[1]院老虎,常玉坤,刘家夫.基于改进YOLOv5s的雾天场景车辆检测方法[J].郑州大学学报(工学版),2023,44(03):37-43.[doi:10.13705/j.issn.1671-6833.2023.03.005]
　YUAN Laohu,CHANG Yukun,LIU Jiafu.Vehicle Detection Method Based on Improved YOLOv5s in Foggy Scene[J].Journal of Zhengzhou University (Engineering Science),2023,44(03):37-43.[doi:10.13705/j.issn.1671-6833.2023.03.005]
点击复制

基于改进YOLOv5s的雾天场景车辆检测方法()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
44
期数:
2023年03期
页码:
37-43
栏目:
出版日期:
2023-04-30

文章信息/Info

Title:
Vehicle Detection Method Based on Improved YOLOv5s in Foggy Scene
作者:
院老虎; 常玉坤; 刘家夫
沈阳航空航天大学 航空宇航学院,辽宁 沈阳 110136

Author(s):
YUAN Laohu; CHANG Yukun; LIU Jiafu
Shenyang University of Aeronautics and Astronautics University of Aeronautics and Astronautics, Liaoning Shenyang 110136

关键词:
深度学习;  YOLOv5s;  车辆检测;  数据增强;  雾气模拟
Keywords:
deep learning;  YOLOv5s;  vehicle detection;  data augmentation;  fog simulation
分类号:
TP391
DOI:
10.13705/j.issn.1671-6833.2023.03.005
文献标志码:
A
摘要:
为了解决现有的目标检测方法在雾天场景下存在识别准确率低、易漏检的问题，提出一种改进 YOLOv5s 的雾天车辆检测方法。首先，以 VisDrone 数据集为基础，通过大气散射模型生成轻雾数据集( LightFogVisDrone) 和浓雾数据集( ThickFogVisDrone) ，并收集真实雾天场景图片组成混合浓度数据集( MixFogData) ; 其 次，对 原 始 YOLOv5s 的 Mosaic 数据增强方式进行改进，由原始 的 4 张 图 片 改 为 9 张图片进行随机剪切，减 少 灰 色 背 景面积，加快模型收敛，提高训练效率，在预测端之前添加 CBAM 注 意 力 机 制，以此来增强模型的图像特 征提取能力，改善遮挡目标与小目标的漏检问题; 最 后，优 化 NMS 非极大抑制值先验框，改善车辆目标的 漏检问题。实验结果表明: 与原 始 YOLOv5s 相 比，改 进 YOLOv5s 在 轻 雾、浓雾和混合雾气状态下的平均 精确率分别提高了 16. 14、16. 16 和 15. 05 百分点。改进 YOLOv5s 对于雾天环境下车辆目标的检测具有 有效性和实用性。
Abstract:
In order to solve the problems of low recognition accuracy and easy omission of existing target detection methods in foggy scenes, an improved vehicle detection method based on YOLOv5s was proposed. Firstly, based on VisDrone data set, LightFogVisDrone and ThickFogVisDrone were generated by atmospheric scattering model,and the MixFogData was composed of real fog scene pictures. Secondly, the Mosaic data enhancement method of the original model was improved from the original 4 pictures to 9 pictures randomly, which reduced the gray background area, accelerated the convergence of the model and improved the training efficiency, and the CBAM attention mechanism module was added before the prediction end to improve the feature extraction ability of the network to tackle the problem of missed detection of occluded targets and small targets. Finally, the prior frame of NMS non-maximum suppression value was optimized to improve the problem of missing detection of vehicle targets. The experimental results showed that, compared with the original YOLOv5s, the average accuracy of the improved YOLOv5s in light fog, dense fog and mixed fog was increased by 16.14, 16.16 and 15.05 percentage points, respectively, which proved that the improved YOLOv5s was effective and practical for vehicle target detection in foggy environment.

参考文献/References:

[1] 贺宇哲, 何宁, 张人, 等. 面向深度学习目标检测模型训练不平衡研究[J]. 计算机工程与应用, 2022, 58(5): 172-178.HE Y Z, HE N, ZHANG R, et al. Research on imba-lanced training of deep learning target detection model[J]. Computer Engineering and Applications, 2022, 58(5): 172-178.

