路侧违停车辆移动影像检测方法

doi:10.11947/j.AGCS.2025.20250038

测绘学报 ›› 2025, Vol. 54 ›› Issue (9): 1712-1726.doi: 10.11947/j.AGCS.2025.20250038

• 地图学与地理信息 • 上一篇

路侧违停车辆移动影像检测方法

唐康¹^,²^,³(), 孙玉¹^,²^,³(), 仲逍阳¹^,²^,³, 高嘉良¹^,²^,³, 陈崇成¹^,²^,³

^1.福州大学空间数据挖掘与信息共享教育部重点实验室，福建　福州　350108
^2.福州大学地理空间信息技术国家地方联合工程研究中心，福建　福州　350108
^3.数字中国研究院（福建），福建　福州　350108

收稿日期:2025-01-23 修回日期:2025-09-01 出版日期:2025-10-10 发布日期:2025-10-10
通讯作者: 孙玉 E-mail:1635138766@qq.com;jade.yusun@outlook.com
作者简介:唐康（2000—），男，硕士生，研究方向为数字地球与智慧社会。E-mail：1635138766@qq.com
基金资助:
国家自然科学基金(42171426);福建省经济联盟科技创新团队项目;地理信息科学与技术全国重点实验室开放基金(2024KFJJ025)

Moving video-based detection for roadside illegally parking vehicles

Kang TANG¹^,²^,³(), Yu SUN¹^,²^,³(), Xiaoyang ZHONG¹^,²^,³, Jialiang GAO¹^,²^,³, Chongcheng CHEN¹^,²^,³

^1.Key Lab of Spatial Data Mining and Information Sharing of Ministry of Education, Fuzhou University, Fuzhou 350108, China
^2.National Engineering Research Center of Geospatial Information Technology, Fuzhou University, Fuzhou 350108, China
^3.Academy of Digital China (Fujian), Fuzhou 350108, China

Received:2025-01-23 Revised:2025-09-01 Online:2025-10-10 Published:2025-10-10
Contact: Yu SUN E-mail:1635138766@qq.com;jade.yusun@outlook.com
About author:TANG Kang (2000—), male, postgraduate, majors in digital Earth and smart society. E-mail: 1635138766@qq.com
Supported by:
The National Natural Science Foundation of China(42171426);Fujian Cultural and Tourism Economy Alliance Science and Technology Innovation Team-Digital Culture and Smart Tourism Fujian Provincial Science and Technology Innovation Team Project;Collaborative Knowledge Graph and Large Language Models for Maritime Shipping Risk Perception and Reasoning-Based Decision-Making(2024KFJJ025)

摘要/Abstract

摘要：

路侧停车区在缓解城市停车压力方面发挥了重要作用，但随着城市机动车保有量的持续增长，城市路侧停车位供需缺口持续扩大，由此导致的违停现象严重影响了交通效率和安全。现有基于定点摄像头或传感器的违停监控系统存在成本高、监控范围受限等问题。为此，本文提出了一种基于移动摄像头的路侧停车区疑似违停车辆检测方法。该方法利用嵌入式设备结合改进的目标检测算法（mAP@50提升3.3%）进行开发，依托自制数据集完成模型训练，实现了疑似违停车辆的实时检测，并有效支持大范围区域的疑似违停监控任务。通过与同时段、同路段的无人机跟踪正射影像对比分析，本文方法的违停检测平均查准率达0.87、查全率达0.88，检测速度高达53.96 fps，充分验证了本文方法的可行性和有效性。

关键词: 嵌入式设备, 目标检测, 疑似违停车辆

Abstract:

Roadside parking zones play a significant role in alleviating urban parking pressure. However, with the continuous growth in urban motor vehicle ownership, the supply-demand gap for roadside parking spaces continues to widen, leading to severe illegally parking that significantly impacts traffic efficiency and safety. Existing illegally parking monitoring systems based on fixed-point cameras or sensors suffer from high costs and limited coverage. To address this, this paper proposes a solution for detecting suspected illegally parked vehicles in roadside parking zones using mobile cameras. The solution is developed using embedded devices combined with an improved object detection algorithm (achieving a 3.3% increase in mAP@50). Trained on a custom-built dataset, it enables real-time detection of suspected illegally parked vehicles and effectively supports large-area monitoring tasks for suspected illegal parking. Comparative analysis with simultaneous drone-tracked orthophoto imagery from the same road sections showed that the solution achieves an average precision of 0.87, a recall of 0.88, and a detection speed of 53.96 frames per second (fps), fully validating the feasibility and effectiveness of the proposed method.

