Backtick among finalists in the Copernicus Hackathon 2020

Detecting snow and clouds in arctic conditions using computer vision
Posted by Michal Stypa on 2020-12-16

A couple of months back, we decided to participate in the Copernicus Hackathon organized by Arctic Business and Innovatum Startup. The objective was to predict road conditions using data from Copernicus satellites. Oskar and Fredrik were up for the task and managed to secure a place among finalists πŸ‘πŸ»

Their solution, ARCIF (Arctic Road Condition Indexing & Forecasts), managed to achieve a 91% (pixel-wise) prediction accuracy for snow and cloud detection, making it possible to determine road conditions when combined with road locations from sources like Open Street Map. You can read more about the competition and the solution in an interview with AI-sweden.

In the usual Backtick manner, we also wrote a blog post about it.

Photos by:
Josefin Widell Hultgren