面向道路异常检测的自适应低光增强：多架构基准与边缘部署研究
Adaptive Low-Light Enhancement for Road Anomaly Detection: A Multi-Architecture Benchmark and Edge Deployment Study
Obed Honour Eje · Oluwatosin Kolade · Glory Bagai · Taliha Abiodun Folorunso
中文摘要
自动驾驶车辆上的道路异常检测系统必须在黑暗、降雨和 Harmattan 尘霾等退化视觉条件下可靠工作。该论文在物理建模的光照与天气退化条件下，对 YOLOv8n、YOLOv12n、RT-DETR 和 Faster R-CNN 四种目标检测架构进行基准测试，并表明统一应用 SCI 低光增强网络在黄昏级光照下反而有害（mAP@50 下降 16.1% 至 30.7%）。提出一种轻量、免训练的亮度自适应路由器，依据每帧 Rec. 601 亮度在不增强、适度 SCI、去噪加强力 SCI 三种模式间选择，在保留低光增益的同时挽回大部分黄昏性能损失，并在 NVIDIA Jetson Nano 边缘硬件上验证完整流水线。
关键要点
- 01问题：自动驾驶道路异常检测在黑暗、降雨、尘霾等退化视觉条件下可靠性不足。
- 02方法：对 YOLOv8n、YOLOv12n、RT-DETR、Faster R-CNN 在物理建模光照与天气退化下做基准测试。
- 03结果：统一应用 SCI 低光增强在黄昏级光照下反而使 mAP@50 下降 16.1% 至 30.7%。
- 04方法：提出免训练亮度自适应路由器，按每帧 Rec. 601 亮度选择三档增强策略。
- 05结果：在 NVIDIA Jetson Nano 边缘硬件上验证完整流水线，挽回大部分黄昏性能损失。
解读
尚无解读。
原始英文摘要
arXiv:2611.08393v1 Announce Type: new Abstract: Road anomaly detection systems on autonomous vehicles must function reliably under degraded visual conditions including darkness, rain, and Harmattan dust haze. This paper benchmarks four object detection architectures, YOLOv8n, YOLOv12n, RT-DETR, and Faster R-CNN, under physically grounded lighting and weather degradation, and shows that uniform application of the SCI low-light enhancement network is actively harmful at dusk-level lighting (-16.1% to -30.7% mAP@50). We propose a lightweight, training-free luminance-adaptive router that selects between no enhancement, moderate SCI, and denoising plus strong SCI based on per-frame Rec. 601 luminance, recovering most of the dusk penalty while preserving low-light gains, and we validate the full pipeline on NVIDIA Jetson Nano edge hardware.