Deep learning is a subset of machine learning based on learning data representations, in contrast to task-specific algorithms. Deep learning models derive inspiration from how information is processed in the nervous system of the human body that consists of trillions of neurons communicating with each other. During a disaster, it is necessary to ensure that containment and rescue operations are conducted as quickly as possible with a primary focus on affected areas, as an improper organization might lead to wastage of resources such as money, materials, and time. To properly plan during disasters, satellite images of the affected location can be analyzed to identify the areas demanding immediate attention. A model can be designed using convolutional neural networks (CNNs) to help categorize the areas by the degree of destruction. To secure data fed into the model, a layer of security can be added between the input and output layers of the CNN. The model can be trained using old satellite images of the cities. New images fed into the model can be analyzed to obtain information on the level of devastation.