基于改进RT-DETR的低光照交通标志检测识别算法
Low-Light Traffic Sign Detection and Recognition Based on an Improved RT-DETR Network
摘要
针对夜间低光照环境下噪声干扰强造成关键特征丢失、交通标志检测精度下降的问题,文章提出一种基于改进实时端到端目标检测器(RT-DETR)的低光照交通标志检测识别算法.在特征提取阶段,引入轻量级网络RepViT重构主干网络,提升对低分辨率模糊目标的特征学习能力.设计低光照小目标增强模块(LSEM),采用通道级特征重标定与空间多尺度感知相协同,有效突出高亮通道响应并抑制暗区噪声.在解码器输入端加入非对称方向特征增强(ADFE)模块,利用大核条形卷积扩大方向感受野,减弱局部暗噪声对边缘特征的干扰,精准检测由反光材质保留的几何边界.在CNTSSS夜间交通标志数据集上的实验结果表明,改进模型的参数量和计算量相比未改进的分别减少了30.7%和27.8%,平均精度均值提升了2.1%,达到90.2%,在夜间交通标志检测任务中表现出较好的检测性能.
To address the problem of reduced traffic sign detection accuracy in intelligent driving systems under nighttime low-light conditions owing to strong noise interference and the loss of critical features,this paper proposes a low-light traffic sign detection and recognition algorithm based on an improved Real-Time Detection Transformer(RT-DETR)framework.In the feature extraction stage,a lightweight RepViT network is introduced to reconstruct the backbone network,thereby enhancing the feature learning capability for low-resolution and blurred targets.In addition,a Low-light Small-target Enhancement Module(LSEM)is designed to combine channel-level feature recalibration with spatial multiscale perception to effectively emphasize high-response channels while suppressing noise in dark regions.Furthermore,an Asymmetric Directional Feature Enhancement(ADFE)module is incorporated at the decoder input stage.By employing large-kernel strip convolutions to expand the directional receptive field,the module reduces the effect of local dark noise on edge features and improves the perception of geometric boundaries preserved by reflective materials.Experimental results on the CNTSSS nighttime traffic sign dataset demonstrate that the proposed model reduces the number of parameters and computational cost by 30.7%and 27.8%,respectively,while improving mAP@50 by 2.1%to 90.2%.The proposed method achieves favorable performance in nighttime traffic sign detection tasks.
年,卷(期):
2026, 47(5)
分类号:
TP391.4
机标分类号:
TP391.41
TN957.54
O481.2
在线出版日期:
2026年11月4日
基金项目:
62465006:国家自然科学基金QKHJCZD(2026)110:贵州省基础研究计划重点项目:贵州省科技计划项目006:贵州省科技计划项目QKH(2025)ZD021:贵州省科技计划项目,