'Multi-temporal InSAR data analysis with Structural Health Monitoring applications' Dr Gabriel Martin Hernandez
Abstract:
Monitoring large areas of the Earth's surface is possible thanks to the availability of remote sensing data obtained by a large collection of diverse satellites orbiting the Earth. Specifically, Multi-temporal Interferometric Synthetic Aperture Radar (MT-InSAR) techniques supply the structural community with time series of Line of Sight (LOS) displacements of terrain or structures such as bridges or buildings resulting from causes such as thermal expansion contraction, and terrain deformation. The analysis of the different deformation signals observed is crucial to identify the different phenomena that causes these deformations. In this talk we will explore the possibility of using different machine learning methods with the aim of automatic identification of different deformation patterns both in buildings and structures, that help to monitor civil infrastructures at scale, providing relevant insights to asset owners regarding the performance of such structures over time. Two different approaches will be described, first an approach based on Deep Learning methods will be introduced. After that a second approach based on a blind source separation algorithm will be proposed to circumvent some limitations of the previously proposed methods.Our results demonstrate that InSAR time-series analysis can benefit from the use of blind source separation approaches, which opens the door for monitoring infrastructure at scale over very large areas.