计算机科学 ›› 2025, Vol. 52 ›› Issue (6A): 240900167-8.doi: 10.11896/jsjkx.240900167
白少康, 王宝会, 陈继轩
BAI Shaokang, WANG Baohui, CHEN Jixuan
摘要： 随着交通事业的不断发展,电动自行车在行驶过程中佩戴头盔的必要性得到了不断验证。同时,电动自行车驾驶佩戴头盔检测在深度学习领域也得到了广泛的研究。目前电动自行车驾驶佩戴头盔在密集场景下存在检测难度大、小目标检测困难等问题,同时为了实现更好在移动端部署需要选择轻量化的模型。因此,提出了一种基于YOLOv8n的电动自行车佩戴头盔检测改进算法。首先YOLOv8n是YOLOv8系列模型中最轻量的模型,能够更好地在移动端部署;此外在YOLOv8n主干网络引入可切换的空洞卷积,在不增加计算量的前提下提升了YOLOv8n在密集场景下提取特征的能力;在YOLOv8n的图像金字塔特征融合网络末尾融合三重注意力机制,加强对YOLOv8n模型对不同尺度特征融合信息中重要特征的提取能力;最后添加大尺寸特征信息与小尺寸特征信息融合,提升对小目标的检测效果。最终在保证YOLOv8n轻量化的同时,使用改进后的模型在自制的验证集中mAP@50、mAP@50-90、召回率分别提升2.7%,2.8%,2.5%,因此所提方法具有一定的实用意义及科研意义。
中图分类号:
| [1]YU Y,PENG J J,LI Q F,et al.Helmet wearing among e-bikers in Shanghai:On-site observational study[J].ShanghaiJournal of Preventive Medicine,2018(9):730-735 [2]LU X,CHEN D,LIU W C,et al.Characteristics of road traffic injuries caused by electric bicycles in Suzhou during 2017-2019[J].J Trauma Surg,2021(5):375-379. [3]KRIZHEVSKY A,SUTSKEVER I,HINTON G.ImageNetClassification with Deep Convolutional Neural Networks[J].Advances in Neural Information Processing Systems,2012,25(2). [4]GIRSHICK R,DONAHUE J,DARRELL T,et al,Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conferenceon ComputerVision and Pattern Recognition.Columbus:IEEE,2014:580-587. [5]REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards Real-Time Object Detection with Region Proposal Networks[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2017,39(6):1137-1149. [6]LIN T Y,DOLLAR P,GIRSHICK R,et al.Feature Pyramid Networks for Object Detection[C]//CVPR.2017:936-944. [7]HE K,GKIOXARI G,PIOTR D,et al.Mask R-CNN[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2017.2980-2988. [8]GE S M,GUO W J,LI C Y,et al.Masked Face Recognition with Generative-to-Discriminative Representations[J].arXiv:2405.16761,2024. [9]WANG G,YUAN Y,LI J,et al.Receptive Multi-GranularityRepresentation for Person Re-Identification[C]//IEEE Transactions on Image Processing.2020:6096-6109. [10]JIJI C V,SHINE L.Automated detection of helmet on motorcyclists from traffic surveillance videos:a comparative analysis using hand-crafted features and CNN[J].Multimedia Tools and Applications,2020:1-21. [11]MISRAA A K,MANOHAR NMISTRY J,AGARWAL M,et al.An automatic detection of helmeted and non-helmeted motorcyclist with license plate extraction using convolutional neural network[C]//Seventh International Conference on Image Processing Theory,Tools and Applications(IPTA).IEEE,2017:1-6. [12]CHATTERJI S,DASGUPTA M,BANDYOPADHYAY O.Automated helmet detection for multiple motorcycle riders using CNN[C]//IEEE Conference on Information and CommunicationTechnology.IEEE,2019:1-4. [13]XU H,DENG Z H.Helmet Wearing Detection Algorithm byImproved YOLOv5[J],Software Guide,2024,8:33-42. [14]ZHU X,LYU S,WANG X,et al.TPH-YOLOv5:ImprovedYOLOv5 Based on Transformer Prediction Head for Object Detection on Drone-captured Scenarios[J].arXiv.2108.11539,2021. [15]XU C G.Rearch real-time detection algorithm for helmets and reflective vests based on YOLOv5s[J].Computer Era,2023,10:110-115. [16]HAN K,WANG Y,TIAN Q,et al.GhostNet:More Features From Cheap Operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).IEEE,2020. [17]WANG J,CHEN K,XU R,et al.CARAFE:Content-Aware ReAssembly of Features[C]//2019 IEEE/CVF International Conference on Computer Vision.2019. [18]QIAO S,CHEN L C,YUILLE A.DetectoRS:Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution[J].arXiv:2006.02334,2020. [19]MISRA D,NALAMADA T,ARASANIPALAI A U,et al.Rotate to Attend:Convolutional Triplet Attention Module[J].ar-Xiv:2010.03045,2020. [20]HU J,SHEN L,SUN G,et al.Squeeze-and-Excitation Networks[C]//2018 IEEE/CVF Conference on Computer Vision and Patern Recognition.2018. [21]WOO S,PARK J,LEE J Y,et al.CBAM:Convolutional Block Attention Module[J].arXiv:1807.06521,2018. [22]PARK J,WOO S,LEE J Y,et al.BAM:Bottleneck Attention Module[J].arXiv:1807.06514,2018. |
|
||
首页