Inspiration: How do we use AI to help expedite climate change mitigation.
What it does: Accurately maps land in near real-time to better understand size and location of plots of land.
How we built it: Mix of Segmentation Anything Model and Sentinel 2B satellite images.
Challenges we ran into: Pixelation of images, connecting with a seamless UI, accounting for seasonality of images.
Accomplishments that we're proud of: The overarching accuracy of the output as a means to remove the time it takes to create field boundaries. Getting these maps to be easily understood by the user was no joke!
What we learned: Application of SAM, consolidation of Sentinel 2B images while ensuring they are geo-referenced.
What's next for Poly Plot: More building! Improve UI, automate the Sentinel 2B image consolidation, create fluid backend to front end pipeline, apply to any and all field boundary dependent businesses!