28595 articles – 22087 references  [version française]

hal-00612491, version 2

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

Koray Kayabol () 1, Josiane Zerubia () 1

N° RR-7700 version 2

  • 1:  ARIANA (INRIA Sophia Antipolis / Laboratoire I3S)

  • INRIA – Université Nice Sophia Antipolis [UNS] – CNRS : UMR7271 France
  • Available versions :  v1 (2011-07-31) v2 (2012-05-02)

    • Bibliographic reference

    • Type of document: Research reports
    • Subject:
      Engineering Sciences/Signal and Image processing
      Computer Science/Signal and Image Processing
    • Title: Unsupervised amplitude and texture based classification of SAR images with multinomial latent model
    • Abstract: 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.
    • Fulltext language: English
    • Production date: 2012-05-02
    • Keyword(s): High resolution SAR – TerraSAR-X – COSMO-SkyMed – classification – texture – multinomial logistic – unsupervised learning – Classification EM – Jensen-Shannon criterion
    • Classification: I.: Computing Methodologies/I.4: IMAGE PROCESSING AND COMPUTER VISION/I.4.6: Segmentation; I.: Computing Methodologies/I.5: PATTERN RECOGNITION/I.5.3: Clustering; I.2.10.8: Texture
    • Internal note: RR-7700 version 2

    Attached file list to this document: 

    PDF
    RR-7700v2.pdf(2.9 MB)
     
    • hal-00612491, version 2
    • http://hal.archives-ouvertes.fr/hal-00612491
    • oai:hal.archives-ouvertes.fr:hal-00612491
    • From: Koray Kayabol <>
    • Submitted on: Wednesday, 2 May 2012 16:15:23
    • Updated on: Wednesday, 2 May 2012 16:19:54