Classification d'images RSO polarimétriques à haute résolution spatiale sur site urbain

Doctoral thesis French OPEN
Soheili Majd , Maryam;
(2014)
  • Publisher: HAL CCSD
  • Subject: [SDU.STU]Sciences of the Universe [physics]/Earth Sciences | [INFO.INFO-IA]Computer Science [cs]/Computer Aided Engineering | High spatial resolution | Image classification | Urban area | Zone urbaine | Information mutuelle | [ INFO.INFO-IA ] Computer Science [cs]/Computer Aided Engineering | Fisher | Genetic Algorithm | Haute résolution spatiale | Polarimétrie radar | [ SDU.STU ] Sciences of the Universe [physics]/Earth Sciences | Maximum de vraisemblance | Maximum Likelihood | Mutual Information | Polarimetric radar | Classification d'images | Algorithme génétique

In this research, our aim is to assess the potential of a one single look high spatial resolution polarimetric radar image for the classification of urban areas. For that purpose, we concentrate on classes corresponding to different kinds of roofs, objects and ground su... View more
  • References (3)

    characteristic functions, ϕ , ϕ , and, vth-order log-cumulants, k B ,v ,kC ,v . ˆ , ˆ1 and ˆ B C

    indicate convolution, inverse convolution and correlation. Adapted from (Nicolas, 2006).

    .......................................................................................................................................... 77 Table 4. 2: Second kind characteristic functions of Beta and Fisher pdf obtained by inverse Mellin convolution and Mellin correlation of the GI [1, M]) and GI [µ, L] gamma distributions...................................................................................................................... 83 Table 4. 3: Second kind characteristic functions of K and Fisher pdf obtained by Mellin convolution of the IG [1, M] Inverse gamma distribution, GI [µ, L] and GI [1, M] gamma distributions ......................................................................................................... 83 Table 4. 4: Mellin statistics of uni-variate distribution (first second characteristic function and log-cumulants) for pdf in amplitude. (.) is the Gamma function, K(.) is the second kind modified Bessel function, ψ(.) is the Digamma function, and ψ(v,.) is the vth-order Polygamma function. Adapted from (Nicolas, 2010; Nicolas, 2006). ............................. 84

  • Metrics
Share - Bookmark