复杂环境下高速服务区禁停检测算法

doi:10.19678/j.issn.1000-3428.0069202

摘要/Abstract

摘要：

高速服务区车辆禁停检测面临场景复杂的挑战, 目前基于单一车型信息的禁停检测算法受到昼夜光照变化、车辆重叠遮挡以及视角倾斜等因素的影响, 无法在连续时间段内准确判断禁停区域内的车辆是否为同一辆车, 存在较高的误检率和漏检率。为此, 提出一种复杂环境下高速服务区禁停检测算法, 首先使用YOLOv5车辆检测算法获取车型与位置信息, 并给出一种改进禁停区域匹配方法, 提升复杂环境下目标和区域匹配准确率; 其次对禁停区域内目标车辆使用ConvNeXt车身颜色识别算法获取颜色信息; 最后设计一种分阶段控制的多维信息匹配与融合策略, 有效降低因光照、遮挡、视角等变化导致的车辆身份判断不准确情况, 从而降低服务区车辆禁停检测的误报率与漏检率。实验结果表明, 所提方法在高速服务区禁停数据集上的禁停误检率由3.56%下降到0.77%, 禁停漏检率由11.3%下降到2.48%, 不仅为服务区管理禁停行为提供了车辆多属性信息和车辆违停时长信息, 而且满足服务区多场景部署的性能要求, 可较好地用于实际应用中。

关键词: ConvNeXt网络, 车身颜色识别, YOLOv5算法, 车辆检测, 禁停检测

Abstract:

Prohibited-parking detection in highway service areas is rendered complicated by complex environments. Current algorithms based on single-vehicle information are affected by factors such as diurnal lighting changes, vehicle overlap and occlusion, and viewing-angle variations, which prevent accurate judgment of whether vehicles within the prohibited parking area are the same over consecutive time periods, thus resulting in high false-positive and -negative rates. Hence, we propose a method for prohibited-parking detection in highway service areas under complex conditions, which involves the following steps: First, the YOLOv5 vehicle-detection algorithm is used to obtain the vehicle type and location information, and an improved matching method for prohibited parking areas is proposed that effectively enhances the accuracy of target and area matching in complex environments. Second, the ConvNeXt vehicle-body-color recognition algorithm is employed to obtain the color information for vehicle targets within the prohibited parking area. Finally, a multistage control strategy for multidimensional information matching and fusion is designed that effectively reduces inaccuracies in vehicle-identity judgment caused by changes in lighting, occlusions, and viewing angles, thereby reducing false reporting and missed-detection rates in prohibited-parking detection. Experimental results show that the proposed method reduces the false-positive rate from 3.56% to 0.77% and the false-negative rate from 11.3% to 2.48% on a highway service area prohibited-parking dataset. Moreover, it not only provides multi-attribute information and the parking duration of vehicles for the management of prohibited-parking behavior in service areas but also satisfies the performance requirements for multiscenario deployment in service areas, thus demonstrating its practical applicability.

Key words: ConvNeXt network, vehicle color recognition, YOLOv5 algorithm, vehicle detection, prohibited parking detection

安国成, 王晓龙, 江波, 幸健. 复杂环境下高速服务区禁停检测算法[J]. 计算机工程, 2025, 51(2): 356-364.

AN Guocheng, WANG Xiaolong, JIANG Bo, XING Jian. Prohibited Parking Detection Algorithm for Highway Service Area in Complex Environment[J]. Computer Engineering, 2025, 51(2): 356-364.

https://www.ecice06.com/CN/Y2025/V51/I2/356

图/表 14

图1 禁停检测算法流程

Fig.1 Procedure of prohibited parking detection algorithm

图2 YOLOv5网络结构

Fig.2 Structure of YOLOv5 network

图3 射线法示意图

Fig.3 Schematic diagram of the ray method

图4 大型车辆检测框中心点不在RoI区域内

Fig.4 The center point of the detection box for large vehicles is not within the ROI

图5 ConvNeXt车身颜色识别算法网络结构

Fig.5 Network structure of ConvNeXt vehicle color recognition algorithm

图6 复杂环境下的高速服务区车辆样本情况

Fig.6 Vehicle sample situation for highway service area in complex environment

图7 服务区车型数据集样本分布

Fig.7 Sample distribution of vehicle types dataset in service area

图8 复杂环境下高速服务区车辆检测结果

Fig.8 Detection results of vehicles in high-speed service areas in complex environments

