基于Dim env-YOLO算法的昏暗场景车辆多目标检测

doi:10.19678/j.issn.1000-3428.0063769

摘要/Abstract

摘要： 低照度的夜间路况复杂，现有夜间车辆识别相关研究较少，且存在识别方法实时性不高、过多占用硬件资源等不足。针对夜间场景车辆识别干扰因素较多、检测效果不佳的问题，提出一种基于YOLOv4的Dim env-YOLO车辆目标检测算法。利用MobileNetV3网络替换原始YOLOv4中的主干网络，以减少模型参数量。在改进的YOLOv4模型上使用图像暗光增强方法，提高车辆目标在昏暗环境中的可识别性。在此基础上，引入注意力机制加强特征信息选择，同时利用深度可分离卷积降低网络计算量。选取北京部分道路的夜间场景图片自制数据集并进行实验验证，结果表明，在存在高斯噪声、模糊扰动、雨雾夜晚等情况下，Dim env-YOLO算法的测试结果较稳定，对于照度低于30 lx的昏暗条件下的车流，其检测mAP值达到90.49%，对于最常见的轿车类别，mAP值达到96%以上，优于Faster-RCNN、YOLOv3、YOLOv4等网络模型在昏暗光照条件下的检测效果。

关键词: 昏暗场景, 车辆检测, 深度可分离卷积, Dim env-YOLO算法, MobileNetV3网络

Abstract: Nighttime road scenes with low illuminance are complex, and visibility is typically poor.However, relatively few works have considered vehicle recognition under these conditions, and existing methods require considerable computational resources and cannot operate in real time.To address these problems, a Dim env-YOLO vehicle target detection algorithm based on YOLOv4 is proposed.Use MobileNetV3 network to replace the backbone network in the original YOLOv4 to reduce the amount of model parameters.The improved YOLOv4 model performs low-light image enhancement to improve the identifiability of vehicle targets in dark environments.An attention mechanism is implemented to enhance the selection of feature information, and depthwise separable convolution is used to reduce computational resource requirements.A dataset of images of nighttime road scenes in Beijing was collected to experimentally evaluate the proposed approach.The results show that the performance of the Dim env-YOLO algorithm was relatively stable under conditions of Gaussian noise, fuzzy disturbance, rain, fog, and so forth, and it achieved an mAP value of up to 90.49% in recognizing vehicle flow under dim condition with illumination of less than 30 lx.For the most common category representing cars, the mAP value of the proposed method exceeded 96%, which outperformed conventional methods including Faster-RCNN, YOLOv3, YOLOv4, and other network models.

Key words: dim scene, vehicle detection, depthwise separable convolution, Dim env-YOLO algorithm, MobileNetV3 network

中图分类号:

TP391.41

郭克友, 王苏东, 李雪, 张沫. 基于Dim env-YOLO算法的昏暗场景车辆多目标检测[J]. 计算机工程, 2023, 49(3): 312-320.

GUO Keyou, WANG Sudong, LI Xue, ZHANG Mo. Multi-Target Detection of Vehicles in Dim Scenes Based on Dim env-YOLO Algorithm[J]. Computer Engineering, 2023, 49(3): 312-320.

