Design and Develop Automatic Light Vehicles (LV) Unit Number Identification by Image Recognition Using Computer Vision Artificial Intelligent
Abstract
This thesis presents the development of an AI-driven system to automate refueling data recording, improving accuracy, efficiency, and cost-effectiveness. The research aims to identify the LV unit and driver's name, verify custodianship cost centers, and record refueling volumes. A comprehensive and accessible refueling database is established for immediate data availability. The study employs the V model for Mechatronics System Design, utilizing the Raspberry Pi 4 model B 8 GB, a 5-inch HDMI screen, and a webcam for capturing LV unit numbers and ID badges. The EAST algorithm locates text, while OCR Tesseract extracts information. The data is stored in a SQLite database using Python. Error rate and OCR accuracy are analyzed, and system duration time is compared with the existing system. Results show that computer vision and AI effectively identify LV unit numbers and driver's names at fuel stations. The proposed system offers cost-effective, accurate refueling data, including date, time, LV number, custodian cost center, and driver. Implementation streamlines operations and provides comprehensive data for decision-making.