Année
2025
Rôle
System design, API and database layer, deployment
Technologies
- YOLOv5
- OCR
- FastAPI
- PostgreSQL
- React
- Docker
Applied computer vision
License Plate Recognition System
A real-time license plate recognition system: a fine-tuned YOLOv5 model with OCR, a FastAPI and PostgreSQL backend and a React frontend, deployed with Docker.
- Developed a real-time license plate recognition system by integrating a fine-tuned YOLOv5 model with OCR for text extraction.
- Focused on system development: designed the pipeline, built REST APIs with FastAPI, managed data in PostgreSQL, and created a ReactJS-based frontend. Deployed the system with Docker for scalability and ease of integration.
- Achieved real-time processing at ~12 FPS with reliable recognition accuracy.
- Enhanced traffic management efficiency and enabled citizens to conveniently check their traffic violation records online.
- Throughput
- ~12 FPS
- real-time processing
