Overview_Eng

Project Introduction

DVA LAB (Drone Video Analysis LAB) aims to conduct AI-based research and technology development utilizing drone video analysis. Our primary research areas include marine environmental protection and AI applications leveraging video analytics. We specifically focus on developing technologies to protect Indo-Pacific bottlenose dolphins in Jeju Island by analyzing drone footage and GPS data. Our research includes measuring distances between dolphins and vessels and detecting illegal dolphin-watching activities.

This document provides an overview of the DIVA service architecture developed by DVA LAB and guides you to quickly locate relevant documents.

📂 Document Structure

The DVA LAB documentation is structured to enable easy access to essential information for users, operators, and developers.

Category	Description	Link
📖 DIVA User Guidebook	Functional descriptions and tutorials for service users.	User Guide (Korean) User Guide (English)
🤖 DVA AI Parts	AI model structures and training methodologies.	AI Parts
🛠 DVA Service Parts	Information regarding service operation and maintenance.	Service Parts
👥 Contributors	Project contributors and guidelines for code contributions.	Contributors

📌 Each document can be accessed quickly via the dropdown menu at the top left corner.

🏗 Service Architecture

• Frontend: React-based UI (Vite + Tailwind + Zustand + TypeScript)

• FE-Admin: Node.js-based administrative interface (Docker + PostgreSQL)

• Backend: FastAPI-based API server (PostgreSQL + Celery + RabbitMQ)

• AI Models: AI models for video analysis and distance measurement (YOLO + SAM + BentoML)

• Infrastructure: Kubernetes cluster management (Helm Chart + Nginx Ingress + MinIO)

• Monitoring: Prometheus & Grafana-based service monitoring

🚀 Getting Started

✅ For Users

• Refer to the User Guide document.

• Try out the service on the DIVA website.

🛠 For Operators

• Read the Service Operation Guide for deployment and maintenance procedures.

• Consult the MLOps System Operation Guide.

💻 For AI Researchers

• Refer to the AI Parts document.

• Explore the models, datasets, and insights gained through the project's trials and errors.

❓ Need Help?

For additional inquiries, please reach out through GitHub Issues.

Documentation is continually updated; please check GitBook regularly for the latest information!