基于改进YOLOv8n算法的城市空间混行交通参与者检测模型

doi:10.16265/j.cnki.issn1003-3033.2024.12.0465

中国安全科学学报 ›› 2024, Vol. 34 ›› Issue (12): 178-186.doi: 10.16265/j.cnki.issn1003-3033.2024.12.0465

基于改进YOLOv8n算法的城市空间混行交通参与者检测模型

周军超¹^,²(), 陈鑫¹, 高建杰³^,^**(), 章杰³^,⁴

¹ 四川轻化工大学机械工程学院,四川自贡 643000
² 成渝地区双城经济圈(泸州)先进技术研究院,四川泸州 646000
³ 四川警察学院智能警务四川省重点实验室,四川泸州 646000
⁴ 长沙理工大学汽车与机械工程学院,湖南长沙 410114

收稿日期:2024-08-17 修回日期:2024-10-16 出版日期:2024-12-28
通信作者:
**高建杰(1985—),男,山东栖霞人,博士,副教授,主要从事城市交通安全管理、城市交通风险判别与检测等方面的研究。E-mail: jianjiecq@163.com。

作者简介:

周军超 (1987—),男,湖北襄阳人,博士,副教授,主要从事城市智能交通安全检测、智能网联汽车安全驾驶与风险识别、新型城市轨道车辆安全与控制等方面的研究。E-mail:zhou1987g@163.com。

章杰,副教授。

基金资助:
智能警务四川省重点实验室开放课题资助(ZNJW2023KFQN005); 智能警务四川省重点实验室开放课题资助(ZNJW2022KFQN004); 泸州市科技计划资助(2023JYJ066)

Urban spatially mixed traffic participants detection model based on improved YOLOv8n

ZHOU Junchao¹^,²(), CHEN Xin¹, GAO Jianjie³^,^**(), ZHANG Jie³^,⁴

¹ School of Mechanical Engineering, Sichuan University of Science & Engineering, Zigong Sichuan 643000, China
² Chengdu-Chongqing Economic Circle (Luzhou) Advanced Technology Research Institute, Luzhou Sichuan 646000, China
³ Intelligent Policing Key Laboratory of Sichuan Province, Sichuan Police College, Luzhou Sichuan 646000, China
⁴ College of Automotive and Mechanical Engineering, Changsha University of Science & Technology, Changsha Hunan 410114, China

Received:2024-08-17 Revised:2024-10-16 Published:2024-12-28

摘要/Abstract

摘要：

为提升智能网联汽车和交通监测系统对交通参与者的识别精度和检测速度,及时应对城市空间混行交通环境中的安全隐患,提出一种基于改进YOLOv8n算法的城市空间混行交通参与者检测模型。首先,在数据输入阶段利用几何变换和像素变换策略来防止过度拟合,提高鲁棒性和泛化性;其次,使用空间到深度的非跨行卷积(SPD-Conv)模块代替所有原始卷积层,提高对低分辨率小目标的特征提取能力;同时,在颈部网络融合结构中加入轻量级坐标注意力机制(CA)模块,在几乎不添加任何计算开销的同时提高对关键信息的识别精度;然后,用边界框损失函数有效交并比(EIoU)替代原损失函数,使模型获得更卓越的收敛速度与识别稳定性;最后,利用平台公开和自建融合的交通参与者数据集(BNS)进行消融和对比试验,运用自动驾驶试验平台进行实景实时检测。结果表明:相比于基线模型YOLOv8n,改进模型SEC-YOLO的每秒传输帧数(FPS)和平均精度均值(mAP)分别提高了7.3%和3.2%;与主流模型对比,mAP和FPS性能值最佳;在自动驾驶试验平台上的实景检测平均准确率为95%。SEC-YOLO算法模型对城市交通参与者的检测准确率更高,鲁棒性和实时性更强。

关键词: YOLOv8n, 空间混行, 交通参与者, 检测模型, 空间到深度的非跨行卷积(SPD-Conv), 坐标注意力机制(CA)

Abstract:

