Alin Achim

Former Postdoc, University of Bristol

Contact :

Last publications in Ariana Research Group :

SAR Image Filtering Based on the Heavy-Tailed Rayleigh Model. A. Achim and E.E. Kuruoglu and J. Zerubia. IEEE Trans. on Image Processing, 15(9): pages 2686-2693, September 2006. Keywords : SAR Images. @ARTICLE{jz_ieee_tr_ip_06, author = {Achim, A. and Kuruoglu, E.E. and Zerubia, J.}, title = {SAR Image Filtering Based on the Heavy-Tailed Rayleigh Model}, year = {2006}, month = {September}, journal = {IEEE Trans. on Image Processing}, volume = {15}, number = {9}, pages = {2686-2693}, pdf = {http://dx.doi.org/10.1109/TIP.2006.877362}, keyword = {SAR Images} } Abstract : Synthetic aperture radar (SAR) images are inherently affected by a signal dependent noise known as speckle, which is due to the radar wave coherence. In this paper, we propose a novel adaptive despeckling filter and derive a maximum a posteriori (MAP) estimator for the radar cross section (RCS). We first employ a logarithmic transformation to change the multiplicative speckle into additive noise. We model the RCS using the recently introduced heavy-tailed Rayleigh density function, which was derived based on the assumption that the real and imaginary parts of the received complex signal are best described using the alpha-stable family of distribution. We estimate model parameters from noisy observations by means of second-kind statistics theory, which relies on the Mellin transform. Finally, we compare the proposed algorithm with several classical speckle filters applied on actual SAR images. Experimental results show that the homomorphic MAP filter based on the heavy-tailed Rayleigh prior for the RCS is among the best for speckle removal Maximum A Posteriori Estimation of Radar Cross Section in SAR Images using the Heavy-Tailed Rayleigh Model. A. Achim and E.E. Kuruoglu and J. Zerubia. In Proc. European Signal Processing Conference (EUSIPCO), Antalya, Turkey, September 2005. @INPROCEEDINGS{achim_eusipco_05, author = {Achim, A. and Kuruoglu, E.E. and Zerubia, J.}, title = {Maximum A Posteriori Estimation of Radar Cross Section in SAR Images using the Heavy-Tailed Rayleigh Model}, year = {2005}, month = {September}, booktitle = {Proc. European Signal Processing Conference (EUSIPCO)}, address = {Antalya, Turkey}, pdf = {http://kilyos.ee.bilkent.edu.tr/~signal/defevent/papers/cr1741.pdf}, keyword = {} } SAR Image Filtering Based on the Heavy-Tailed Rayleigh Model. A. Achim and E.E. Kuruoglu and J. Zerubia. Research Report 5493, INRIA, France, February 2005. Keywords : Synthetic Aperture Radar (SAR), MAP estimation, Alpha-stable distribution, Mellin transform. @TECHREPORT{5493, author = {Achim, A. and Kuruoglu, E.E. and Zerubia, J.}, title = {SAR Image Filtering Based on the Heavy-Tailed Rayleigh Model}, year = {2005}, month = {February}, institution = {INRIA}, type = {Research Report}, number = {5493}, address = {France}, url = {https://hal.inria.fr/inria-00070514}, pdf = {https://hal.inria.fr/file/index/docid/70514/filename/RR-5493.pdf}, ps = {https://hal.inria.fr/docs/00/07/05/14/PS/RR-5493.ps}, keyword = {Synthetic Aperture Radar (SAR), MAP estimation, Alpha-stable distribution, Mellin transform} } Résumé : Les images issues d'un radar à synthèse d'ouverture (RSO) sont affectées de manière inhérente par un bruit dépendant du signal, généralement connu sous le nom de bruit de chatoiement et qui est dû à la cohérence de l'onde radar. Dans ce rapport, nous proposons un nouveau filtre adaptatif pour débruiter les images RSO et nous déduisons un estimateur du maximum a posteriori (MAP) pour la section efficace du diagramme de gain en radar. On utilise d'abord une transformée logarithmique afin de changer le bruit multiplicatif en bruit additif. Nous modélisons la section efficace à l'aide d'une densité de probabilité récemment introduite - la densité de Rayleigh à queue lourde, qui a été obtenue en supposant que les parties réelles et imaginaires du signal complexe reçu peuvent être mieux caractérisées à l'aide de la famille des distributions alpha-stables. Nous estimons les paramètres du modèle à partir d'observations bruitées en faisant appel à la théorie statistique de deuxième espèce qui est fondée sur la transformée de Mellin. Enfin, nous faisons la comparaison entre la méthode que nous proposons et d'autres filtres classiques pour le débruitage d'images RSO. Nos résultats expérimentaux démontrent que le filtre MAP homomorphique fondé sur le modèle de Rayleigh à queue lourde est parmi les meilleurs pour enlever le bruit de chatoiement. Abstract : Synthetic aperture radar (SAR) images are inherently affected by a signal dependent noise known as speckle, which is due to the radar wave coherence. In this report, we propose a novel adaptive despeckling filter and derive a maximum a posteriori (MAP) estimator for the radar cross section (RCS). We first employ a logarithmic transformation to change the multiplicative speckle into additive noise. We model the RCS using the recently introduced heavy-tailed Rayleigh density function, which was derived based on the assumption that the real and imaginary parts of the received complex signal are best described using the alpha-stable family of distribution. We estimate model parameters from noisy observations by means of second-kind statistics theory, which relies on the Mellin transform. Finally, we compare our proposed algorithm with several classical speckle filters applied on actual SAR images. Experimental results show that the homomorphic MAP filter based on the heavy-tailed Rayleigh prior for the RCS is among the best for speckle removal.

All publications in Ariana Research Group