Key words: embedded devices, obeject detection, suspected illegally parked vehicles

中图分类号:

P237

唐康, 孙玉, 仲逍阳, 高嘉良, 陈崇成. 路侧违停车辆移动影像检测方法[J]. 测绘学报, 2025, 54(9): 1712-1726.

Kang TANG, Yu SUN, Xiaoyang ZHONG, Jialiang GAO, Chongcheng CHEN. Moving video-based detection for roadside illegally parking vehicles[J]. Acta Geodaetica et Cartographica Sinica, 2025, 54(9): 1712-1726.

图/表 20

图1

图2

图3

图4

图5

表1

图6

表2

表3

图7

图8

表4

表5

表6

表7

图9

表8

图10

图11

表9

参考文献 25

[1]	HUANG Y H, HSIEH C H. A decision support system for available parking slots on the roadsides in urban areas[J]. Expert Systems with Applications, 2022, 205: 117668.
[2]	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[J]. 南京信息工程大学学报, 2024, 16(3): 341-351.
	ZHUANG Jianjun, XU Ziheng, ZHANG Ruoyu. Illegal parking detection based on improved YOLOv5 model and ray method[J]. Journal of Nanjing University of Information Science & Technology, 2024, 16(3): 341-351.
[3]	XIE Xuemei, WANG Chenye, CHEN Shu, et al. Real-time illegal parking detection system based on deep learning[C]//Proceedings of 2017 International Conference on Deep Learning Technologies. Chengdu: ACM Press, 2017: 23-27.
[4]	SHARMA N, BARAL S, PAING M P, et al. Parking time violation tracking using YOLOv8 and tracking algorithms[J]. Sensors, 2023, 23(13): 5843.
[5]	陈伟, 王晓龙, 张晏玮, 等. 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.
	CHEN Wei, WANG Xiaolong, ZHANG Yanwei, et al. Vehicle violation detection based on improved YOLOv8 in highway service areas[J]. Computer Engineering, 2024, 50(4): 11-19.
[6]	RANJAN A, MISRA P, VASAN A, et al. City scale monitoring of on-street parking violations with Street HAWK[C]//Proceedings of the 6th ACM International Conference on Systems for Energy-Efficient Buildings, Cities, and Transportation. New York: ACM Press, 2019: 31-40.
[7]	PENG Xinggan, SONG Rongzihan, CAO Qi, et al. Real-time illegal parking detection algorithm in urban environments[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(11): 20572-20587.
[8]	程梁华, 黄瑞雪, 沈鑫. 适于高动态视频场景下的城市道路违停检测算法[J]. 计算机科学, 2024, 51(12): 190-198.
	CHENG Lianghua, HUANG Ruixue, SHENG Xin. Urban illegal on-road parking detection algorithm for high dynamic video scenarios[J]. Computer Science, 2024, 51(12): 190-198.
[9]	田爱军, 蔡旭阳, 陈玮, 等. 无人机端路面车辆违停检测及取证系统[J]. 测控技术, 2021, 40(5): 67-74.
	TIAN Aijun, CAI Xuyang, CHEN Wei, et al. Vehicle illegal parking detection and evidence collection system on UAV[J]. Measurement & Control Technology, 2021, 40(5): 67-74.
[10]	Ultralytics . YOLOv5[EB/OL]. [2024-11-01]. http://github.com/ultralytics/yolov5.
[11]	WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of 2018 Computer Vision-ECCV. Cham: Springer, 2018: 3-19.
[12]	MA Siliang, XU Yong. Mpdiou: a loss for efficient and accurate bounding box regression[EB/OL]. [2024-11-01]. https://arxiv.org/abs/2307.07662.
[13]	ZHENG Z, WANG P, LIU W, et al. Distance-IoU loss: faster and better learning for bounding box regression[C]//Proceedings of 2020 AAAI conference on artificial intelligence. [S. l.]: PKP, 2020, 34(7): 12993-13000.
[14]	LIU Shu, QI Lu, QIN Haifang, et al. Path aggregation network for instance segmentation[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake city: IEEE, 2018: 8759-8768.
[15]	WOJKE N, BEWLEY A, PAULUS D. Simple online and realtime tracking with a deep association metric[C]//Proceedings of 2017 IEEE International Conference on Image Processing. Beijing: IEEE, 2017: 3645-3649.
[16]	BEWLEY A, GE Zongyuan, OTT L, et al. Simple online and realtime tracking[C]//Proceedings of 2016 IEEE International Conference on Image Processing. Phoenix: IEEE, 2016: 3464-3468.
[17]	LIU Fei, ZENG Zhiyuan, JIANG Rong. A video-based real-time adaptive vehicle-counting system for urban roads[J]. PLoS One, 2017, 12(11): e0186098.
[18]	GEIGER A, LENZ P, STILLER C, et al. Vision meets robotics: the KITTI dataset[J]. The International Journal of Robotics Research, 2013, 32(11): 1231-1237.
[19]	TONG Zanjia, CHEN Yuhang, XU Zzewei, et al. Wise-IoU: bounding box regression loss with dynamic focusing mechanism[EB/OL]. [2024-11-01]. https://arxiv.org/abs/2301.10051.
[20]	ZHANG Hao, XU Cong, ZHANG Shuaijie. Inner-IoU: more effective intersection over union loss with auxiliary bounding box[EB/OL]. [2024-11-01]. https://arxiv.org/abs/2311.02877.
[21]	HOU Qibin, ZHOU Daquan, FENG Jiashi. Coordinate attention for efficient mobile network design[C]//Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville: IEEE, 2021: 13708-13717.
[22]	OUYANG Daliang, HE Su, ZHANG Guozhong, et al. Efficient multi-scale attention module with cross-spatial learning[C]//Proceedings of 2023 IEEE International Conference on Acoustics, Speech and Signal Processing. Rhodes Island: IEEE, 2023: 1-5.
[23]	WAN Dahang, LU Rongsheng, SHEN Siyuan, et al. Mixed local channel attention for object detection[J]. Engineering Applications of Artificial Intelligence, 2023, 123: 106442.
[24]	REDMON Joseph, FARHADI Ali. Yolov3: an incremental improvement[EB/OL]. [2024-11-01]. https://arxiv.org/abs/1804.02767.
[25]	WANG Ao, CHEN Hui, LIU Lihao, et al. Yolov10: real-time end-to-end object detection[EB/OL]. [2024-11-01]. https://arxiv.org/abs/2405.14458.

