.png){kind=icon}
.png){kind=icon}
Return
Roadside LiDAR-Filtering Method Based on a Partial Background Optimization Method
.png){kind=icon}
PRE
AI
DOI：10.1061/JTEPBS.TEENG-9160.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Background filtering is a key step for traffic information acquisition using roadside light detection and ranging (LiDAR) sensors because their performance directly affects the accuracy of downstream perception tasks. In this paper, we proposed a novel background construction-based filtering method that incorporates a region-specific optimization framework to improve the accuracy of background construction, thereby enhancing the overall performance of background filtering. Specifically, we introduced a region partitioning strategy based on grid partitioning and scanning frequency features, establishing a stronger connection with real-world traffic scenarios. Subsequently, we applied different strategies to different regions, including point distribution-based strategies, occupancy score-based strategies, and statistical analysis-based strategies, to construct a more accurate and adaptive background model. Finally, we evaluated our method on an open-source data set. The experimental results demonstrated that the proposed method can construct accurate background point clouds and outperform the other existing filtering methods in mean intersection over union (mIOU) and overall accuracy (OA).
Keywords:
Roadside light detection and ranging (LiDAR)
Background construction
Background filtering
Background optimization
Online filtering
Journal
J
IF:
2.1
Papers:
130
Citations:
1.9K