University of Surrey

Test tubes in the lab Research in the ATI Dance Research


Franceschetti, G, , Guida, R, Iodice, A, , Riccio, D, , Ruello, G, and Stilla, U, (2007) ELECTROMAGNETIC FEATURE EXTRACTION FROM REAL SAR IMAGES OF BUILT-UP AREAS In: 27th EARSeL Symposium: Geoinformation in Europe, 2007-06-04 - 2007-06-07, Bolzano, Italy.

Full text not available from this repository.


Nowadays, monitoring and management of built-up areas is becoming a crucial endeavor for all civil and politic communities. In the upcoming future, important contributions to this matter will come from scientific community. In fact, new powerful spaceborne sensors for monitoring urban areas have been designed and are ready for launch. In particular, a Synthetic Aperture Radar (SAR) called TerraSAR-X with outstanding performance of resolution will be soon launched by the German Aerospace Center (DLR) [1]. Availability of more detailed SAR images could improve, at least in theory, human interpretation as more information is contained in the image itself. But the extraction of this both geometric and electromagnetic information, is still a hard task. With regards geometric features of urban areas, many steps have been done in building shape and dimensions retrieval and are widely documented in literature [2],[3]. But extraction concerning electromagnetic features of a building (i.e. dielectric constants of its parts) is not yet a concrete result even if it would be of paramount importance for the impact that a similar retrieval would have on applications. For example, let us think of the possibility of checking the presence of water infiltrations in building walls or of amianthus, widely employed in the past in building industry as insulating material but now forbidden by the law in many countries for its demonstrated carcinogenic properties. In this paper, capability of retrieving electromagnetic information from high resolution SAR images is inspected. To our aim, we need scattering models taking into considerations the electromagnetic properties of the building materials when the radar signal interacts with a building in the scene. Some empirical models have been developed for radar backscattering from soils, crops, forest vegetation etc., but for urban geometries they are still inadequate. Despite of this, a complete electromagnetic backscattering model for a canonical isolated structure placed on a rough terrain [4] is here employed in simulation examples. In order to foresee our future ability to retrieve electromagnetic information from high resolution SAR images of urban areas, some experiments have been jointly carried out by the remote sensing groups at the University of Naples (Italy) and the University of Munich (Germany). The test area is located in Munich and shows the buildings of Technische Universität and the Alte Pinakothek in the quarter of Schwabing. The radar functioning mode parameters that have been set are responsible for the high resolution got in the final SAR image which needs to be interpreted after being processed. In this paper, the authors first simulate, thanks to the availability of a Digital Elevation Model (DEM), the SAR raw signal relative to the same area illuminated during the flight campaign. In this first simulation only contributions of the first order to the backscattered signal are considered (i.e., single scattering) and no difference in materials or in soil roughness is taken into account. In this way, the attention mainly focuses on the geometrical properties of the scene and on their effects on the SAR image, which will be discussed at the conference. Subsequently, a different raw signal simulation has been realized in which only some buildings, with a particular electromagnetic behaviour, have been placed in the scene. Contributions to the backscattered field till the third order have been considered (i.e., single scattering, wall-ground and ground-wall double scattering, and wall-ground-wall triple scattering), but, to do this, some simplifications on the geometrical models needed to be accomplished according to the simulator requirements. On the other hand, in this way different materials could be assumed even for the same building, thus obtaining a more realistic simulation. The interested area has been visited several times to collect some samples of materials and achieve a better simulated representation of the area itself. Finally, the SAR simulated images are compared to the real one. Based on this comparison, some previously unexplained features of the real SAR image are satisfactorily interpreted. At the conference the authors will discuss the simulation-based interpretation of the considered SAR image, particularly analyzing the possibility to retrieve electromagnetic information about the buildings in the scene when a model-based approach is employed. The most interesting results will be widely inspected and presented. [1]Roth, R. Weninghaus “Status of the TerraSAR-X Mission”, IEEE International Geoscience and Remote Sensing Symposium (IGARSS) on CD. [2]M.Quartulli, M.Datcu, “Stochastic Geometrical Modeling for Built-Up Area Understanding from a Single SAR Intensity Image with Meter Resolution”, IEEE Trans. Geosc. Remote Sensing, vol.42, pp.1996 2003, 2004. [3]R.Bolter, “Reconstruction of Man-Made Objects from High Resolution SAR Images”, IEEE Aerospace Conference Proceedings, vol.3, pp.287-292, 2000. [4]G.Franceschetti, A.Iodice, D.Riccio, G.Ruello “SAR raw signal simulation for urban structures”, IEEE Trans. Geosc. Remote Sensing, vol.41, pp.1986 1995, 2003.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Divisions : Surrey research (other units)
Authors :
Franceschetti, G,
Iodice, A,
Riccio, D,
Ruello, G,
Stilla, U,
Date : 4 June 2007
Depositing User : Symplectic Elements
Date Deposited : 17 May 2017 12:34
Last Modified : 23 Jan 2020 17:56

Actions (login required)

View Item View Item


Downloads per month over past year

Information about this web site

© The University of Surrey, Guildford, Surrey, GU2 7XH, United Kingdom.
+44 (0)1483 300800