GreenGuard

A tech-powered initiative to safeguard urban greenery and empower citizens for environmental impact.

Inspiration

The rapid urbanization of cities often comes at the cost of green cover. During one of our morning walks, we noticed that a previously lush area had turned into concrete within a few months—yet there had been no awareness or warning. This sparked the idea:

What if we could use technology to keep an eye on our trees? We wanted to build something that doesn’t just analyze the environment but actively involves citizens and municipalities in urban sustainability efforts.

What it does

Displays green cover heatmaps using satellite NDVI data Detects tree loss in urban neighborhoods using ML models Suggests tree-planting spots based on open urban spaces Lets users "Adopt a Tree" and monitor its growth Sends alerts to users and authorities about tree cover changes

How we built it

Frontend: Streamlit (Web UI) React Native (Mobile UI) Mapping Google Maps API Sentinel-2 imagery Leaflet.js for map rendering Machine Learning UNet model for tree cover segmentation PyTorch for training NDVI computation and preprocessing Backend: Flask for ML model serving Node.js + Express for main API PostgreSQL + PostGIS for spatial storage

Challenges we ran into

Cloudy satellite images affecting model accuracy Inconsistent resolution of free satellite data Difficulties segmenting small patches of vegetation Designing engaging user flows for tree adoption

Accomplishments that we're proud of

Achieved 85%+ accuracy in tree cover change detection Created an intuitive, interactive heatmap visualizer Enabled community engagement via tree adoption system Used only open data sources and free-tier cloud services

What we learned

Using NDVI and satellite data for environmental tracking Training deep learning segmentation models for geospatial use Importance of geospatial databases like PostGIS Designing for both usability and impact in climate-tech

What's next for GreenGuard

Add drone and street-level camera integration Expand to global cities with modular architecture Launch a gamified mobile app experience Release an open API for NGOs and researchers Partner with municipalities for ground action

Built With

• django
• google
• javascript
• maps
• python
• tensorflow
• torch