Deformation monitoring by multipass synthetic aperture radar (SAR) interferometry is widely recognized as the only method to assess millimeter-level deformation over a large area from space. Recently, it is demonstrated that urban infrastructures can be monitored in a semantic level in SAR interferometry (InSAR) by fusing the InSAR point cloud with the classification labels derived from optical images [1], [2]. This paper proposes an algorithm for object-level joint InSAR deformation reconstruction using these classification labels. We derived an object-based multi-baseline InSAR reconstruction model, and proposed an efficient algorithm for bridge detection in optical images.
«
Deformation monitoring by multipass synthetic aperture radar (SAR) interferometry is widely recognized as the only method to assess millimeter-level deformation over a large area from space. Recently, it is demonstrated that urban infrastructures can be monitored in a semantic level in SAR interferometry (InSAR) by fusing the InSAR point cloud with the classification labels derived from optical images [1], [2]. This paper proposes an algorithm for object-level joint InSAR deformation reconstruct...
»