Search

Decorrelation is a major drawback in Synthetic Aperture Radar (SAR) interferometry (InSAR) processing, causing poor density of measurement points, mainly outside of urbanized areas. SAR Tomography is one of the most suitable multi-temporal InSAR technique for built environment monitoring. Although typically capable of identifying a large number of scatterers, SAR Tomography detectors suffer the assumption of fully coherent scatterer response, which can be restrictive in real cases. In this work we address the test of a detector of partially correlated scatterers on a real COSMO-SkyMed (CSK) dataset, which was previously derived in the literature [1] as a Rao test for the single-look case and is here extended to the multi-look scenario. Experimental results are compared with those achieved by the Multi-look Generalized Likelihood Ratio Test (MGLRT). Both Partial Correlation and MGLRT have been implemented with a fixed multi-look degree.

Interferometric and tomographic processing at full resolution of data acquired by very high resolution SAR systems allow providing accurate 3-D reconstruction of built environments, as well as deformation monitoring. In this framework, the detection of Persistent Scatterers (PSs) is a fundamental aspect. The exploitation of multiple observation looks (multilook), is an effective strategy which allows improving the coverage of monitored pixels in natural scenarios. In the framework of SAR tomographic processing, the Multilook Generalized Ratio Test (MGLRT) detection scheme has been proposed in literature. In this work we address the potentialities and main issues related to the exploitation of a multiresolution approach for MGLRT detection based on the identification of statistical homogeneous looks.

Approaches based on the coherent processing of Synthetic Aperture Radar (SAR) data archives acquired over long temporal intervals allows 3D reconstruction/imaging and monitoring of urban areas with very high accuracy on the velocity, e.g. sub-centimetric per year. The technology is rather mature and developed in some cases also at the level of operative services. Nevertheless research is still going on to improve the technology in terms of coverage and reliability of estimated parameters for correct localization and accurate monitoring of targets. With this regard, the application to regions characterized by a mixture of urban and rural areas is of interest. This work discusses the different techniques for filtering and selection of scattering mechanisms and for single and multi-look approaches for the multi-resolution detection of Persistent Scatterers.

Persistent scatterers (PS) interferometry tools are extensively used for the monitoring of slow, long-term ground deformation. High spatial resolution is typically required in urban areas to cope with the variability of the signal, whereas in rural regions, multilook shall be implemented to improve the coverage of monitored areas. Along this line, SqueeSAR and later Component extrAction and sElection SAR (CAESAR) were introduced for the monitoring of both persistent and (decorrelating) distributed scatterers (DS). Multilook generalized likelihood ratio test (MGLRT) is a detector derived in the context of tomographic SAR processing that has been investigated for a fixed multilook degree. In this work, we address MGLRT and CAESAR in the multiresolution context characterized by a spatially variable multilook degree. We compare the two schemes for the multiresolution selection of PS and DS, highlighting the pros and cons of each scheme, particularly the peculiarities of CAESAR that have important implications at the implementation stage. A performance analysis of both detectors in case of model mismatch is also addressed. Experiments carried out with data acquired by the COSMO-SkyMed constellation support both the theoretical argumentation and the results achieved by resorting to Monte Carlo simulations.

Synthetic Aperture Radar (SAR) tomography presents the advantage of multiple stable targets detection within same pixel. Fast-sup-GLRT (generalized likelihood ratio test based on support estimation) algorithm proved to be an ideal compromise between detection capabilities and computational complexity. In this work, a multi-look version of this detector which exploits the advantages of Capon estimation is examined. Statistical analysis of estimation and detection processes are conducted to compare the performances of sequential non-linear least-squares (NLLS) search and Capon filtering of projected data for double PS identification. Main objective is to exploit the super-resolution advantages of NLLS method without the risk of multiple stable targets classification from the same scattering contribution. For the last desiderate, an additional verification is included within the detection step.

Persistent scatterer (PS) interferometry and more recently synthetic aperture radar tomography have shown to be powerful tools in urban scenarios for providing 3-D point clouds in the reconstruction of buildings as well as in the monitoring of their possible slow temporal deformations. The detection of PSs represents a fundamental aspect, which in the literature has been mainly addressed at full resolution (single-look detection), thus considering only the scatterer coherence properties along the different acquisitions. In this paper, we investigate the benefits offered by the usage of multiple observation looks. Multi-look generalized likelihood ratio test detection schemes are derived and analyzed in terms of detection performances. The analysis shows that even a slight multi-look can provide a dramatic improvement on the detection capability both on simulated and real data, especially in the areas characterized by a low signal-to-noise ratio and in the presence of a limited number of acquisitions.

