OrbitalDoc

• 

  Dark Mode Main Page

• 

  Light Mode Main Page

• 

  ImageSlider for two images(Salt lake City)

• 

  Shows a Summary of the two images displayed

🌍 Inspiration Natural disasters, deforestation, and climate change often go unnoticed until it's too late. We wanted to build a tool that makes satellite image analysis accessible to anyone—especially researchers, journalists, and emergency responders—so they can track changes in the environment quickly and visually.

🔍 What it does OrbitalDoc allows users to:

Select a region on the map

Choose two dates for satellite image comparison

Instantly visualize before-and-after imagery

Get AI-driven predictions of substantial environmental change

See change scores, cloud coverage, and pixel change counts

Toggle between dark and light mode

View everything with a clean and responsive UI

🛠️ How we built it Frontend: React + Leaflet for interactive maps, custom UI components, and image sliders

Backend: Flask handles API requests, satellite fetching via SentinelHub, image processing with OpenCV & NumPy, and ML predictions

ML Model: A CNN trained on labeled before-after image pairs, deployed via TensorFlow

Tools: Netlify (frontend), GitHub, Python, JS, HTML/CSS

🧱 Challenges we ran into Accessing and formatting satellite data efficiently

Managing dependencies and model file locations between frontend/backend

Building a change detection model with a limited dataset in a short time

Handling CORS and image serving cleanly in Flask

🏆 Accomplishments that we're proud of End-to-end functional app with live satellite fetching + AI detection

Custom-built and trained our own change detection CNN

Fully responsive and user-friendly design

All core features (and more) completed within the hackathon timeline

📚 What we learned How to integrate AI models into full-stack apps

Building scalable APIs with Flask

Responsive design + accessibility considerations

Real-world use of geospatial tools like SentinelHub

🚀 What's next for OrbitalDoc Add support for multi-date timelines and GIF exports

Enhance AI model with larger training datasets and multi-class detection

Deploy the Flask backend for full production use

Add user login & saved analysis reports

Explore wildfire detection and disaster monitoring integrations

Built With

• https://github.com/muazamsyed98/orbitaldoc