Hyperspectral image unmixing using a multiresolution sticky HDP - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue IEEE Transactions on Signal Processing Année : 2012

Hyperspectral image unmixing using a multiresolution sticky HDP

Résumé

This paper is concerned with joint Bayesian endmember extraction and linear unmixing of hyperspectral images using a spatial prior on the abundance vectors.We propose a generative model for hyperspectral images in which the abundances are sampled from a Dirichlet distribution (DD) mixture model, whose parameters depend on a latent label process. The label process is then used to enforces a spatial prior which encourages adjacent pixels to have the same label. A Gibbs sampling framework is used to generate samples from the posterior distributions of the abundances and the parameters of the DD mixture model. The spatial prior that is used is a tree-structured sticky hierarchical Dirichlet process (SHDP) and, when used to determine the posterior endmember and abundance distributions, results in a new unmixing algorithm called spatially constrained unmixing (SCU). The directed Markov model facilitates the use of scale-recursive estimation algorithms, and is therefore more computationally efficient as compared to standard Markov random field (MRF) models. Furthermore, the proposed SCU algorithm estimates the number of regions in the image in an unsupervised fashion. The effectiveness of the proposed SCU algorithm is illustrated using synthetic and real data.
Fichier principal
Vignette du fichier
Mittelman_5345.pdf (2.9 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04240678 , version 1 (13-10-2023)

Identifiants

Citer

Roni Mittelman, Nicolas Dobigeon, Alfred Hero. Hyperspectral image unmixing using a multiresolution sticky HDP. IEEE Transactions on Signal Processing, 2012, 30 (4), pp.1656--1671. ⟨10.1109/TSP.2011.2180718⟩. ⟨hal-04240678⟩
28 Consultations
6 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More