Image Classification of Damage of Houses hit by the hurricane based on Convolution Neural Network

The world is confronting multiple huge hurricanes every year, and people are suffering. So, it is necessary to design a scheme to save as many as people possible during the catastrophe. In this paper, we will introduce a CNN neuron model to identify a set of pictures to types of before the hurricane and after the hurricane. Then, we need to experiment with the drop-out layer and image augmentation to refine the model. First, we compare the accuracy under three situations (with drop-out layer, strengthen the data, and none). Then we work out the best way of training the data. We finally got the classification accurate with 96.97% in number.