ETD system

Electronic theses and dissertations repository


Tesi etd-06012011-113952

Thesis type
Tesi di laurea specialistica
Spaceborne Radar Tomography for Urban Applications
Corso di studi
relatore Dott. Lombardini, Fabrizio
correlatore Ing. Cai, Francesco
relatore Prof. Berizzi, Fabrizio
Parole chiave
  • Spaceborne Tomography
  • Model Order Selection
  • Autofocusing
  • Synthetic Aperture Radar
Data inizio appello
Data di rilascio
Riassunto analitico
Three-dimensional (3-D) Synthetic Aperture Radar (SAR) imaging is a recent
technique, based on coherent SAR data combination, and aims to obtain a full
3-D analysis in space. It is a multibaseline (MB) extension of the SAR interferometry concept and off ers new options for the analysis and monitoring of ground scenes by means of the capability of separating the scattering phenomena along
the height dimension. In the last few years, the monitoring of complex urban
or infrastructure areas with layover "garbled" scatterers by means of MB SAR
array has grown into a valuable and indispensable tool. However, Tomo-SAR is typically aff ected by problems of unsatisfactory height sidelobe behavior and resolution, due to the low number and irregular sampling of baselines. In this context, superresolution techniques posses interesting capabilities, but they require
a multilook processing that reduces horizontal resolution. The main objective of
this work, is the implementation of a tomographic processing able to work at full
horizontal resolution, desirable in an urban scenario. This method exploits an
interpolation step in order to reconstruct the data in a way that allows the application of the superresolution techniques even with single-look data. This new Tomo-SAR processor off ers both height superresolution and sidelobe cleaning at
full horizontal resolution. Results from ERS satellite data are presented.
Moreover, an hybrid 3D adaptive tomography-complex data domain model fi tting detection method is proposed, in order to individuate permanent single and double scatterer, by selection of two thresholds based on radar reflectivity(SNR)
and fi tting error. A fi rst validation of results is proposed by geolocalization of Single and Double scatterers.
The Thesis is organized as follows:
Chapter 1: a short introduction to layover problem and SAR Tomography.
Chapter 2:Spectral estimation of array complex data. Two processing algorithms are briefly discussed: Adaptive (Capon) and Model-based (MUSIC, MUltiple SIgnal Classi cation).
Chapter 3: implementation of a Single-Look MB SAR Tomography with full analysis and optimization of several parameters.
Chapter 4: Model Order Selection (MOS) by exploiting LS fi tting model algorithm. Results are geolocalized and superimposed on an ortophoto.
Chapter 5: array phase errors estimation and calibration using automatic SAR
focusing (named Autofocusing) processing. This last part of the work was conducted with the collaboration of University of Rennes-1 (France).