ST137: ADVOS - AUTOMATED DETECTION OF VEHICLES WITHOUT OKU STICKERS USING ARTIFICIAL INTELLIGENCE

MUHAMMAD IMAN KHAIRULNAZARULLAH BIN CHE MOHD HASSAN POLITEKNIK SULTAN MIZAN ZAINAL ABIDIN

In line with the growing public awareness of the rights of Persons with Disabilities (OKU), the need to regulate the use of designated parking spaces requires close attention and monitoring. However, manual enforcement methods remain inefficient as they rely on continuous supervision by authorities. This project aims to develop an automated system for detecting vehicles without OKU stickers using Artificial Intelligence (AI) based on computer vision. The method involves collecting and labeling images using the Roboflow platform and training an object detection model using YOLOv5. The trained model is then integrated with a Raspberry Pi and a camera to perform real-time detection. The system is also equipped with a flashing light as a visual warning indicator and sends automatic notifications to enforcement officers via the Telegram application when a vehicle without an OKU sticker is detected. The developed model demonstrates excellent accuracy, achieving a 99.5% accuracy rate in detecting OKU stickers on vehicles. Overall, the ADVOS system has the potential to assist local authorities in enhancing the effectiveness of automated and intelligent enforcement of OKU parking space usage.