Skip to Main content Skip to Navigation
New interface

Ship detection and characterization from SAR imagery linked with cooperative vessel tracking data

Ramona-Maria Pelich 1, 2 
Lab-STICC - Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance
Abstract : SAR imagery presents an increased interest in maritime surveillance applications. The research work completed in this thesis is dedicated to vessels detection and signature characterization from data acquired by different spaceborne SAR sensors. Firstly, we assess the performances of different ship detectors based on adaptive threshold algorithms. The detection algorithms are based on various clutter distributions and assessed automatically with a systematic methodology. Evaluation using large datasets of medium resolution SAR images and AIS (automatic identification system) data as ground truths allows to evaluate the efficiency of each detector. Depending on the datasets used for testing, the detection algorithms offer different advantages and disadvantages. The systematic method used in discriminating real detected targets and false alarms in order to determine the detection rate, allows us to perform an appropriate and consistent comparison of the detectors. The impact of SAR sensors characteristics (incidence angle, polarization, frequency and spatial resolution) is fully assessed, the vessels length being also considered. Experiments are conducted on Radarsat-2 and CosmoSkymed ScanSAR datasets and AIS data acquired by coastal stations. Secondly, the effects of stationary-based processing of moving ship signatures in SAR imagery are assessed and a methodology that makes it possible to estimate and compensate them is introduced. SAR imaging of moving targets usually results in residual chirps in the azimuthal SLC processed signal. The Fractional Fourier Transform (FrFT) allows to represent the SAR signal in a rotated joint time¿frequency plane and performs an optimal processing and analyse of chirp signals. Employing the FrFT reduces the effects of residual chirps achieving compensation of the along-track defocus of a moving target and estimation of the target¿s azimuthal speed itself. Experiments are conducted on Radarsat-2 Multilook Fine and Ultrafine SAR images. Evaluation using a large number of ship signatures allows to assess the efficiency of the proposed method. Comparisons with AIS data as ground truth and with a method based on the assessment of the temporal correlation between a sequence of sublook images are carried out for a complete performance analysis. Finally, the use of complex dual-polarization data for SAR vessel detection is assessed. As a first step, an intercomparison between the individual use of each polarimetric channel is considered, as well as the fusion of the detection results corresponding to the two polarimetric channels. In a second phase, the fusion of both polarization channels before the detection step is assessed. When dealing with amplitude data only, we propose to employ a method based on the generalized temporal moments (Hölder means), in order to fuse the information of both polarization channels. When dealing with complex data, the coherence coefficient or target dual-polarimetric decompositions, which may provide additional information in comparison with single channel imagery, are employed.
Complete list of metadata

Cited literature [123 references]  Display  Hide  Download
Contributor : Bibliothèque Télécom Bretagne Connect in order to contact the contributor
Submitted on : Thursday, February 11, 2016 - 11:32:46 AM
Last modification on : Monday, March 14, 2022 - 11:08:11 AM
Long-term archiving on: : Saturday, November 12, 2016 - 5:30:13 PM


2015telb0368_Pelich Ramona-Mar...
Files produced by the author(s)


  • HAL Id : tel-01272674, version 1


Ramona-Maria Pelich. Ship detection and characterization from SAR imagery linked with cooperative vessel tracking data. Signal and Image processing. Télécom Bretagne; Université de Bretagne Occidentale, 2015. English. ⟨NNT : ⟩. ⟨tel-01272674⟩



Record views


Files downloads