图9 车身颜色样本分布

Fig.9 Sample distribution of vehicle color samples

图10 样本尺寸密度图

Fig.10 Density of samples size

参考文献 28

1	郭克友, 王苏东, 李雪, 等. 基于Dim env-YOLO算法的昏暗场景车辆多目标检测. 计算机工程, 2023, 49 (3): 312- 320. doi: 10.19678/j.issn.1000-3428.0063769
	GUO K Y , WANG S D , LI X , et al. Multi-target detection of vehicles in dim scenes based on dim env-YOLO algorithm. Computer Engineering, 2023, 49 (3): 312- 320. doi: 10.19678/j.issn.1000-3428.0063769
2	李松江, 耿兰兰, 王鹏. 基于改进Yolov4的车辆目标检测. 计算机工程, 2023, 49 (4): 272- 280. doi: 10.19678/j.issn.1000-3428.0062943
	LI S J , GENG L L , WANG P . Vehicle target detection based on improved Yolov4. Computer Engineering, 2023, 49 (4): 272- 280. doi: 10.19678/j.issn.1000-3428.0062943
3	庄建军, 徐子恒, 张若愚. 基于改进的YOLOv5模型和射线法的车辆违停检测. 南京信息工程大学学报(自然科学版), 2024, 16 (3): 341- 351. doi: 10.13878/j.cnki.jnuist.20230402001
	ZHUANG J J , XU Z H , ZHANG R Y . Illegal parking detection based on improved YOLOv5 model and ray method. Journal of Nanjing University of Information Science & Technology(Natural Science Edition), 2024, 16 (3): 341- 351. doi: 10.13878/j.cnki.jnuist.20230402001
4	赵逸如, 刘正熙, 熊运余, 等. 基于目标检测和语义分割的人行道违规停车检测. 现代计算机, 2020, (9): 82- 88. URL
	ZHAO Y R , LIU Z X , XIONG Y Y , et al. Detection of illegal sidewalk parking based on object detection and semantic segmentation. Modern Computer, 2020, (9): 82- 88. URL
5	KE X , ZHANG Y F . Fine-grained vehicle type detection and recognition based on dense attention network. Neurocomputing, 2020, 399, 247- 257. doi: 10.1016/j.neucom.2020.02.101
6	TARIQ A, KHAN M Z, GHANI KHAN M U. Real time vehicle detection and color recognition using tuned features of faster-RCNN[C]//Proceedings of the 1st International Conference on Artificial Intelligence and Data Analytics. Washington D. C., USA: IEEE Press, 2021: 262-267.
7	BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOv4: optimal speed and accuracy of object detection[EB/OL]. [2023-12-10]. https://arxiv.org/abs/2004.10934v1.
8	REDMON J, FARHADI A. YOLOv3: an incremental improvement[EB/OL]. [2023-12-10]. https://arxiv.org/abs/1804.02767.
9	CHEN L C , PAPANDREOU G , KOKKINOS I , et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40 (4): 834- 848. doi: 10.1109/TPAMI.2017.2699184
10	詹泽辉, 钟铭恩, 袁彬淦, 等. 随机平视摄像条件下的路边车辆违停检测[J/OL]. 上海交通大学学报: 1-10[2023-12-10]. https://doi:10.16183/j.cnki.jsjtu.2023.578.
	ZHAN Z H, ZHONG M E, YUAN B A, et al. Detection of roadside vehicle parking violations under random horizontal camera condition[J/OL]. Journal of Shanghai Jiao Tong University: 1-10[2023-12-10]. https://doi:10.16183/j.cnki.jsjtu.2023.578. (in Chinese)
11	LIU Z, MAO H Z, WU C Y, et al. A ConvNet for the 2020s[EB/OL]. [2023-12-10]. https://arxiv.org/abs/2201.03545.
12	李建威, 吕晓琪, 谷宇. 基于改进ConvNeXt的皮肤镜图像分类方法. 计算机工程, 2023, 49 (10): 239-246, 254.
	LI J W , (LÜ/LV/LU/LYU) X Q , GU Y . Dermoscopy image classification method based on improved ConvNeXt. Computer Engineering, 2023, 49 (10): 239-246, 254.
13	TANG K H, HUANG J Q, ZHANG H W. Long-tailed classification by keeping the good and removing the bad momentum causal effect[C]//Proceedings of NIPS'20. Cambridge, USA: MIT Press, 2020: 267-277.
14	JEONG Y , PARK K H , PARK D . Homogeneity patch search method for voting-based efficient vehicle color classification using front-of-vehicle image. Multimedia Tools and Applications, 2019, 78 (20): 28633- 28648. URL
15	CHEN P , BAI X , LIU W Y . Vehicle color recognition on urban road by feature context. IEEE Transactions on Intelligent Transportation Systems, 2014, 15 (5): 2340- 2346. URL
16	HU M D, BAI L, LI Y, et al. Vehicle 24-color long tail recognition based on smooth modulation neural network with multi-layer feature representation[EB/OL]. [2023-12-10]. https://arxiv.org/abs/2107.09944.
17	LIU X C , LIU W , MEI T , et al. A deep learning-based approach to progressive vehicle re-identification for urban surveillance. Berlin, Germany: Springer, 2016.
18	IOFFE S. Batch renormalization: towards reducing minibatch dependence in batch-normalized models[EB/OL]. [2023-12-10]. https://arxiv.org/abs/1702.03275.
19	LIU Z, LIN Y T, CAO Y, et al. Swin transformer: hierarchical vision transformer using shifted windows[C]//Proceedings of IEEE/CVF International Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2022: 9992-10002.
20	DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[EB/OL]. [2023-12-10]. https://arxiv.org/abs/1810.04805.
21	BA J L, KIROS J R, HINTON G E. Layer normalization[EB/OL]. [2023-12-10]. https://arxiv.org/abs/1607.06450.
22	HE K M, ZHANG X, REN S Q, et al. Deep residual learning for image recognition[EB/OL]. [2023-12-10]. https://arxiv.org/pdf/1512.03385.pdf.
23	HENDRYCKS D, GIMPEL K. Gaussian error linear units[EB/OL]. [2023-12-10]. https://arxiv.org/abs/1606.08415.
24	RADFORD A , WU J , CHILD R , et al. Language models are unsupervised multitask learners. OpenAI Blog, 2019, 1 (8): 9.
25	HOWARD A G, ZHU M L, Chen B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications[EB/OL]. [2023-12-10]. https://arxiv.org/abs/1704.04861.
26	CHOLLET F. Xception: deep learning with depthwise separable convolutions[C]//Proceedings of 2017 IEEE Conference on Computer Vision. Washington D. C., USA: IEEE Press, 2017: 332-341.
27	NAIR V, HINTON G E. Rectified linear units improve restricted boltzmann machines[C]//Proceedings of the 27th International Conference on International Conference on Machine Learning. Washington D. C., USA: IEEE Press, 2010: 807-814.
28	VASWANI A, SHAZEER N M, PARMAR N, et al. Attention is all you need[EB/OL].[2023-12-10]. https://arxiv.org/abs/1706.03762.