https://www.ecice06.com/CN/Y2023/V49/I3/312

图/表 21

20230314190956

20230314191000

20230314191003

20230314191007

20230314191010

20230314191013

20230314191016

20230314191019

20230314191022

20230314191026

20230314191029

20230314191033

20230314191036

20230314191039

20230314191043

20230314191053

20230314191056

20230314191100

20230314191104

20230314191107

20230314191111

参考文献

[1] 李国进, 胡洁, 艾矫燕.基于改进SSD算法的车辆检测[J].计算机工程, 2022, 48(1):266-274. LI G J, HU J, AI J Y.Vehicle detection based on improved SSD algorithm[J].Computer Engineering, 2022, 48(1):266-274.(in Chinese)
[2] TIAN D X, ZHANG C, DUAN X T, et al.An automatic car accident detection method based on cooperative vehicle infrastructure systems[J].IEEE Access, 2019, 7:127453-127463.
[3] 刘鹏.公路隧道非对称光强空间分布效能分析研究[D].重庆:重庆交通大学, 2020. LIU P.Analysis and research on spatial distribution efficiency of asymmetric light intensity of highway tunnel[D].Chongqing:Chongqing Jiaotong University, 2020.(in Chinese)
[4] FENG D, HAASE-SCHÜTZ C, ROSENBAUM L, et al.Deep multi-modal object detection and semantic segmentation for autonomous driving:datasets, methods, and challenges[J].IEEE Transactions on Intelligent Transportation Systems, 2021, 22(3):1341-1360.
[5] 黄靖淞, 左颢睿, 张建林.轻量化目标检测算法研究及应用[J].计算机工程, 2021, 47(10):236-241. HUANG J S, ZUO H R, ZHANG J L.Research and application of lightweight object detection algorithm[J].Computer Engineering, 2021, 47(10):236-241.(in Chinese)
[6] 李柯泉, 陈燕, 刘佳晨, 等.基于深度学习的目标检测算法综述[J].计算机工程, 2022, 48(7):1-12. LI K Q, CHEN Y, LIU J C, et al.Survey of deep learning-based object detection algorithms[J].Computer Engineering, 2022, 48(7):1-12.(in Chinese)
[7] MEKHALFI M L, NICOLÒ C, BAZI Y, et al.Detecting crop circles in Google earth images with Mask R-CNN and YOLOv3[J].Applied Sciences, 2021, 11(5):2238.
[8] GAO X W, LI S Q, JIN B Y, et al.Intelligent crack damage detection system in shield tunnel using combination of retinanet and optimal adaptive selection[J].Journal of Intelligent & Fuzzy Systems, 2021, 40(3):4453-4469.
[9] HOESER T, KUENZER C.Object detection and image segmentation with deep learning on earth observation data:a review-part I:evolution and recent trends[J].Remote Sensing, 2020, 12(10):1667.
[10] HU X W, XU X M, XIAO Y J, et al.SINet:a scale-insensitive convolutional neural network for fast vehicle detection[J].IEEE Transactions on Intelligent Transportation Systems, 2019, 20(3):1010-1019.
[11] JU M, LUO H, WANG Z, et al.The application of improved YOLOv3 in multi-scale target detection[J].Applied Sciences, 2019, 9(18):3775.
[12] ZHANG Z, XIA S H, CAI Y X, et al.A soft-YoloV4 for high-performance head detection and counting[J].Mathematics, 2021, 9(23):3096.
[13] ZHANG S F, WEN L Y, BIAN X, et al.Single-shot refinement neural network for object detection[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:4203-4212.
[14] GUO K, HE C, YANG M, et al.A pavement distresses identification method optimized for YOLOv5s[J].Scientific Reports, 2022, 12:3542.
[15] HE K M, GKIOXARI G, DOLLÁR P, et al.Mask R-CNN[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:10-15.
[16] PEI D S.A fast RetinaNet fusion framework for multi-spectral pedestrian detection[J].Infrared Physics & Technology, 2020, 105:103178.
[17] REDMON J, DIVVALA S, GIRSHICK R, et al.You only look once:unified, real-time object detection[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2016:779-788.
[18] HE K M, ZHANG X Y, REN S Q, et al.Deep residual learning for image recognition[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2016:770-778.
[19] LIU S, QI L, QIN H F, et al.Path aggregation network for instance segmentation[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:8759-8768.
[20] 李贝贝, 彭力, 戴菲菲.结合马氏距离与自编码器的网络流量异常检测方法[J].计算机工程, 2022, 48(4):133-142. LI B B, PENG L, DAI F F.Abnormal network traffic detection method combining mahalanobis distance and autoencoder[J].Computer Engineering, 2022, 48(4):133-142.(in Chinese)
[21] 刘通, 程江华, 华宏虎, 等.结合YdUaVa颜色模型和改进MobileNetV3的视频烟雾检测方法[J].国防科技大学学报, 2021, 43(5):80-85. LIU T, CHENG J H, HUA H H, et al.Video smoke detection method combining YdUaVa color model and improved MobileNetV3[J].Journal of National University of Defense Technology, 2021, 43(5):80-85.(in Chinese)
[22] 许德刚, 王露, 李凡.深度学习的典型目标检测算法研究综述[J].计算机工程与应用, 2021, 57(8):10-25. XU D G, WANG L, LI F.Review of typical object detection algorithms for deep learning[J].Computer Engineering and Applications, 2021, 57(8):10-25.(in Chinese)
[23] 常青, 韩文, 王清华, 等.改进YOLO轻量化网络的行人检测算法[J].光学技术, 2022, 48(1):80-85. CHANG Q, HAN W, WANG Q H, et al.Pedestrian detection algorithm based on improved YOLO lightweight network[J].Optical Technique, 2022, 48(1):80-85.(in Chinese)
[24] 李琳.低光照图像目标检测关键技术研究[D].合肥:合肥工业大学, 2021. LI L.Research on key technology of low illumination image object detection[D].Hefei:Hefei University of Technology, 2021.(in Chinese)
[25] 赵恒.基于公路交通场景的低质量图像增强技术研究[D].西安:长安大学, 2021. ZHAO H.Research on low quality image enhancement technology based on road traffic scenes[D].Xi'an:Changan University, 2021.(in Chinese)
[26] ZHU X Z, HU H, LIN S, et al.Deformable ConvNets V2:more deformable, better results[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2019:9300-9308.
[27] 周卓峰.基于轻量化YOLO的交通标志检测方法研究[D].西安:长安大学, 2021. ZHOU Z F.Research on traffic sign detection method based on lightweight YOLO[D].Xi'an:Changan University, 2021.(in Chinese)
[28] 牟亮, 赵红, 李燕, 等.基于梯度压缩的YOLO v4算法车型识别[J].工程科学学报, 2022, 44(5):940-950. MU L, ZHAO H, LI Y, et al.Vehicle recognition based on gradient compression and YOLO v4 algorithm[J].Chinese Journal of Engineering, 2022, 44(5):940-950.(in Chinese)
[29] 江金洪, 鲍胜利, 史文旭, 等.基于YOLO v3算法改进的交通标志识别算法[J].计算机应用, 2020, 40(8):2472-2478. JIANG J H, BAO S L, SHI W X, et al.Improved traffic sign recognition algorithm based on YOLO v3 algorithm[J].Journal of Computer Applications, 2020, 40(8):2472-2478.(in Chinese)

选择文件类型/文献管理软件名称

选择包含的内容