87-89,共3页
Digital Technology & Application
基金
基于ORB特征的单机器人同时定位与地图创建算法研究与实践(KYCX20-2915)。
关键词
单目相机
3D点云
3D融合
稀疏点云
Monocular camera
3D point cloud
3D fusion
Sparse point cloud