[2] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
[3] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2016: 779-788.
[4] 院老虎, 翟柯嘉, 张泽鹏, 等. 基于模拟雾天遥感数据集的飞机目标检测研究[J]. 南京邮电大学学报(自然科学版), 2021, 41(3): 77-84.YUAN L H, ZHAI K J, ZHANG Z P, et al. Aircraft target detection based on fog simulation remote sensing image dataset[J]. Journal of Nanjing University of Posts and Telecommunications (Natural Science Edition), 2021, 41(3): 77-84.
[5] 王启明,何梓林,张栋林,等.基于YOLOv3的雾天场景行人车辆检测方法研究 [J/OL]. 控制工程, 2022,4(2):1-8(2022-02-17)[2022-11-14].https:∥doi.org/10.14107/j.cnki.kzgc.20211118.WANG Q M, HE Z L, ZHANG D L, et al. Research on pedestrian and vehicle detection method in foggy scene based on YOLOv3 [J/OL]. Control Engineering, 2022,4(2):1-8(2022-02-17)[2022-11-14]. https:∥doi.org/10.14107/j.cnki.kzgc.20211118.
[6] 陈琼红, 冀杰, 种一帆, 等. 基于AOD-Net和SSD的雾天车辆和行人检测[J]. 重庆理工大学学报(自然科学), 2021, 35(5): 108-117.CHEN Q H, JI J, CHONG Y F, et al. Vehicle and pedestrian detection based on AOD-Net and SSD algorithm in hazy environment[J]. Journal of Chongqing University of Technology (Natural Science), 2021, 35(5): 108-117.
[7] 李北明, 金荣璐, 徐召飞, 等. 基于特征蒸馏的改进Ghost-YOLOv5红外目标检测算法[J]. 郑州大学学报(工学版), 2022, 43(1): 20-26.LI B M, JIN R L, XU Z F, et al. An improved Ghost-YOLOv5 infrared target detection algorithm based on feature distillation[J]. Journal of Zhengzhou University (Engineering Science), 2022, 43(1): 20-26.
[8] ZHU P F, WEN L Y, DU D W, et al. Detection and tracking meet drones challenge[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(11): 7380-7399.
[9] 梁增龑, 刘本永. 基于主分量分析和大气散射模型的彩色图像雾霾快速去除算法[J]. 计算机应用, 2015, 35(2): 531-534.LIANG Z Y, LIU B Y. Fast algorithm for color image haze removal using principle component analysis and atmospheric scattering mode[J]. Journal of Computer Applications, 2015, 35(2): 531-534.
[10] 高隽, 褚擎天, 张旭东, 等. 结合光场深度估计和大气散射模型的图像去雾方法[J]. 光子学报, 2020, 49(7): 0710001.GAO J, CHU Q T, ZHANG X D, et al. Image dehazing method based on light field depth estimation and atmospheric scattering model[J]. Acta Photonica Sinica, 2020, 49(7): 0710001.
[11] HE K M, SUN J, TANG X O. Single image haze removal using dark channel prior[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(12): 2341-2353.
[12] WANG C Y, MARK LIAO H Y, WU Y H, et al. CSPNet: a new backbone that can enhance learning capability of CNN[C]∥2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Piscataway: IEEE, 2020: 1571-1580.
[13] LIN T Y, DOLLwidth=11,height=14,dpi=110R P, GIRSHICK R, et al. Feature py-ramid networks for object detection[C]∥2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2017: 936-944.
[14] LIU S, QI L, QIN H F, et al. Path aggregation network for instance segmentation[C]∥2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 8759-8768.
[15] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[M]∥Computer Vision-ECCV 2018. Cham: Springer International Publishing, 2018: 3-19.

相似文献/References:

