Challenge Overview
Natural disasters such as hurricanes, tornadoes, earthquakes, and floods inflict significant damage on infrastructure each year costing lives, time, and billions of dollars. There is a need to rapidly create maps and assess damage to assess the scale of damage and direct first responders. Utilizing the datasets and winning algorithms of previous SpaceNet challenges applied to a real- world disaster response scenario we hope to find answers to these questions: ��� How has building and road extraction improved from prior challenges? ��� What is the impact on performance for a multi class feature extraction challenge? ��� How accurately can roads obstructed from flood waters be characterized from pre- event roads and post- event satellite imagery? from satellite imagery? ��� Do flood assessment algorithms generalize to other geographic areas? TODO
Please read above