Visible to the public Vehicle Violation Detection System Based on Improved YOLOv5 Algorithm

TitleVehicle Violation Detection System Based on Improved YOLOv5 Algorithm
Publication TypeConference Paper
Year of Publication2022
AuthorsSun, Xiaohan, Zhang, Yanju, Huang, Xiaobin, Wang, Fangzhou, Mo, Zugang
Conference Name2022 18th International Conference on Computational Intelligence and Security (CIS)
Date Publisheddec
Keywordscomposability, compositionality, Computational Intelligence, cryptography, license plate recognition, Object recognition, pubcrawl, Quadrilateral Detection, security, Violation Judgment, YOLOv5
AbstractThis paper proposes a vehicle violation determination system based on improved YOLOv5 algorithm, which performs vehicle violation determination on a single unit at a single intersection, and displays illegal photos and license plates of illegal vehicles on the webpage. Using the network structure of YOLOv5, modifying the vector output of the Head module, and modifying the rectangular frame detection of the target object to quadrilateral detection, the system can identify vehicles and lane lines with more flexibilities.
DOI10.1109/CIS58238.2022.00038
Citation Keysun_vehicle_2022