|
| 基于RGB-D相机的改进ORB-SLAM2算法地图构建研究 |
| Study on Map Construction of Improved ORB-SLAM2 Algorithm Based on RGB-D Camera |
| |
| DOI：10.16018/j.cnki.cn32-1650/n.202501008 |
| 中文关键词: ORB-SLAM2 稠密建图 八叉树地图 关键帧选取 移动机器人 |
| 英文关键词: ORB-SLAM2 dense mapping octree map keyframe selection mobile robot |
| 基金项目: |
|
| 摘要点击次数: 131 |
| 全文下载次数: 83 |
| 中文摘要: |
| 针对传统ORB-SLAM2算法在动态场景下的关键帧信息不足问题，提出了一种基于RGB-D相机的改进关键帧选取算法。通过增加帧间运动量和转角因数筛选关键帧，并优化特征点跟踪，提高了关键帧选取的准确率；在原有三线程基础上，新增稠密建图和八叉树线程，构建稠密点云地图并压缩成八叉树地图；最后在TUM数据集上验证了算法的有效性。结果表明，改进关键帧选取的ORB-SLAM2算法在保证系统实时性的同时，显著提升了系统跟踪定位精度，降低了轨迹误差，构建的八叉树地图内存占用低，可直接于机器人导航规划。 |
| 英文摘要: |
| Aiming at the problem of insufficient keyframe information of traditional ORB-SLAM2 algorithm in dynamic scenes, an improved keyframe selection algorithm based on RGB-D camera is proposed. By increasing the motion amount between frames and the rotation angle factor to screen keyframes, and optimizing the feature point tracking, the accuracy of keyframe selection is improved. On the basis of the original three threads, a dense mapping thread and an octree thread are newly added to construct a dense point cloud map and compress it into an octree map. Finally, the effectiveness of the algorithm is verified on the TUM dataset. The results show that the ORB-SLAM2 algorithm with improved key frame selection can, while ensuring the real-time performance of the system, significantly enhance the tracking and positioning accuracy of the system, reduce the trajectory error. The constructed octree map has a low memory footprint and can be directly applied to robot navigation planning. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |