arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2405.15664cs.ROcs.CV
收藏

GroundGrid:LiDAR点云地面分割与地形估计

GroundGrid:LiDAR Point Cloud Ground Segmentation and Terrain Estimation

Dahlem Center for Machine Learning and Robotics (DCMLR)(达勒姆机器学习与机器人学中心);Freie Universität Berlin(柏林自由大学)

Nicolai Steinke, Daniel Göhring, Raùl Rojas

AI总结:

GroundGrid利用2D高程图实现LiDAR点云地面分割与地形估计,在SemanticKITTI上以94.78%平均IoU和171Hz性能超越现有方法。

摘要:

精确的点云地面分割是自动驾驶车辆中几乎所有LiDAR传感器感知任务的关键前提。特别是从点云中聚类和提取物体通常依赖于准确的地面点移除。对周围地形的正确估计对于地表可驾驶性、路径规划和障碍物预测等方面至关重要。在本文中,我们提出了我们的系统GroundGrid,它依赖于2D高程图来解决地形估计和点云地面分割问题。我们评估了GroundGrid的地面分割和地形估计性能,并将其与其他最先进的方法进行了比较,使用了SemanticKITTI数据集和一种依赖于机载LiDAR扫描的新型评估方法。结果表明,GroundGrid能够以94.78%的平均IoU超越其他最先进的系统,同时保持171Hz的高运行性能。源代码可在https://github.com/dcmlr/groundgrid获取。

英文摘要:

The precise point cloud ground segmentation is a crucial prerequisite of virtually all perception tasks for LiDAR sensors in autonomous vehicles. Especially the clustering and extraction of objects from a point cloud usually relies on an accurate removal of ground points. The correct estimation of the surrounding terrain is important for aspects of the drivability of a surface, path planning, and obstacle prediction. In this article, we propose our system GroundGrid which relies on 2D elevation maps to solve the terrain estimation and point cloud ground segmentation problems. We evaluate the ground segmentation and terrain estimation performance of GroundGrid and compare it to other state-of-the-art methods using the SemanticKITTI dataset and a novel evaluation method relying on airborne LiDAR scanning. The results show that GroundGrid is capable of outperforming other state-of-the-art systems with an average IoU of 94.78% while maintaining a high run-time performance of 171Hz. The source code is available at https://github.com/dcmlr/groundgrid

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