Natural disasters detection and classification based on deep learning

With the technological development that we are witnessing at all times, and the height of the spread of artificial intelligence in life, all fields have become dependent on it to solve problems. In recent years, most countries have perceived large natural disasters in a very harmful way, including floods and fires, and they sometimes spread to residential areas. These disasters caused great damage from suffocation, burns, drownings, deaths, and others, or take safety precautions. Therefore, in the interest of the safety of the environment and people, and taking advantage of the advantages of artificial intelligence, we proposed the idea of detecting disasters, protecting the environment, and preserving people's lives so that safety measures are taken. The proposed system detects the presence of disaster after image processing and then classified it if it is flood or fire based on Convolutional Neural Network. We achieved an excellent accuracy rate of 99.77% compared with other state of art works.