After looking online most solar calculators for existing homes are outrageously complicated and time-consuming. This does not fit the possibilities of extracting all the demanded information from high-quality satellite images. To reduce the hurdles for possible homeowners to get a quote the only thing that should be of interest is the address.

What it does

We take that address and detect the corresponding roof and the available area on it to then give the homeowner a quote on how much money they might be able to save by going solar.

How we built it

With the use of OpenCV functionality for segmenting the images, which we got trough MapBox. The information we could retrieve is used to create a plot.

Challenges we ran into

It was difficult to consistently segment out the roof for complex buildings especially including adjacent garages and similar.

Accomplishments that we're proud of

That it works!

What we learned

A bunch of OpenCV, React and Django!

What's next for MINimum CO2

The predictions need to be improved by detecting objects on the roof, that impede the positioning of solar panels.

Built With

Share this project: