《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (8): 2378-2385.DOI: 10.11772/j.issn.1001-9081.2021061005

• 人工智能 • 上一篇    

基于改进注意力机制的交通标志检测算法

张新宇1,2, 丁胜1,2(), 杨治佩1,2   

  1. 1.武汉科技大学 计算机科学与技术学院, 武汉 430065
    2.智能信息处理与实时工业系统湖北省重点实验室(武汉科技大学), 武汉 430065
  • 收稿日期:2021-06-15 修回日期:2021-10-03 接受日期:2021-10-18 发布日期:2022-01-25 出版日期:2022-08-10
  • 通讯作者: 丁胜
  • 作者简介:张新宇(1996—),男,河南焦作人,硕士研究生,主要研究方向:计算机视觉、深度学习;
    丁胜(1975—),男,湖北武汉人,副教授,博士,CCF会员,主要研究方向:计算机视觉;
    杨治佩(1996—),男,甘肃庆阳人,硕士研究生,主要研究方向:计算机视觉、深度学习。
  • 基金资助:
    国家自然科学基金资助项目(61806150)

Traffic sign detection algorithm based on improved attention mechanism

Xinyu ZHANG1,2, Sheng DING1,2(), Zhipei YANG1,2   

  1. 1.School of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan Hubei 430065,China
    2.Hubei Province Key Laboratory of Intelligent Information Processing and Real?time Industrial System (Wuhan University of Science and Technology),Wuhan Hubei 430065,China
  • Received:2021-06-15 Revised:2021-10-03 Accepted:2021-10-18 Online:2022-01-25 Published:2022-08-10
  • Contact: Sheng DING
  • About author:ZHANG Xinyu, born in 1996, M. S. candidate. His research interests include computer vision, deep learning.
    DING Sheng, born in 1975, Ph. D., associate professor. His research interests include computer vision.
    YANG Zhipei, born in 1996, M. S. candidate. His research interests include computer vision, deep learning.
  • Supported by:
    National Natural Science Foundation of China(61806150)

摘要:

针对交通标志在某些场景中存在分辨率过低、被覆盖等环境因素影响导致在目标检测任务中出现漏检、误检的情况,提出一种基于改进注意力机制的交通标志检测算法。首先,针对交通标志因破损、光照等环境影响造成图像分辨率低从而导致网络提取图像特征信息有限的问题,在主干网络中添加注意力模块以增强目标区域的关键特征;其次,特征图中相邻通道间的局部特征由于感受野重叠而存在一定的相关性,用大小为k的一维卷积代替通道注意力模块中的全连接层,以达到聚合不同通道信息和减少额外参数量的作用;最后,在路径聚合网络(PANet)的中、小尺度特征层引入感受野模块来增大特征图的感受野以融合目标区域的上下文信息,从而提升网络对交通标志的检测能力。在中国交通标志检测数据集(CCTSDB)上的实验结果表明,所提出的YOLOv4(You Only Look Once v4)改进算法在引进极少的参数量与原算法检测速度相差不大的情况下,平均精确率均值(mAP)达96.88%,mAP提升了1.48%;而与轻量级网络YOLOv5s相比,在单张检测速度慢10 ms的情况下,所提算法mAP比YOLOv5s高3.40个百分点,检测速度达到40?frame/s,说明该算法完全满足目标检测实时性的要求。

关键词: 注意力机制, 一维卷积, 感受野模块, 特征提取网络, YOLOv4

Abstract:

In some scenes, the low resolution, coverage and other environmental factors of traffic signs lead to missed and false detections in object detection tasks. Therefore, a traffic sign detection algorithm based on improved attention mechanism was proposed. First of all, in response to the problem of low image resolution due to damage, lighting and other environmental impacts of traffic signs, which leaded to the limited extraction of image feature information by the network, an attention module was added to the backbone network to enhance the key features of the object area. Secondly, the local features between adjacent channels in the feature map had a certain correlation due to the overlap of the receptive fields, a one-dimensional convolution of size k was used to replace the fully connected layer in the channel attention module to aggregate different channel information and reduce the number of additional parameters. Finally, the receptive field module was introduced in the medium- and small-scale feature layers of Path Aggregation Network (PANet) to increase the receptive field of the feature map to fuse the context information of the object area and improve the network’s ability to detect traffic signs. Experimental results on CSUST Chinese Traffic Sign Detection Benchmark (CCTSDB) dataset show that the proposed improved You Only Look Once v4 (YOLOv4) algorithm achieve an average detection speed with a small amount of parameters introduced and the detection speed is not much different from that of the original algorithm. The mean Accuracy Precision (mAP) reached 96.88%, which was increased by 1.48%; compared with the lightweight network YOLOv5s, with the single frame detection speed of 10?ms slower, the mAP of the proposed algorithm is 3.40 percentage points higher than that of YOLOv5s, and the speed reached 40?frame/s, indicating that the algorithm meets the real-time requirements of object detection completely.

Key words: attention mechanism, one-dimensional convolution, receptive field block, feature extraction network, You Only Look Once v4 (YOLOv4)

中图分类号: