Crowdsourced-based deep convolutional networks for urban flood depth mapping

Successful flood recovery and evacuation require access to reliable depth information. Most existing mapping tools do not provide real-time maps of inundated streets in around residential areas. In this paper, a deep convolutional network is used determine with high spatial resolution by analyzing crowdsourced images submerged traffic signs. Testing the model on photos from recent U.S. Canada yields mean absolute error 6.978 in., which par previous studies, thus demonstrating applicability approach low-cost, accurate, risk mapping.