Hurricane damage detection using machine learning and deep learning techniques: A review

Abstract

Hurricane is one of the most disastrous natural disasters causing immense harm to the ecosystem and economic system worldwide. It is also known as a tropical cyclone. Heavy rainfall and high winds accompanying hurricane inflict damage to property as well as loss of human life. Hence, appropriate steps need to be taken to mitigate the damage caused by the disaster. Recently, social media platforms are used that help in providing immediate relief to the people affected by the disaster. Since, the difficulty arises in analysis of high volume data of social media, satellite imagery is also being used for damage detection due to its ability to cover large spatial and temporal areas. But manual damage detection is error prone. Therefore, machine learning and deep learning which automated methods are used for detection of damage. This paper includes the use of machine learning and deep learning for detection of damage caused by natural disasters with a special focus on hurricane damage.