虚拟线位置	总车辆数	触发数	触发比
0.6×width	61	58	0.95
0.7×width	61	60	0.98
0.8×width	61	34	0.56

Tire_1	Tire_2	Tire_3	Tire_4	疑似违停	合理停泊
前	后				√
前		后		√
	前&后			√
前			后	√
	前	后		√
	后	前		√
后	前			√
后			前	√
		后	前	√
			前&后	√

配置	Jetson Orin NX	PC
OS	Ubuntu 20.04 LTS	Windows 10
CPU	6-core NVIDIA Arm Cortex A78AE v8.2 64 bit CPU 2MB L2+4MB L3	Intel（R）Core（TM）i7-11700
GPU	1024-core NVIDIA Ampere GPU with 32 Tensor Cores	GTX1660Super
Memory	16 GB 128 bit LPDDR5 102.4 GB/S	32GB（2933 MHz/3200 MHz）

标签类别	标签数量	占比/（%）
Vehicle	1980	32
Tire_1	1267	20
Tire_2	1017	16
Tire_3	710	11
Tire_4	1280	20

边框损失函数	查准率	召回率	mAP@50
CIoU	0.911	0.908	0.941
MPDIoU	0.923	0.922	0.951
Wise-IoU	0.928	0.898	0.942
Inner-IoU	0.902	0.904	0.939

路侧违停车辆移动影像检测方法

Moving video-based detection for roadside illegally parking vehicles

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 20

参考文献 25

相关文章 14

编辑推荐

Metrics

本文评价

注意力机制	查准率	召回率	mAP@50
无	0.923	0.922	0.951
CA	0.925	0.935	0.965
EMA	0.935	0.934	0.968
MLCA	0.936	0.931	0.963
CBAM	0.940	0.939	0.972

