Vehicle tracking algorithms using YOLOv8-OD and DeepSORT
TONG Yuan
FEI Shumin
Abstract：To address the limitations of traditional multi-object tracking algorithms in terms of detection accuracy,tracking precision,and robustness,this paper proposes a novel method based on the Tracking-By-Detection paradigm for vehicle flow monitoring.The method employs the YOLOv8 object detection al-gorithm to achieve rapid localization and identification of vehicle targets,and integrates an improved deep learning-based DeepSORT multi-object tracking algorithm to ensure accurate and real-time tracking and counting of vehicles.Experimental results demonstrate that the proposed method achieves high detection accuracy when handling fast-moving vehicles,with an average precision of 94.7%.This end-to-end ap-proach exhibits good feasibility and effectiveness in batch processing applications of vehicle video data.
Keywords：YOLOv8DeepSORTdeep learningvehicle tracking
Publication Date：2026-02-25
Online Publishing Date：2026-08-28(First online date of this platform, not the publication date of the document)
Pages：8( 24-31 )