In order to improve the recognition accuracy and detection speed of traffic participants by intelligent networked vehicles and traffic monitoring systems so that they can timely respond to the safety hazards in the mixed traffic environment in urban space, a mixed traffic participant detection model in urban space based on the improved YOLOv8n algorithm was proposed. Firstly, geometric transformation and pixel transformation enhancement strategies were employed in the data input stage to prevent overfitting and improve robustness, and generalization. Secondly, the SPD-Conv module was used to replace all original convolution layers of the YOLOv8n algorithm, which enhances the feature extraction capability for low-resolution small targets. Meanwhile, the CA module was added to the fusion structure of the neck network of the YOLOv8n algorithm to improve the recognition accuracy of key information with almost no additional computational overhead. Then, the boundary box loss function EIoU was used to replace the original loss function, enabling the model to achieve superior convergence speed and recognition stability. Finally, the ablation and comparison experiments were carried out with the public and self-built integrated traffic participant dataset, and the real-time detection experiment was carried out with the automatic driving experiment platform. The experimental results show that compared to the YOLOv8n model, the improved SEC-YOLO model has increased mAP and FPS by 3.2% and 7.9% respectively. The SEC-YOLO model outperforms mainstream models in terms of mAP and FPS as well. The average accuracy of real-scene detection on the automatic driving experimental platform is around 95%. The SEC-YOLO algorithm model achieves higher detection accuracy for urban traffic participants, with stronger robustness and real-time performance.

Key words: YOLOv8n, spatial mixing, traffic participants, detection model, space-to-depth convolution (SPD-Conv), coordinate attention (CA)

中图分类号:

X951交通运输安全

周军超, 陈鑫, 高建杰, 章杰. 基于改进YOLOv8n算法的城市空间混行交通参与者检测模型[J]. 中国安全科学学报, 2024, 34(12): 178-186.

ZHOU Junchao, CHEN Xin, GAO Jianjie, ZHANG Jie. Urban spatially mixed traffic participants detection model based on improved YOLOv8n[J]. China Safety Science Journal, 2024, 34(12): 178-186.

