Vehicle Detection Based on Improved YOLOV5s in Complex Weather

Yusen Ma, Ye Tian, Ning Hao, Xinan Zhang, Yujun Shen

Research output: Chapter in Book/Conference paperConference paperpeer-review

Abstract

Vehicle detection has become a crucial task in various intelligent transportation systems. However, the current vehicle detection algorithm's accuracy is limited under complex weather conditions, such as illumination, haze, rain, and snow, which affect the image quality, making it difficult to detect the vehicle accurately. To overcome these limitations, we propose an improved YOLOV5s algorithm for vehicle detection under complex weather conditions. By embedding the Selective Kernel Attention Mechanism in the feature layer after the fusion of low-level features and high-level features, the ability of the neural network to obtain effective information is improved. Experimental results on the BDD100K dataset show that the proposed algorithm achieves a mean Average Precision of 58.2%, which is 1.3% higher than that of the original YOLOV5s algorithm. And the recall increased by 1.5%. The Precision decreases by only 0.7%. These results demonstrate the effectiveness and practicability of the improved YOLOV5s algorithm for vehicle target detection under complex weather conditions.

Original languageEnglish
Title of host publicationComputational and Experimental Simulations in Engineering - Proceedings of ICCES 2023—Volume 3
EditorsShaofan Li
PublisherSpringer Science and Business Media B.V.
Pages577-585
Number of pages9
ISBN (Print)9783031449468
DOIs
Publication statusPublished - 2024
Externally publishedYes
Event29th International Conference on Computational and Experimental Engineering and Sciences, ICCES 2023 - Shenzhen, China
Duration: 26 May 202329 May 2023

Publication series

NameMechanisms and Machine Science
Volume146
ISSN (Print)2211-0984
ISSN (Electronic)2211-0992

Conference

Conference29th International Conference on Computational and Experimental Engineering and Sciences, ICCES 2023
Country/TerritoryChina
CityShenzhen
Period26/05/2329/05/23

Fingerprint

Dive into the research topics of 'Vehicle Detection Based on Improved YOLOV5s in Complex Weather'. Together they form a unique fingerprint.

Cite this