View in EDS

Impact of LiDAR beam loss on 3D object detection: A systematic analysis of vulnerable road user safety.

Saved in:
Bibliographic Details
Title: Impact of LiDAR beam loss on 3D object detection: A systematic analysis of vulnerable road user safety.
Authors: Feng, Weicong1 (AUTHOR) wfeng@ccny.cuny.edu, Li, Yiqiao2 (AUTHOR), Wei, Jie1 (AUTHOR)
Source: Pattern Recognition Letters. Jul2026, Vol. 205, p155-161. 7p.
Subjects: LIDAR, Road users, Autonomous vehicles, Detection algorithms, Object recognition (Computer vision)
Abstract: • Measure 3D detector degradation under systematically controlled LiDAR beam loss. • Generalize results by reporting statistics from repeated, randomized beam removals. • Derive VRU thresholds for the number of damaged beams and the best detector. • Compare effects across detectors, classes, and datasets. • Show beam-loss impact is location-sensitive and non-uniform across beams. Light Detection and Ranging (LiDAR) sensors provide high-resolution 3D perception and are widely used in autonomous driving and intelligent transportation systems. However, the reliability of LiDAR can be compromised by partial beam loss caused by sensor degradation, occlusion, or adverse environmental conditions. This degradation poses safety concerns, particularly for vulnerable road users (VRUs), which are smaller and harder to detect than other classes of vehicles. While previous research has focused on global beam-density reduction and weather-related degradation, the consequences of losing vertical beams at specific locations within the LiDAR field of view, a practical issue associated with sensor configuration, are still insufficiently investigated. To address this knowledge gap, we investigate how the loss of vertical LiDAR beams affects object detection performance across six state-of-the-art detection models using the KITTI and nuScenes datasets. Vertical beam loss is progressively simulated to assess its overall effect on performance, category-specific sensitivity (e.g., vehicles vs. VRUs), and the influence of beam loss location. We also identify thresholds beyond which detection performance deteriorates sharply, potentially compromising VRUs detection. Our empirical findings highlight that the perceptual degradation under vertical beam loss is highly dependent on the location of the beam loss and the selection of detection models. These findings provide quantitative guidelines for determining the minimum LiDAR beam configurations that balance sensor cost, reliability, and safety, offering practical insights for manufacturers, transportation agencies, and policymakers. [ABSTRACT FROM AUTHOR]
Copyright of Pattern Recognition Letters is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:• Measure 3D detector degradation under systematically controlled LiDAR beam loss. • Generalize results by reporting statistics from repeated, randomized beam removals. • Derive VRU thresholds for the number of damaged beams and the best detector. • Compare effects across detectors, classes, and datasets. • Show beam-loss impact is location-sensitive and non-uniform across beams. Light Detection and Ranging (LiDAR) sensors provide high-resolution 3D perception and are widely used in autonomous driving and intelligent transportation systems. However, the reliability of LiDAR can be compromised by partial beam loss caused by sensor degradation, occlusion, or adverse environmental conditions. This degradation poses safety concerns, particularly for vulnerable road users (VRUs), which are smaller and harder to detect than other classes of vehicles. While previous research has focused on global beam-density reduction and weather-related degradation, the consequences of losing vertical beams at specific locations within the LiDAR field of view, a practical issue associated with sensor configuration, are still insufficiently investigated. To address this knowledge gap, we investigate how the loss of vertical LiDAR beams affects object detection performance across six state-of-the-art detection models using the KITTI and nuScenes datasets. Vertical beam loss is progressively simulated to assess its overall effect on performance, category-specific sensitivity (e.g., vehicles vs. VRUs), and the influence of beam loss location. We also identify thresholds beyond which detection performance deteriorates sharply, potentially compromising VRUs detection. Our empirical findings highlight that the perceptual degradation under vertical beam loss is highly dependent on the location of the beam loss and the selection of detection models. These findings provide quantitative guidelines for determining the minimum LiDAR beam configurations that balance sensor cost, reliability, and safety, offering practical insights for manufacturers, transportation agencies, and policymakers. [ABSTRACT FROM AUTHOR]
ISSN:01678655
DOI:10.1016/j.patrec.2026.04.030