Spatial Information Enhancement with Multi-Scale Feature Aggregation for Long-Range Object and Small Reflective Area Object Detection from Point Cloud
Remote Sensing · 18.07.2024 · 16 (14) · 2631
DOI: 10.3390/rs16142631 ↗Kurz gefasst
Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 18.07.2024. Verfasst von Hanwen Li und 4 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Research Project on Laser object Feature Extraction and Recognition.
Abstract (Originalsprache)
Accurate and comprehensive 3D objects detection is important for perception systems in autonomous driving. Nevertheless, contemporary mainstream methods tend to perform more effectively on large objects in regions proximate to the LiDAR, leaving limited exploration of long-range objects and small objects. The divergent point pattern of LiDAR, which results in a reduction in point density as the distance increases, leads to a non-uniform point distribution that is ill-suited to discretized volumetric feature extraction. To address this challenge, we propose the Foreground Voxel Proposal (FVP) module, which effectively locates and generates voxels at the foreground of objects. The outputs are subsequently merged to mitigating the difference in point cloud density and completing the object shape. Furthermore, the susceptibility of small objects to occlusion results in the loss of feature space. To overcome this, we propose the Multi-Scale Feature Integration Network (MsFIN), which captures contextual information at different ranges. Subsequently, the outputs of these features are integrated through a cascade framework based on transformers in order to supplement the object features space. The extensive experimental results demonstrate that our network achieves remarkable results. Remarkably, our approach demonstrated an improvement of 8.56% AP on the SECOND baseline for the Car detection task at a distance of more than 20 m, and 9.38% AP on the Cyclist detection task.
Förderung
- Research Project on Laser object Feature Extraction and Recognition
Zitieren
APA
Li, H., Tao, H., Deng, Q., Xiao, S., & Zhou, J. (2024). Spatial Information Enhancement with Multi-Scale Feature Aggregation for Long-Range Object and Small Reflective Area Object Detection from Point Cloud. Remote Sensing, 16(14), 2631. https://doi.org/10.3390/rs16142631
Vancouver
Li H, Tao H, Deng Q, Xiao S, Zhou J. Spatial Information Enhancement with Multi-Scale Feature Aggregation for Long-Range Object and Small Reflective Area Object Detection from Point Cloud. Remote Sensing. 2024;16(14):2631. doi:10.3390/rs16142631
BibTeX
@article{Li2024,
author = {Li, Hanwen and Tao, Huamin and Deng, Qiuqun and Xiao, Shanzhu and Zhou, Jianxiong},
title = {Spatial Information Enhancement with Multi-Scale Feature Aggregation for Long-Range Object and Small Reflective Area Object Detection from Point Cloud},
journal = {Remote Sensing},
year = {2024},
volume = {16},
number = {14},
pages = {2631},
doi = {10.3390/rs16142631}
}
EUVIMED