Inspiration
My little cousin is a cute kid, but she's a troublemaker. She bought a drone on my Amazon account and thought I wouldn't notice.
Well, joke's on her. Now I have a drone and had to figure out something to build with it.
That's how ERFMap was born.
What is it?
ERFMap (Environment Revision Flyover Map) is an end-to-end surveying and green space design tool.
Professional land surveys can cost anywhere from $500 to over $1,500. Small communities shouldn't have to choose between improving their public spaces and breaking their budgets.
Our solution? A toy drone and a measuring tape.
What it does
Fly your drone over an area, take one reference measurement, and let ERFMap do the rest.
Using computer vision, deep learning, and 3D reconstruction, ERFMap identifies buildings, paths, trees, grass, benches, and other features to generate a scaled survey with relative elevation contour lines.
But that's just the beginning. You can:
- Design: Add, remove, and reposition features.
- Analyze: Estimate costs, ecological impact, and community satisfaction.
- Explore: Walk through your environment in 3D.
- Collaborate: Share and improve designs with others.
- Export: Save your surveys and 3D models.
I could explain everything ERFMap does for another 15 minutes, but by then it could have already generated your survey. Just check it out!
How we built it
We combined Python, computer vision, deep learning, and photogrammetry to turn ordinary drone footage into usable environmental data.
NodeODM/OpenDroneMap handles 3D reconstruction, while our image processing and classification pipeline identifies environmental features. Our backend connects everything to an interactive web frontend where users can survey, design, and explore their environments.
The biggest challenge was getting all these technologies working together seamlessly.
Challenges we ran into
Our drone died halfway through development. Yes, the drone that inspired the project.
Luckily, someone lent us theirs.
Beyond that, estimating elevation and reconstructing accurate environments without LiDAR or expensive surveying equipment was much harder than expected.
Accomplishments that we're proud of
Contour lines.
Nearly 24 hours of struggling with 3D reconstruction and elevation calculations, but we finally got them working.
We're also proud of combining surveying, classification, 3D modeling, and environmental design into one platform.
What's next for ERFMap?
More training, better accuracy, and bigger ambitions.
We want to improve our deep learning models, strengthen reconstruction accuracy, and expand our environmental simulations.
Ultimately, we hope ERFMap becomes a practical, affordable tool for civil engineers, green space designers, and communities everywhere.
All because my little cousin bought a drone she wasn't supposed to.
Built With
- javascript
- node.js
- opendronemap
- python
