COMPARISON OF YOLO AND OCR PERFORMANCE IN VEHICLE LICENSE PLATE DETECTION ON A SMART GATE SYSTEM BASED ON VIDEO ANALYTICS UNDER VARIOUS LIGHTING CONDITIONS

• Main Author: Jerry Josia Partogi, Sitanggang
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/85143
• Institution: Institut Teknologi Bandung

This study explores the comparison of YOLO and Optical Character Recognition (OCR) technologies in a vehicle license plate detection system applied to a smart gate. The primary focus of this research is on the implementation and evaluation of these two technologies in detecting vehicle license plates under various lighting conditions. The developed system utilizes YOLO technology for object detection and OCR for character extraction from vehicle license plates passing through the smart gate. The analysis was conducted to measure the accuracy of license plate detection based on lighting variations during day and night. The results of the study indicate differences in performance between YOLO and OCR under different lighting conditions, emphasizing how each technology adapts to the visual challenges encountered. This research demonstrates that the developed system is capable of accurately detecting vehicle license plates under various lighting conditions and viewing angles. The implementation of this system is expected to enhance the effectiveness of security at gate locations requiring strict supervision, as well as reduce the risk of violations and vehicle theft.