改进YOLOv5的小目标交通标志实时检测算法

doi:10.3778/j.issn.1002-8331.2206-0503

摘要/Abstract

摘要： 在真实场景下准确实时检测小目标交通标志对自动驾驶有重要意义，针对YOLOv5算法检测小目标交通标志精度低的问题，提出一种基于改进YOLOv5的小目标交通标志实时检测算法。借鉴跨阶段局部网络思想，在YOLOv5的空间金字塔池化上设置新的梯度路径，强化特征提取能力；在颈部特征融合中增设深、浅卷积特征的可学习自适应权重，更好地融合深层语义和浅层细节特征，提高小目标交通标志的检测精度。为验证所提算法的优越性，在TT100K交通标志数据集上进行了实验验证。实验结果表明所提算法在小目标交通标志上的平均精度均值（mean average precision，mAP）为77.3%，比原始YOLOv5提升了5.4个百分点，同时也优于SSD、RetinaNet、YOLOX、SwinTransformer等算法的检测结果。所提算法的运行速度为46.2?frame/s，满足检测实时性的要求。

关键词: 交通标志检测, 自适应特征融合, 小目标, 特征提取, YOLOv5

Abstract: Accurate real-time detection of small target traffic signs in real scenarios is important for autonomous driving. To address the problem of low accuracy of YOLOv5 algorithm in detecting small target traffic signs, a real-time detection algorithm for small target traffic signs based on improved YOLOv5 is proposed. Drawing on the idea of cross-stage local networks, another gradient path is set on the spatial pyramid pooling of YOLOv5 to strengthen the feature extraction capability; the learnable adaptive weights of deep and shallow convolutional features are added to the neck feature fusion to better fuse deep semantic and shallow detail features and improve the detection accuracy of small target traffic signs. To verify the superiority of the proposed algorithm, experimental validation is carried out on the TT100K traffic sign dataset. The experimental results show that the mean average precision（mAP） of the proposed algorithm on small-target traffic signs is 77.3%, which is 5.4 percentage points better than the original YOLOv5, and also outperforms the detection results of SSD, RetinaNet, YOLOX and Swin Transformer. The proposed algorithm runs at 46.2 frame/s, meeting the requirements for real-time detection.

Key words: traffic signs detection, adaptive feature fusion, small object, feature extraction, YOLOv5

胡均平, 王鸿树, 戴小标, 高小林. 改进YOLOv5的小目标交通标志实时检测算法[J]. 计算机工程与应用, 2023, 59(2): 185-193.

HU Junping, WANG Hongshu, DAI Xiaobiao, GAO Xiaolin. Real-Time Detection Algorithm for Small-Target Traffic Signs Based on Improved YOLOv5[J]. Computer Engineering and Applications, 2023, 59(2): 185-193.

参考文献

[1] DALAL N，TRIGGS B.Histograms of oriented gradients for human detection[C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition，Piscataway，2005：886-893.
[2] LOWE D G.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision，2004，60（2）：91-110.
[3] 伍晓晖，田启川.交通标志识别方法综述[J].计算机工程与应用，2020，56（10）：20-26.
WU X H，TIAN Q C.Survey of traffic sign recognition[J].Computer Engineering and Applications，2020，56（10）：20-26.
[4] REN S，HE K，GIRSHICK R，et al.Faster R-CNN：towards real-time object detection with region proposal networks[C]//Advances in Neural Information Processing Systems，2015：91-99.
[5] LIU W，ANGUELOV D，ERHAN D，et al.SSD：single shot multibox detector[C]//European Conference on Computer Vision，Amsterdam，2016：21-37.
[6] LIN T Y，GOYAL P，GIRSHICK R，et al.Focal loss for dense object detection[C]//Proceedings of the IEEE International Conference on Computer Vision，Venice，2017：2980-2988.
[7] REDMON J，FARHADI A.YOLO9000：better，faster，stronger[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，Honolulu，2017：7263-7271.
[8] REDMON J，FARHADI A.Yolov3：an incremental improvement[J].arXiv：1804.02767，2018.
[9] BOCHKOVSKIY A，WANG C Y，LIAO H Y M.Yolov4：optimal speed and accuracy of object detection[J].arXiv：2004.10934，2020.
[10] 郭璠，张泳祥，唐琎，等.YOLOv3-A：基于注意力机制的交通标志检测网络[J].通信学报，2021，42（1）：87-99.
GUO F，ZHANG Y X，TANG J，et al.YOLOv3-a：a traffic sign detection network based on attention mechanism[J].Journal on Communications，2021，42（1）：87-99.
[11] LIU Z，QI M，SHEN C，et al.Cascade saccade machine learning network with hierarchical classes for traffic sign detection[J].Sustainable Cities and Society，2021，67：102700.
[12] LIU Z，LI D，GE S S，et al.Small traffic sign detection from large image[J].Applied Intelligence，2020，50（1）：1-13.
[13] HE K，ZHANG X，REN S，et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2015，37（9）：1904-1916.
[14] LIU S，QI L，QIN H，et al.Path aggregation network for instance segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，Salt Lake City，2018：8759-8768.
[15] GHIASI G，CUI Y，SRINIVAS A，et al.Simple copy-paste is a strong data augmentation method for instance segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition，2021：2918-2928.
[16] WANG C Y，LIAO H Y M，WU Y H，et al.CSPNet：a new backbone that can enhance learning capability of CNN[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops，Seattle，2020：390-391.
[17] ZHU Z，LIANG D，ZHANG S，et al.Traffic-sign detection and classification in the wild[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，Las Vegas，2016：2110-2118.
[18] GE Z，LIU S，WANG F，et al.Yolox：exceeding YOLO series in 2021[J].arXiv：2107.08430，2021.

[19] LIU Z，LIN Y，CAO Y，et al.Swin transformer：hierarchical vision transformer using shifted windows[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision，Montreal，2021：10012-10022.

[20] TARACHANDY S M，ARAVINTH J.Enhanced local features using ridgelet filters for traffic sign detection and recognition[C]//2021 Second International Conference on Electronics and Sustainable Communication Systems（ICESC），2021：1150-1156.