High-Speed Outlier Removal Filter for LiDAR Sensor Point Cloud Data
This paper proposes an innovative high-speed outlier removal filter designed to efficiently eliminate noise in light detection and ranging (LiDAR) sensor point cloud data, which is essential for real-time autonomous vehicle applications. Traditional filtering methods rely on exhaustive neighbor sear...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10802883/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This paper proposes an innovative high-speed outlier removal filter designed to efficiently eliminate noise in light detection and ranging (LiDAR) sensor point cloud data, which is essential for real-time autonomous vehicle applications. Traditional filtering methods rely on exhaustive neighbor searches, which result in long computation times that remain constant regardless of the increase in noise. The proposed high-speed outlier removal filter addresses these limitations by utilizing the sequential rotational measurement characteristics of LiDAR sensors, enabling noise identification without requiring neighbor searches. Experimental results demonstrate that the proposed filter achieves noise removal performance comparable to that of existing methods while greatly reducing the processing time. Specifically, the proposed filter was found to be up to 92 times faster than dynamic statistical outlier removal and 24 times faster than dynamic radius outlier removal in 3D LiDAR data collected under snowy conditions. Moreover, the superiority of the proposed algorithm was demonstrated by evaluating its noise removal and original signal preservation performance under simulated noise conditions with varying noise ratios on roads with multiple vehicles, and comparing the results with those of existing methods. These findings suggest that the proposed filter has the potential to enhance the real-time processing capabilities of autonomous vehicles, particularly in adverse weather conditions. |
|---|---|
| ISSN: | 2169-3536 |