Contribution to the Detection of Cyclones by Coupling the Convolutional Neural Network With the Barycentric Calculation of an Image

The material and human losses caused by the passage of a cyclone are becoming increasingly significant for countries located in tropical areas. Under the effect of climate change, we see from the analysis of data published each year that the power of these tropical cyclones multiplies over time. The direct impacts left by the passage of a strong storm, such as flooding, destruction of infrastructure and especially human loss, increase each year.