Epicenter Detection integrating Sentinel-1 InSAR data and Deep Learning

This master thesis explores a new approach to improving seismic monitoring, which is often characterized by poor efficiency and accuracy based on current methods. Traditional approaches to generate interferograms from satellite radar data can take a long time and human analysis. In addition, the complexity of SAR image processing makes it difficult to obtain timely information as a disaster response. The automation of this process is due to the need for faster and more accurate seismic assessments, and the desire to eliminate delays caused by manual data interpretation. This could reduce the possibility of human error and improve the overall reliability of seismic risk assessment. The aim is to incorporate InSAR data and machine learning technologies into a system that functions efficiently and minimizes the human intervention. The study highlights the potential of integrating remote sensing and deep learning models to optimise earthquake monitoring, in order to achieve faster and more accurate information for disaster preparedness and response. This research aims to lay the foundations for future improvements that will allow real-time seismic monitoring.