@article{M28AFA32C,
title = "Implementation of a Deep Learning Image Analysis-Based Disabled Parking Space Management System in an Embedded System",
journal = "The Journal of Korean Institute of Communications and Information Sciences",
year = "2024",
issn = "1226-4717",
doi = "10.7840/kics.2024.49.7.1012",
author = "Chae-Yul Woo, Soon-Ryang Kwon",
keywords = "Embedded System, IoT Sensor, Deep Learning, Video Analysis, Management of Disabled Parking Spaces",
abstract = "Managing parking spaces for people with disabilities is one of the socially significant issues. To address this issue, there is a need to develop a system that can monitor and manage disabled parking spaces in real-time efficiently, utilizing IoT sensors, deep learning, and video analysis technologies.
This paper aims to implement a disabled parking space management system using deep learning-based video analysis in an embedded environment. To achieve this goal, we defined 15 types of South Korean vehicle license plates and constructed a corresponding dataset. Furthermore, to improve the real-time license plate recognition accuracy, we trained license plate detection and character recognition models based on YOLOv8, applying the most superior performing model to the embedded system. By enhancing Korean character recognition rates and implementing character combination algorithms, the system's stability and accuracy were improved. This approach enables efficient management of illegal parking for various license plates in real disabled parking spaces."
}