- Team Marketing
- October 8, 2024
Evaluating and Improving RoSELS for Road Surface Extraction from 3D Automotive LiDAR Point Cloud Sequences | Technical Paper
Navigable space determination is a difficult problem encountered in robotics and intelligent vehicle technology and it requires integrated solutions using advances in computer vision and computational geometry.
The conventional data processing workflow uses semantic segmentation to identify road points from three-dimensional (3D) automotive LiDAR point clouds, which have to be extended to determine its boundary points.
The boundary points are critical in getting an initial coarse extraction of its surface. Hence, to filter the curb or edge points from the road points, the semantic segmentation results are post-processed.