%0 Conference Proceedings %T A non-Gaussian statistical modeling of SIFT and DT-CWT for radar target recognition %+ Lab-STICC_ENSTAB_MOM_PIM %+ Pôle STIC_REMS %+ Laboratoire de Recherche Informatique et Télécommunications (LRIT) %+ Laboratoire de Recherche en Informatique et Télécommunications [Rabat] (GSCM-LRIT) %A Karine, Ayoub %A Toumi, Abdelmalek %A Khenchaf, Ali %A Hassouni, Mohammed El %< avec comité de lecture %B AICCSA 2016 %C Agadir, Tunisia %8 2016-11-29 %D 2016 %K Automatic target recognition %K SAR image %K SIFT %K DTCWT %K non-Gaussian distributions %K SVM. %Z Engineering Sciences [physics]/Signal and Image processing %Z Engineering Sciences [physics]/ElectromagnetismConference papers %X The work presented in this paper is part of the filedof automatic recognition of radar targets. Thus, for assistancein target recognition, we propose a new approach to extractefficient feature from synthetic aperture radar (SAR) images.The proposed approach deals with a combination of two featuredescriptors obtained from two methods. In the first method,we perform the dual-tree complex wavelet transform (DT-CWT)on SAR image, and then, the complex subbands magnitudesare modeled by a non-Gaussian statistical model. In the secondmethod, we use the scale invariant feature transform (SIFT). Dueto the fact that SIFT descriptor is limited to a huge dimension,we propose to model its statistical behavior using a non-Gaussianstatistical model in order to overcome this limit. The combinationof the resulting Weibull or Gamma statistical parameters for theboth DT-CWT and SIFT methods are selected as a feature vector.To validate our appraoch, the classification results are providedusing Polynomial kernel based support vector machines (SVM)classifier. The experimental results using SAR images databaseshow the benefits of the proposed approach to extract featuredescriptor. %G English %L hal-01406126 %U https://hal.science/hal-01406126 %~ UNIV-BREST %~ INSTITUT-TELECOM %~ ENSTA-BRETAGNE %~ CNRS %~ UNIV-UBS %~ ENSTA-BRETAGNE-STIC %~ ENIB %~ LAB-STICC_ENIB %~ LAB-STICC %~ INSTITUTS-TELECOM