[1]	林芷薇, 杨祖元, 王斯秋, 杨超. 基于多尺度线性全局注意力的运动员检测算法[J]. 计算机工程, 2024, 50(7): 352-359.
[2]	陈伟, 王晓龙, 张晏玮, 安国成, 江波. 基于改进YOLOv8的高速公路服务区车辆违停检测[J]. 计算机工程, 2024, 50(4): 11-19.
[3]	陈伟, 王晓龙, 张晏玮, 安国成, 江波. 基于改进YOLOv8的高速公路服务区车辆违停检测[J]. 计算机工程, 2024, 50(4): 11-19.
[4]	王林, 赵莉, 王无为. 高动态场景下无人机空对空目标检测[J]. 计算机工程, 2024, 50(12): 265-275.
[5]	刘俊豪, 王美林, 谢兴, 宋烨兴, 许莉花. 基于改进YOLOv5的皮革瑕疵检测算法[J]. 计算机工程, 2023, 49(8): 240-249.
[6]	陈露萌, 曹彦彦, 黄民, 谢鑫钢. 基于改进YOLOv5的火焰检测方法[J]. 计算机工程, 2023, 49(8): 291-301, 309.
[7]	李松江, 耿兰兰, 王鹏. 基于改进Yolov4的车辆目标检测[J]. 计算机工程, 2023, 49(4): 272-280.
[8]	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.
[9]	李建威, 吕晓琪, 谷宇. 基于改进ConvNeXt的皮肤镜图像分类方法[J]. 计算机工程, 2023, 49(10): 239-246, 254.
[10]	毛雨晴, 赵奎. 基于改进YOLOv5的多任务安全人头检测算法[J]. 计算机工程, 2022, 48(8): 136-143.
[11]	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.
[12]	江枭宇, 李忠兵, 张军豪, 彭娇, 文婷. 基于NCS2神经计算棒的车辆检测方法[J]. 计算机工程, 2021, 47(3): 298-303.
[13]	杨先凤,吴姝泓. 基于自适应阈值与双特征的ViBe运动车辆检测算法[J]. 计算机工程, 2018, 44(10): 241-245,251.
[14]	吴骏,李文杰,耿磊,肖志涛,张芳,李月龙. 基于单目视觉的前方车辆检测与测距[J]. 计算机工程, 2017, 43(2): 26-32.
[15]	杨先凤,杨燕. 一种基于HOG-LBP 的高效车辆检测方法[J]. 计算机工程, 2014, 40(9): 210-214.

选择文件类型/文献管理软件名称

选择包含的内容