.png){kind=icon}
.png){kind=icon}
Return
Uncertainty-Aware LiDAR Object Registration Algorithm for Urban Semantic Mapping
.png){kind=icon}
PRE
AI
DOI：10.1109/TASE.2025.3632337.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
LiDAR directly acquires 3D point measurements of physical surfaces, making it well-suited for interpreting objects in tasks such as object reconstruction and semantic mapping. However, each object is represented by a sparse point set, and the observations are inherently partial due to self-occlusion, which poses significant challenges for object registration. The uncertainty of the modeled surface depends on the local point density, and outliers from non-overlapping areas are often involved in the registration process. To address these issues, we propose a stochastic object registration method that models surface uncertainty using a Gaussian process and an overlap determination scheme based on object detector outputs. Both the overlap determination scheme and the stochastic object registration method are verified through Monte Carlo simulations on numerous real-world urban object point clouds, demonstrating performance improvements of 52% and 12%, respectively, compared to the baseline methods. To validate the full pipeline in real-world conditions, the proposed registration method is integrated with PointPillars, achieving an 8% accuracy improvement over the baseline method and demonstrating a 98.8% convergence ratio in LiDAR-based vehicle registration. Note to Practitioners—This paper presents a registration algorithm using LiDAR for object mapping in urban environments. Two major challenges in this task are the non-overlapping regions within point clouds and the empty spaces between points, which result from differences in the observed LiDAR poses and the limited number of points representing objects within large-scale point clouds, respectively. To address these issues, this work proposes two key components: 1) an overlap determination scheme based on an object detector, and 2) a probabilistic formulation of the objective function using Gaussian processes to model surface uncertainty. Each module is validated under various initial error conditions, and the comprehensive mapping system, integrated with the object detector, demonstrates superior performance over existing methods in terms of robustness and accuracy in diverse urban semantic mapping scenarios.
Keywords:
LiDAR perception
object registration
semantic mapping
uncertainty estimation
Journal
IF:
6.4
Papers:
4.9K
Citations:
1.6W