基于改进YOLOv5s-CBAM-ASFF算法的交通标志识别研究

  • 打印
  • 收藏
收藏成功


打开文本图片集

关键词:深度学习 交通标志识别注意力机制多尺度特征融合YOLOv5s中图分类号:U463.6;TP391.41 文献标志码:A DOI: 10.20104/j.cnki.1674-6546.20240263

【Abstract】To achieve more efficient detection of smalltraffic sign targetsundercomplex urban stret background conditions,this paperproposesanimprovedYOLOv5salgorithm.Thisenhancementisachievedbyicorporatinga Convolution Block Attention Module (CBAM) Spatial Channel Attention Mechanism,an Adaptive Spatial Feature Fusion (ASFF) module, andanimproved loss function fordetection boxes.Thevalidationresultsonthe T1OoKtrafic signdatasetdemonstrate that the proposed algorithm achieves a mean Average Precision (mAP) of 84.5% in traffic sign recognition.

Keywords:Deep learning,Traffic signrecognition,Attention mechanism,Multi-scale feature fusion,O5s

【引用格式】付蓉萍,付建胜,梁旺阳.基于改进YOLOv5s-CBAM-ASFF算法的交通标志识别研究[J].汽车工程师,2025(8): 22-28.FURP,FUJS,LIANGWY.Research on Traffic SignRecognition Based on the ImprovedYOLOv5s-CBAM-ASFFAlgorithm[J].Automotive Engineer,2025(8):22-28.

1前言

在城市街道场景中,交通标志的识别往往存在复杂的背景干扰和标志被部分遮挡等因素,严重影响检测速度和准确性。(剩余10950字)

monitor