模型	查准率	召回率	mAP@50	模型大小/MB	推理用时/ms
YOLOv5	0.911	0.908	0.941	13.69	20.9
YOLOv8	0.912	0.930	0.951	21.46	24.5
YOLOv9	0.890	0.902	0.949	14.52	52.2
YOLOv10	0.815	0.864	0.898	15.75	33
YOLOv11	0.902	0.878	0.938	18.28	28.4
YOLOv3-tiny	0.902	0.863	0.927	23.23	18
本文模型	0.940	0.939	0.972	14.05	22

CBAM	MPDIoU	FLOPs	mAP@50	Time/ms
		15.8×10⁹	0.941	20.9
√		16.4×10⁹	0.962	22
	√	15.8×10⁹	0.951	20.9
√	√	16.4×10⁹	0.972	22

ID	查准率	召回率	检测速度/fps	图像分辨率/像素
A	1.00	0.85	55.28	940×544
B	0.9	0.9	53.17
C	0.94	0.89	52.27
D	0.62	0.89	55.13
平均	0.87	0.88	53.96

[1]	谢亚坤, 赵耀纪, 涂佳星, 夏瑞丰, 冯德俊, 刘苏凝, 陈虹宇, 朱军. 融合边缘与全局特征的遥感影像显著性目标检测方法[J]. 测绘学报, 2025, 54(7): 1265-1279.
[2]	杨军, 解恒静, 范红超, 闫浩文. 遥感影像目标检测多尺度熵神经网络架构搜索[J]. 测绘学报, 2024, 53(7): 1384-1400.
[3]	黄启灏, 靳国旺, 熊新, 王丽美, 李佳豪. 通道剪枝与知识蒸馏相结合的轻量化SAR目标检测[J]. 测绘学报, 2024, 53(4): 712-723.
[4]	刘家兴, 黄玉春, 石文轩, 叶曦, 杨鹤. 顾及拓扑结构的道路标线提取[J]. 测绘学报, 2024, 53(11): 2213-2227.
[5]	陈玮彤, 许鑫, 朱长青, 任娜. 基于后门水印和感兴趣区加密的遥感目标检测数据集版权保护算法[J]. 测绘学报, 2024, 53(11): 2086-2098.
[6]	李景文, 韦晶闪, 姜建武, 陆妍玲, 刘垒, 唐一飞, 李旭. 多视角监控视频中动态目标的时空信息提取方法[J]. 测绘学报, 2022, 51(3): 388-400.
[7]	陈占龙, 李双江, 徐永洋, 徐道柱, 马超, 赵军利. 高分影像密集建筑物Correg-YOLOv3检测方法[J]. 测绘学报, 2022, 51(12): 2531-2540.
[8]	黄远程, 薛园园, 李朋飞. 高光谱影像子空间分析孤立森林异常目标探测方法[J]. 测绘学报, 2021, 50(3): 416-425.
[9]	郑鑫, 潘斌, 张健. 可变形网络与迁移学习相结合的电力塔遥感影像目标检测法[J]. 测绘学报, 2020, 49(8): 1042-1050.
[10]	董志鹏, 王密, 李德仁, 王艳丽, 张致齐. 遥感影像目标的尺度特征卷积神经网络识别法[J]. 测绘学报, 2019, 48(10): 1285-1295.
[11]	陈丁, 万刚, 李科. 多层特征与上下文信息相结合的光学遥感影像目标检测[J]. 测绘学报, 2019, 48(10): 1275-1284.
[12]	邓志鹏, 孙浩, 雷琳, 周石琳, 邹焕新. 基于多尺度形变特征卷积网络的高分辨率遥感影像目标检测[J]. 测绘学报, 2018, 47(9): 1216-1227.
[13]	戴玉超, 张静, Fatih PORIKLI, 何明一. 深度残差网络的多光谱遥感图像显著目标检测[J]. 测绘学报, 2018, 47(6): 873-881.
[14]	胡风明; 杨汝良; 范学花; 商建; . 利用方向性粗糙度特征对SAR图像目标检测的研究[J]. 测绘学报, 2009, 38(3): 0-201.