Ship Wake Detection in X-band SAR Images Using Sparse GMC Regularization

Author(s):  
Oktay Karakus ◽  
Alin Achim
Keyword(s):  
X Band ◽  
Author(s):  
Björn Tings ◽  
Sven Jacobsen ◽  
Stefan Wiehle ◽  
Egbert Schwarz ◽  
Holger Daedelow

Recent studies investigated the detectability of ship wake signatures on SAR imagery using a large number of SAR images collocated with Automatic Identification System data for training machine learning models. These detectability models are in agreement with oceanographic expectations from preceding studies and can therefore be used for comparing the performance of different SAR sensors in terms of wake detectability. Previous model comparisons showed better wake detection performance of TerraSAR-X (TS-X) than of RADARSAT-2 (RS2) and Sentinel-1 (S1). A comparison between CosmoSkymed (CSK) and RS2 is performed here, to examine the hypothesis that X-Band is generally better for wake detection than C-Band. Finally, this hypothesis is not confirmed, as the detectability models for TS-X, CSK and RS2 reveal similar performances. A comparison of wake detection performance should take the individual wake components into account separately.


2020 ◽  
Vol 58 (3) ◽  
pp. 1665-1677 ◽  
Author(s):  
Oktay Karakus ◽  
Igor Rizaev ◽  
Alin Achim

Sign in / Sign up

Export Citation Format

Share Document