图/表 14

图1

图2

图3

图4

图5

图6

表1

图7

表2

表3

图8

图9

图10

图11

参考文献 17

[1]	刘晓芳, 吝涛, 于琪, 等. 城镇街巷空间人群通行疏散及碰撞风险定量模拟[J]. 中国安全科学学报, 2023, 33(11):133-141. doi: 10.16265/j.cnki.issn1003-3033.2023.11.2158
	LIU Xiaofang, LIN Tao, YU Qi, et al. Quantitative simulation of pedestrian evacuation and collision risk in urban street space[J]. China Safety Science Journal, 2023, 33(11):133-141. doi: 10.16265/j.cnki.issn1003-3033.2023.11.2158
[2]	张军, 尹柳, 巩欣飞, 等. 基于改进YOLOv5的小目标烟雾检测算法[J]. 中国安全科学学报, 2023, 33(11):75-81. doi: 10.16265/j.cnki.issn1003-3033.2023.11.0854
	ZHANG Jun, YIN Liu, GONG Xinfei, et al. Small target smoke detection algorithm based on improved YOLOv5[J]. China Safety Science Journal, 2023, 33(11):75-81. doi: 10.16265/j.cnki.issn1003-3033.2023.11.0854
[3]	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]. Proceedings of the IEEE Conference on Computer Cision and Cattern Recognition, 2016: 779-788.
[4]	LIN Tianyi, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]. Proceedings of the IEEE International Conference on Computer Vision, 2017: 2980-2988.
[5]	LIU Wu, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]. Proceedings of the 2016 European Conference on Computer Vision, 2016: 21-37.
[6]	夏正洪, 魏汝祥, 李彦冬. 改进 YOLOv3 算法下通航机场场面运动目标检测[J]. 中国安全科学学报, 2023, 33(2):82-88. doi: 10.16265/j.cnki.issn1003-3033.2023.02.0149
	XIA Zhenghong, WEI Ruxiang, LI Yandong. Moving target detection of general aviation airport based on improved YOLOv3 algorithm[J]. China Safety Science Journal, 2023, 33(2):82-88. doi: 10.16265/j.cnki.issn1003-3033.2023.02.0149
[7]	SHU Jijun, LING Qinghua, HAN Fei. An improved algorithm for small object detection based on YOLO v4 and multi-scale contextual information[J]. Computers and Electrical Engineering, 2023, 105:84-90.
[8]	WANG Shouhe, WANG Yongda, CHANG Yujian, et al. EBSE-YOLO: high precision recognition algorithm for small target foreign object detection[J]. IEEE Access, 2023, 11:57951-57 964.
[9]	ZHOU Wei, WANG Chuan, XIA Jie, et al. Monitoring-based traffic participant detection in urban mixed traffic: a novel dataset and a tailored detector[J]. IEEE Transactions on Intelligent Transportation Systems, 2024, 25(1):189-202.
[10]	王泽宇, 徐慧英, 朱信忠, 等. 基于YOLOv8改进的密集行人检测算法:MER-YOLO[J]. 计算机工程与科学, 2024, 46(6):1050-1062.
	WANG Zeyu, XU Huiying, ZHU Xinzhong, et al. An improved dense pedestrian detection algorithm based on YOLOv8: MER-YOLO[J]. Computer Engineering and Science, 2024, 46(6):1050-1062.
[11]	OUYANG Daliang, HE Su, ZHANG Guozhong, et al. Efficient multi-scale attention module with cross-spatial learning[C]. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2023: 1-5.
[12]	YI Hao, LIU Bo, ZHAO Bin, et al. Small object detection algorithm based on improved YOLOv8 for remote sensing[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17:1734-1747.
[13]	YU Fu, CHEN Hui, WANG Xia, et al. BDD100K: a diverse driving dataset for heterogeneous multitask learning[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2020: 2636-2645.
[14]	CATAK O, TAN Yue, ALI S. Prediction surface uncertainty quantification in object detection models for autonomous driving[C]. IEEE International Conference on Artificial Intelligence Testing, 2021: 93-100.
[15]	SUNKARA R, LUO Tao. No more strided convolutions or pooling: a new CNN building block for low-resolution images and small objects[C]. Joint European Conference on Machine Learning and Knowledge Discovery in Databases, 2022: 443-459.
[16]	HOU Qibin, ZHOU Daquan, FENG Jiashi. Coordinate attention for efficient mobile network design[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2021: 13 713-13 722.
[17]	YANG Zuomin, WANG Xialun, LI Jiang. EIoU: an improved vehicle detection algorithm based on vehiclenet neural network[J]. Journal of Physics Conference Series, 2021, 1924(1):DOI:10.1088/1742-6596/1924/1/012001.

软件		硬件
操作系统	Windows10	CPU	12代酷睿i5-12490F
模型构建	PyTorch	显卡	Nvidia GeForce RTX3060 Ti
CUDA版本	11.3	内存	32GB
编译语言	Python	—	—
开发平台	PyCharm	—	—

模型	SPD- Conv	CA	EIoU	P/%	R/%	mAP@ 0.5/%	mAP@0.5~ 0.95/%	FLOPs/10⁹
YOLOv8n	—	—	—	90.7	91.6	91.3	48.4	8.5
改进1	√	—	—	92.8	93.8	93.4	52.9	8.8
改进2	—	√	—	90.8	93.4	92.6	50.5	8.5
改进3	—	—	√	90.9	92.7	91.7	49.3	8.6
SEC-YOLO	√	√	√	93.1	95.9	94.5	54.2	8.9

模型	R/%	mAP@ 0.5/%	mAP@ 0.5~0.95/%	FPS
Faster R-CNN	85.2	84.3	42.1	14.7
YOLOv8n	91.6	91.3	48.4	87.4
SSD	83.1	84.4	37.7	44.8
YOLOv5n	89.8	88.5	45.3	63.5
YOLOv7	86.4	90.1	47.9	72.2
RetinaNet	79.8	83.2	39.5	24.9
SEC-YOLO	95.9	94.5	54.2	93.8

基于改进YOLOv8n算法的城市空间混行交通参与者检测模型

Urban spatially mixed traffic participants detection model based on improved YOLOv8n

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 17

相关文章 1

编辑推荐

Metrics

本文评价