Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT Algorithm

To investigate whether the skid resistance of the ramp meets the requirements of vehicle driving safety and stability, the simulation using the ideal driver model is inaccurate. Therefore, considering the driver’s driving habits, this paper proposes the use of Unmanned aerial vehicles (UAVs) for the...

Full description

Saved in:
Bibliographic Details
Main Authors: Binshuang Zheng, Jing Zhou, Zhengqiang Hong, Junyao Tang, Xiaoming Huang
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/9/2788
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850275330515795968
author Binshuang Zheng
Jing Zhou
Zhengqiang Hong
Junyao Tang
Xiaoming Huang
author_facet Binshuang Zheng
Jing Zhou
Zhengqiang Hong
Junyao Tang
Xiaoming Huang
author_sort Binshuang Zheng
collection DOAJ
description To investigate whether the skid resistance of the ramp meets the requirements of vehicle driving safety and stability, the simulation using the ideal driver model is inaccurate. Therefore, considering the driver’s driving habits, this paper proposes the use of Unmanned aerial vehicles (UAVs) for the collection and extraction of vehicle driving information. To process the collected UAV video, the Google Collaboration platform is used to modify and compile the “You Only Look Once” version 5 (YOLOv5) algorithm with Python 3.7.12, and YOLOv5 is retrained with the captured video. The results show that the precision rate P and recall rate R have satisfactory results with an F1 value of 0.86, reflecting a good P-R relationship. The loss function also stabilized at a very low level after 70 training epochs. Then, the trained YOLOv5 is used to replace the Faster R-CNN detector in the DeepSORT algorithm to improve the detection accuracy and speed and extract the vehicle driving information from the perspective of UAV. By coding, the coordinate information of the vehicle trajectory is extracted, the trajectory is smoothed, and the frame difference method is used to calculate the real-time speed information, which is convenient for the establishment of a real driver model.
format Article
id doaj-art-794859ea07294c9ab9c1a192ecc9090c
institution OA Journals
issn 1424-8220
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-794859ea07294c9ab9c1a192ecc9090c2025-08-20T01:50:46ZengMDPI AGSensors1424-82202025-04-01259278810.3390/s25092788Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT AlgorithmBinshuang Zheng0Jing Zhou1Zhengqiang Hong2Junyao Tang3Xiaoming Huang4Research and Development Center of Transport Industry of New Generation of Artificial Intelligence Technology, Zhejiang Scientific Research Institute of Transport, No. 705 Dalongjuwu Rd., Hangzhou 311305, ChinaSchool of Modern Posts, Nanjing University of Posts and Telecommunications, Nanjing 210023, ChinaChina State Construction Engineering (Hong Kong) Limited, Hong Kong 999077, ChinaSchool of Transportation, Southeast University, Nanjing 211189, ChinaSchool of Transportation, Southeast University, Nanjing 211189, ChinaTo investigate whether the skid resistance of the ramp meets the requirements of vehicle driving safety and stability, the simulation using the ideal driver model is inaccurate. Therefore, considering the driver’s driving habits, this paper proposes the use of Unmanned aerial vehicles (UAVs) for the collection and extraction of vehicle driving information. To process the collected UAV video, the Google Collaboration platform is used to modify and compile the “You Only Look Once” version 5 (YOLOv5) algorithm with Python 3.7.12, and YOLOv5 is retrained with the captured video. The results show that the precision rate P and recall rate R have satisfactory results with an F1 value of 0.86, reflecting a good P-R relationship. The loss function also stabilized at a very low level after 70 training epochs. Then, the trained YOLOv5 is used to replace the Faster R-CNN detector in the DeepSORT algorithm to improve the detection accuracy and speed and extract the vehicle driving information from the perspective of UAV. By coding, the coordinate information of the vehicle trajectory is extracted, the trajectory is smoothed, and the frame difference method is used to calculate the real-time speed information, which is convenient for the establishment of a real driver model.https://www.mdpi.com/1424-8220/25/9/2788rampUAV videovehicle recognitionYOLOv5 algorithmvehicle track information
spellingShingle Binshuang Zheng
Jing Zhou
Zhengqiang Hong
Junyao Tang
Xiaoming Huang
Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT Algorithm
Sensors
ramp
UAV video
vehicle recognition
YOLOv5 algorithm
vehicle track information
title Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT Algorithm
title_full Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT Algorithm
title_fullStr Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT Algorithm
title_full_unstemmed Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT Algorithm
title_short Vehicle Recognition and Driving Information Detection with UAV Video Based on Improved YOLOv5-DeepSORT Algorithm
title_sort vehicle recognition and driving information detection with uav video based on improved yolov5 deepsort algorithm
topic ramp
UAV video
vehicle recognition
YOLOv5 algorithm
vehicle track information
url https://www.mdpi.com/1424-8220/25/9/2788
work_keys_str_mv AT binshuangzheng vehiclerecognitionanddrivinginformationdetectionwithuavvideobasedonimprovedyolov5deepsortalgorithm
AT jingzhou vehiclerecognitionanddrivinginformationdetectionwithuavvideobasedonimprovedyolov5deepsortalgorithm
AT zhengqianghong vehiclerecognitionanddrivinginformationdetectionwithuavvideobasedonimprovedyolov5deepsortalgorithm
AT junyaotang vehiclerecognitionanddrivinginformationdetectionwithuavvideobasedonimprovedyolov5deepsortalgorithm
AT xiaominghuang vehiclerecognitionanddrivinginformationdetectionwithuavvideobasedonimprovedyolov5deepsortalgorithm