The general problem of multiple signal sources detection using an array of sensors, with the particular case of Synthetic Aperture Radar (SAR) Tomography is approached in this paper. A detection algorithm which generates accurate results but it's computationally demanding, sup-GLRT (Generalized Likelihood Ratio Test with support estimation), is discussed. Main objective consists in applying matrix algebra for optimization of 2D Non-Linear Least Squares (NLLS) search implementation, the most demanding part of the detector. Results of the detection algorithm on a dataset of high resolution SAR images are presented. Performance of the proposed optimization method is assessed by comparison of algorithm implementation times.

Sparse Representation and Compressive Sensing (CS) theory has gained an increasing interest during the last years in many application fields, including SAR tomography. The latter offers the possibility to perform a focusing, beyond the classical 2D (azimuth-range) domain, along other dimensions: f.i., elevation and velocity. The literature, however, lacks of an assessment of the improvement of CS over classical point cloud detection schemes based on the Generalized Likelihood Ratio test which, in the simplest form, use basic beamforming (matched filter) detection based schemes. This work aims to provide a contribution along this line.

Interferometric and particularly tomographic processing at full resolution of data acquired by very high resolution SAR systems allows accurate 3D imaging and monitoring of buildings and infrastructures. Recently it has been shown that implementing a light data multilooking leads to a dramatical increase of the density of monitored pixels. In this work we investigate different options for performing the the multilook operation in SAR Tomography and we analyze their performances in terms of detection capabilities of single scatterers.

In this chapter, experimental data from a Ka-band high-resolution radar system are used to evaluate the performance of several schemes for coherent detection of distributed targets. The importance of this analysis relies on the possibility of exploiting detectors for range-spread targets in modern applications for homeland security. A careful border control is crucial to avoid illegal immigration or drug trade by means of very small boats (such as inflatable or wooden boats). It is thus of great interest to assess the capability of a high-resolution radar system to detect such kind of small targets which, at the resolutions of the data available in this chapter (0.10m, 0.20m, and 1 m), appear range-spread.

Synthetic aperture radar (SAR) tomography is a technique that extends the concept of SAR interferometry for the accurate localization and monitoring of ground scatterers. Data that are being acquired by the new high resolution SAR sensors offer new perspectives in the 3-D reconstruction and monitoring of urban areas and, particularly, of individual buildings. SAR tomography allows increasing the density of measurements by handling situations where multiple stable scatterers interfere in the same resolution cell. The detection of reliable, i.e., persistent, scatterers is however a challenging issue. In this paper, we investigate three detection approaches: The first is based on a modification of information theoretical criteria; the last two are based on the generalized likelihood ratio test. Theoretical performances are analyzed in details on simulated data, and results of the application to real data from both medium and very high resolution sensors are also provided.

Multidimensional synthetic aperture radar (SAR) imaging is a technique based on coherent SAR data combination for space (full 3-D) and space deformation-velocity (4-D) analysis. It is an extension of the concepts of SAR interferometry and differential interferometry SAR and offers new options for the analysis and monitoring of ground scenes. In this paper, we consider the problem of detecting single scatterers for localization and monitoring issues. To this end, we resort to a constant false alarm rate (CFAR) detection scheme which can be synthesized according to three different design criteria: generalized likelihood ratio test, Rao test, and Wald test. At the analysis stage, the performance of the aforementioned detector is compared to that of a previously proposed CFAR scheme, based on the multi-interferogram complex coherence and widely used in persistent scatterer interferometry. The analysis is conducted both on simulated and on real SAR data, acquired by ERS-1/2 satellites. Finally, Cramer-Rao lower bounds for the estimation of the scatterer elevation and velocity are provided.

This work is focused on the detection of single and multiple scatterers in multiview/multitemporal SAR data in order to locate and monitor a high number of ground structures with low signal misinterpretation. This issue is addressed here by combining amplitude and phase data, differently from common techniques for the detection of single scatterers which rely on phase coherence measures. Experiments with real satellite C-band data are presented with both full resolution and multilook processing.

Multidimensional synthetic aperture radar (SAR) imaging is a technique based on coherent SAR data combination for space (full 3-D) and space/deformation-velocity (4-D) analysis. It extends SAR interferometry and differential interferometry concepts, offering new options for the analysis and monitoring of ground scenes. In this paper, we consider the problem of detecting scatterers showing partial correlation properties induced by simultaneous acquisitions from satellite formations or an uneven temporal distribution of satellite constellations. To this end, we first provide a general model for the pixel imaged by a multidimensional SAR system. Then, we design a constant false alarm rate (CFAR) decision rule accounting for the presence of partially coherent scatterers. At the analysis stage, we assess the performance of the new detector also in comparison with a previously proposed CFAR scheme, developed in the context of SAR tomography for fully coherent scatterers.