---

ASTRA: Autonomous unmanned aerial System for synThetic aperture Radar, photogrAmmetry and digital image correlation

Abstract

The main goal of the project is the development of an autonomous Unmanned Aerial System (UAS) platform able to perform Synthetic Aperture Radar (SAR) Interferometry, photogrammetry, and Digital Image Correlation for monitoring both slopes (e.g. landslides and mines) and infrastructures (e.g. bridges, towers, urban sites).

The key idea of this project is to exploit the huge potential of autonomous UAS as a platform that could overcome the gap between space/air-borne platforms and terrestrial installations. Space/air-borne platforms are able to cover large areas, but their use has strong limitations in terms of flexibility and cost. Terrestrial installations are much cheaper and can acquire data at much higher sampling time, but their field of view is limited by orography. UAS platforms could be the answer for many applicative cases when a satellite or an airplane does not provide adequate time coverage, and a terrestrial installation is not effective. In this context the capability of autonomous flight is a qualifying feature. SAR acquisitions are as much better as the flight is well-controlled. Ideally, the flight should be a repeatable straight line. An autonomous flight could be much closer to this condition than a flight piloted by an human operator.

Another key idea of this project is the integration with photogrammetry provided by aboard cameras. Photogrammetry can provide 3D maps of the monitored area, i.e. a digital elevation model (DEM). This can be useful by itself, but it is especially useful as an ancillary tool for SAR Interferometry. Furthermore, photogrammetry can help to accurately trace the path of the UAS. The accuracy of positioning is a critical issue for obtaining high quality SAR images. Finally, the cameras aboard UAS could be used for acquiring georeferenced images of the field of view. These images could be used for detecting medium-long term displacements as pixel changes using the image processing techniques referring to Digital Image Correlation. 

Last but not least, a further key idea of this project is the use of Digital Image Correlation technique also for radar images with the aim to detect displacements in medium-long terms.  

This technique is really complementary to SAR interferometry, as the latter is able to detect small displacements (sub-centimeter), but it is affected by a fundamental phase ambiguity; on the other hand, Digital Image Correlation cannot reach the accuracy of interferometry but it is a robust technique for detecting large displacement in medium-long term.

Finally, a qualifying point of this project is the in-field validation. The equipment will be tested in realistic test sites that will be selected during the project.