[1]袁航,钟发海,聂上上,等.基于卷积神经网络的道路拥堵识别研究[J].郑州大学学报(工学版),2019,40(02):21.[doi:10.13705/j.issn.1671-6833.2019.02.008]
　LUO Ronghui,YUAN Hang,ZHONG Fahai,et al.The Research of Traffic Jam Detection Based on Convolutional Neural Network[J].Journal of Zhengzhou University (Engineering Science),2019,40(03):21.[doi:10.13705/j.issn.1671-6833.2019.02.008]
[2]朱俊丞,杨之乐,郭媛君,等.深度学习在电力负荷预测中的应用综述[J].郑州大学学报(工学版),2019,40(05):12.[doi:10.13705/j.issn.1671-6833.2019.05.005]
　Zhu Juncheng,Young Joy,Guo Yuanjun,et al.A review of the application of deep learning in power load forecasting[J].Journal of Zhengzhou University (Engineering Science),2019,40(03):12.[doi:10.13705/j.issn.1671-6833.2019.05.005]
[3]黄文锋,徐珊珊,孙燚,等.基于多分辨率卷积神经网络的火焰检测[J].郑州大学学报(工学版),2019,40(05):79.[doi:10.13705/j.issn.1671-6833.2019.05.022]
　Huang Wenfeng,Susan Hsu,Sun Yi,et al.Fire Detection Based on Multi-resolution Convolution Neural Network in Various Scenes[J].Journal of Zhengzhou University (Engineering Science),2019,40(03):79.[doi:10.13705/j.issn.1671-6833.2019.05.022]
[4]陈义飞、郭胜、潘文安、陆彦辉.基于多源传感器数据融合的三维场景重建[J].郑州大学学报(工学版),2021,42(02):81.[doi:10.13705/j.issn.1671-6833.2021.02.008]
　Chen Yifei,Guo Sheng,Pan Wenan,et al.3D Scene Reconstruction Based on Multi-source Sensor Data Fusion[J].Journal of Zhengzhou University (Engineering Science),2021,42(03):81.[doi:10.13705/j.issn.1671-6833.2021.02.008]
[5]李学相,曹淇,刘成明.基于无配对生成对抗网络的图像超分辨率重建[J].郑州大学学报(工学版),2021,42(05):1.[doi:10.13705/j.issn.1671-6833.2021.05.018]
　LI Xuexiang,CAO Qi,LIU Chengming.Image Super-resolution Based on No Match Generative Adversarial Network[J].Journal of Zhengzhou University (Engineering Science),2021,42(03):1.[doi:10.13705/j.issn.1671-6833.2021.05.018]
[6]王希鹏,李永,李智,等.融合图像深度的抗遮挡目标跟踪算法[J].郑州大学学报(工学版),2021,42(05):19.[doi:10.13705/j.issn.1671-6833.2021.05.011]
　Wang Xipeng,Li Yong,Li Zhi,et al.Anti-occlusion Target Tracking Algorithm Based on Image Depth[J].Journal of Zhengzhou University (Engineering Science),2021,42(03):19.[doi:10.13705/j.issn.1671-6833.2021.05.011]
[7]卢晨辉,冯硕,易爱华,等.基于深度学习的加油站销量预测与营销策略应用研究[J].郑州大学学报(工学版),2022,43(01):1.[doi:10.13705/j.issn.1671-6833.2022.01.014]
　LU Chenhui,FENG Shuo,YI Aihua,et al.Gasoline Station Sales Prediction Method Based on Deep Learning and Its Application of Promotion Strategy[J].Journal of Zhengzhou University (Engineering Science),2022,43(03):1.[doi:10.13705/j.issn.1671-6833.2022.01.014]
[8]陈浩杰,黄锦,左兴权,等.基于宽度&深度学习的基站网络流量预测方法[J].郑州大学学报(工学版),2022,43(01):7.[doi:10.13705/j.issn.1671-6833.2022.01.011]
　CHEN Haojie,HUANG Jin,ZUO Xingquan,et al.Base Station Network Traffic Prediction Method Based on Wide & Deep Learning[J].Journal of Zhengzhou University (Engineering Science),2022,43(03):7.[doi:10.13705/j.issn.1671-6833.2022.01.011]
[9]成科扬,荣兰,蒋森林,等.基于深度学习的遥感图像超分辨率重建技术综述[J].郑州大学学报(工学版),2022,43(05):8.[doi:10.13705/j.issn.1671-6833.2022.05.013]
　CHENG Keyang,RONG Lan,JIANG Senlin,et al.Overview of Methods for Remote Sensing Image Super-resolution Reconstruction Based on Deep Learning[J].Journal of Zhengzhou University (Engineering Science),2022,43(03):8.[doi:10.13705/j.issn.1671-6833.2022.05.013]
[10]高宇飞,马自行,徐 静,等.基于卷积和可变形注意力的脑胶质瘤图像分割[J].郑州大学学报(工学版),2024,45(02):27.[doi:10.13705/j.issn.1671-6833.2023.05.007]
　GAO Yufei,MA Zixing,XU Jing,et al.Brain Glioma Image Segmentation Based on Convolution and Deformable Attention[J].Journal of Zhengzhou University (Engineering Science),2024,45(03):27.[doi:10.13705/j.issn.1671-6833.2023.05.007]
[11]王 瑜,毕 玉,石健彤,等.基于注意力与多级特征融合的 YOLOv5 算法[J].郑州大学学报(工学版),2024,45(03):38.[doi:10. 13705 / j. issn. 1671-6833. 2023. 06. 009]
　LIU Xin,XU Hongzhen,LIU Aihua,et al.Geological Named Entity Recognition Based on MacBERT and R-Drop[J].Journal of Zhengzhou University (Engineering Science),2024,45(03):38.[doi:10. 13705 / j. issn. 1671-6833. 2023. 06. 009]

更新日期/Last Update: 2023-05-08