Cross-Section-Based Method for LiDAR Dataset Generation With Multipath-Resilient Ground Truth

The demand for high-quality light detection and ranging (LiDAR) datasets is increasing as LiDAR technology is being used in various applications, including autonomous vehicles, robotics, and 3-D mapping. However, generating accurate ground-truth (GT) data for LiDAR datasets remains a challenge due to issues like multipath interference (MPI) and other disturbances. The method for generating high-quality GT data for LiDAR applications based on cross section (CS) is introduced in this article. The key concept is based on precisely registering a digital twin, CAD-based, which is later converted to a mesh, with LiDAR depth images. By utilizing selective point usage in its CSs, the method demonstrates greater robustness to MPI compared with standard approaches. The technique is evaluated using a dataset of LiDAR-acquired point clouds of a living room scene. The results show that the proposed technique achieves significantly better accuracy in affected regions than standard iterative closest point (ICP)-based methods. In addition, this article proposes a new set of metrics for evaluating the quality of GT data, which is more robust to MPI than standard metrics such as root-mean-square error (RMSE) and Chamfer distance (CD). The proposed technique is a valuable tool for generating large-scale, high-quality datasets for LiDAR applications. Finally, we compare our dataset with the latest available datasets.