InsClustering: Instantly Clustering LiDAR Range Measures for Autonomous Vehicle
This addresses the need for real-time processing in autonomous driving systems, though it appears incremental as it focuses on speeding up existing clustering steps.
The paper tackles the problem of fast 3D point cloud segmentation for autonomous vehicles by proposing a method that processes raw LiDAR data with a delay of less than 1ms for full 360-degree scans, while maintaining good accuracy in field tests.
LiDARs are usually more accurate than cameras in distance measuring. Hence, there is strong interest to apply LiDARs in autonomous driving. Different existing approaches process the rich 3D point clouds for object detection, tracking and recognition. These methods generally require two initial steps: (1) filter points on the ground plane and (2) cluster non-ground points into objects. This paper proposes a field-tested fast 3D point cloud segmentation method for these two steps. Our specially designed algorithms allow instantly process raw LiDAR data packets, which significantly reduce the processing delay. In our tests on Velodyne UltraPuck, a 32 layers spinning LiDAR, the processing delay of clustering all the $360^\circ$ LiDAR measures is less than 1ms. Meanwhile, a coarse-to-fine scheme is applied to ensure the clustering quality. Our field experiments in public roads have shown that the proposed method significantly improves the speed of 3D point cloud clustering whilst maintains good accuracy.