FalconVision

FalconVision is an offline-capable drone search system designed to locate missing persons in mountainous terrain using computer vision and YOLOv11. It includes a streamlined interface built with Electron and Vite for video analysis and model customization.

Features

• ✈️ Local drone video analysis (no internet required)
• 🔍 Person detection using fine-tuned YOLOv11 models
• 🔠 Choose between pre-trained models or custom YOLO weights
• ⏳ Adjustable frame skipping for faster processing
• ⌚ Live progress bar and detection results
• 📄 PDF export of all detected frames, including confidence, timestamps and locations
• 📄 Multilingual UI (German / English toggle)

Technologies

• Frontend: React + Vite + Tailwind + Electron
• Backend: Python + Flask + Ultralytics YOLOv11

Installation

This repo contains the full source code. For installation instructions and binaries, see the Releases page.

If you're building locally:

# Install dependencies
cd frontend && npm install
cd ../backend && python3 -m venv venv && source venv/bin/activate && pip install -r requirements.txt

# Start backend
cd backend && python app.py

# Start frontend (in another terminal)
cd frontend && npm run start

Project Structure

FalconVision/
├── backend/       # Flask + YOLOv11 inference server
├── frontend/      # Electron + React video analysis interface
├── uploads/       # Temporary folder for uploaded video frames

Build the Electron App

To create a cross-platform distributable version:

cd frontend
npm run dist

This creates .AppImage, .exe and other binaries under frontend/dist.

---

License

MIT License. Built with ❤️ by Ruchit Kumar Patel and Frans Dressler @ FalconVision.