Automating administrative tasks can save significant time and resources. I developed a system to detect vehicle parking passes with a high accuracy of 96%, leveraging YOLOv3 for real-time object detection and Google Cloud AI's OCR capabilities for text recognition.
System Architecture
The backend of this project is built with Python and Flask, where YOLO handles the detection of parking passes. The system is trained on custom-labeled data to ensure high accuracy. Once detected, the Google OCR API extracts pass identification numbers, automating the record-keeping process.
Real-Time Visualization
To provide a comprehensive solution for school administration, I developed a frontend using HTML/CSS and JavaScript. This interface allows administrators to visualize real-time detection results, making the system not only accurate but also user-friendly.
Conclusion
The YOLOv3-based vehicle parking pass detector showcases the power of combining object detection with optical character recognition. This project highlights the potential for automation in administrative tasks, leading to increased efficiency and accuracy in data management.