Unsupervised amplitude and texture based classification of SAR images with multinomial latent model

نویسندگان

  • Koray Kayabol
  • Josiane Zerubia
چکیده

We combine both amplitude and texture statistics of the Synthetic Aperture Radar (SAR) images for classification purpose. We use Nakagami density to model the class amplitudes and a non-Gaussian Markov Random Field (MRF) texture model with t-distributed regression error to model the textures of the classes. A non-stationary Multinomial Logistic (MnL) latent class label model is used as a mixture density to obtain spatially smooth class segments. The Classification Expectation-Maximization (CEM) algorithm is performed to estimate the class parameters and to classify the pixels. We resort to Integrated Classification Likelihood (ICL) criterion to determine the number of classes in the model. We obtained some classification results of water, land and urban areas in both supervised and unsupervised cases on TerraSAR-X, as well as COSMO-SkyMed data. Key-words: High resolution SAR, TerraSAR-X, COSMO-SkyMed, classification, texture, multinomial logistic, Classification EM, Jensen-Shannon criterion ∗ Koray Kayabol carried out this work during the tenure of an ERCIM ”Alain Bensoussan” Postdoctoral Fellowship Programme. Classification non supervisée d’images RSO fondée sur l’amplitude et la texture à l’aide d’une modèle multinomial latent Résumé : Nous combinons les statistiques fondées sur l’amplitude et la texture d’images Radar à Synthése d’Ouverture (RSO) à des fins de classification. Nous utilisons la densité de Nakagami afin de modéliser les amplitudes des classes et un champ de Markov non-gaussien pour modéliser la texture, en utilisant l’erreur de régression t-distribuée afin de modéliser les textures des classes. Un modéle non-stationnaire Logistique Multinomial (LMn) d’étiquettes de structure latente est utilisé comme densité du mélange afin d’obtenir des segments de classe lissés spatialement. L’algorithme de Classification Espérance-Maximisation (CEM) est utilisé pour estimer les paramétres des classes et classer les pixels. Nous avons recours au critère ICV (Integrated Classification Vraisemblance) pour déterminer le nombre de classes dans le modèle. Nous avons obtenu des résultats de classification pour l’eau, les sols et les zones urbaines dans les cas supervisé ou non-supervisé sur des données TerraSAR-X ainsi que COSMO-SkyMed. Mots-clés : RSO haute résolution, TerraSAR-X, COSMO-SkyMed, classification, texture, modèle logistique multinomial, Classification EM, critére de Jensen-Shannon Unsupervised classification of SAR images 3

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Unsupervised Classification of SAR Images Using Hierarchical Agglomeration and EM

We implement an unsupervised classification algorithm for high resolution Synthetic Aperture Radar (SAR) images. The foundation of algorithm is based on Classification Expectation-Maximization (CEM). To get rid of two drawbacks of EM type algorithms, namely the initialization and the model order selection, we combine the CEM algorithm with the hierarchical agglomeration strategy and a model ord...

متن کامل

Unsupervised Learning of a Finite Discrete Mixture Model Based on the Multinomial Dirichlet Distribution: Application to Texture Modeling

This paper presents a new finite mixture model based on the Multinomial Dirichlet distribution (MDD). For the estimation of the parameters of this mixture we propose an unsupervised algorithm based on the Maximum Likelihood (ML) and Fisher scoring methods. This mixture is used to produce a new texture model. Experimental results concern texture images summarizing and are reported on the Vistex ...

متن کامل

Unsupervised Texture Image Segmentation Using MRFEM Framework

Texture image analysis is one of the most important working realms of image processing in medical sciences and industry. Up to present, different approaches have been proposed for segmentation of texture images. In this paper, we offered unsupervised texture image segmentation based on Markov Random Field (MRF) model. First, we used Gabor filter with different parameters’ (frequency, orientatio...

متن کامل

Unsupervised Texture Image Segmentation Using MRFEM Framework

Texture image analysis is one of the most important working realms of image processing in medical sciences and industry. Up to present, different approaches have been proposed for segmentation of texture images. In this paper, we offered unsupervised texture image segmentation based on Markov Random Field (MRF) model. First, we used Gabor filter with different parameters’ (frequency, orientatio...

متن کامل

Automatic classification of Non-alcoholic fatty liver using texture features from ultrasound images

Background: Accurate and early detection of non-alcoholic fatty liver, which is a major cause of chronic diseases is very important and is vital to prevent the complications associated with this disease. Ultrasound of the liver is the most common and widely performed method of diagnosing fatty liver. However, due to the low quality of ultrasound images, the need for an automatic and intelligent...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2018