Contextual unmixing of geospatial data based on Markov random fields and conditional random fields

Ryuei Nishii, Tomohiko Ozaki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

In supervised and unsupervised image classification, it is known that contextual classification methods based on Markov random fields (MRFs) improve the performance of non-contextual classifiers. In this paper, we consider the unsupervised unmixing problem based on MRFs. The exact solutions maximizing local conditional densities are derived, and they show excellent performance for unximing of data sets. Furthermore a new stochastic model based on conditional random fields is proposed for unmixing of hyperspectral data. The approximation formula of its normalizing factor is also derived.

Original languageEnglish
Title of host publicationWHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing
Subtitle of host publicationEvolution in Remote Sensing
DOIs
Publication statusPublished - Dec 21 2009
EventWHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing - Grenoble, France
Duration: Aug 26 2009Aug 28 2009

Publication series

NameWHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing

Other

OtherWHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing
CountryFrance
CityGrenoble
Period8/26/098/28/09

Fingerprint

Image classification
Stochastic models
Classifiers

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition
  • Signal Processing

Cite this

Nishii, R., & Ozaki, T. (2009). Contextual unmixing of geospatial data based on Markov random fields and conditional random fields. In WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing [5289004] (WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing). https://doi.org/10.1109/WHISPERS.2009.5289004

Contextual unmixing of geospatial data based on Markov random fields and conditional random fields. / Nishii, Ryuei; Ozaki, Tomohiko.

WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing. 2009. 5289004 (WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nishii, R & Ozaki, T 2009, Contextual unmixing of geospatial data based on Markov random fields and conditional random fields. in WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing., 5289004, WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing, WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing, Grenoble, France, 8/26/09. https://doi.org/10.1109/WHISPERS.2009.5289004
Nishii R, Ozaki T. Contextual unmixing of geospatial data based on Markov random fields and conditional random fields. In WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing. 2009. 5289004. (WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing). https://doi.org/10.1109/WHISPERS.2009.5289004
Nishii, Ryuei ; Ozaki, Tomohiko. / Contextual unmixing of geospatial data based on Markov random fields and conditional random fields. WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing. 2009. (WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing).
@inproceedings{b6205f11d1b84711a8d6d8f6c938c62b,
title = "Contextual unmixing of geospatial data based on Markov random fields and conditional random fields",
abstract = "In supervised and unsupervised image classification, it is known that contextual classification methods based on Markov random fields (MRFs) improve the performance of non-contextual classifiers. In this paper, we consider the unsupervised unmixing problem based on MRFs. The exact solutions maximizing local conditional densities are derived, and they show excellent performance for unximing of data sets. Furthermore a new stochastic model based on conditional random fields is proposed for unmixing of hyperspectral data. The approximation formula of its normalizing factor is also derived.",
author = "Ryuei Nishii and Tomohiko Ozaki",
year = "2009",
month = "12",
day = "21",
doi = "10.1109/WHISPERS.2009.5289004",
language = "English",
isbn = "9781424446872",
series = "WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing",
booktitle = "WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing",

}

TY - GEN

T1 - Contextual unmixing of geospatial data based on Markov random fields and conditional random fields

AU - Nishii, Ryuei

AU - Ozaki, Tomohiko

PY - 2009/12/21

Y1 - 2009/12/21

N2 - In supervised and unsupervised image classification, it is known that contextual classification methods based on Markov random fields (MRFs) improve the performance of non-contextual classifiers. In this paper, we consider the unsupervised unmixing problem based on MRFs. The exact solutions maximizing local conditional densities are derived, and they show excellent performance for unximing of data sets. Furthermore a new stochastic model based on conditional random fields is proposed for unmixing of hyperspectral data. The approximation formula of its normalizing factor is also derived.

AB - In supervised and unsupervised image classification, it is known that contextual classification methods based on Markov random fields (MRFs) improve the performance of non-contextual classifiers. In this paper, we consider the unsupervised unmixing problem based on MRFs. The exact solutions maximizing local conditional densities are derived, and they show excellent performance for unximing of data sets. Furthermore a new stochastic model based on conditional random fields is proposed for unmixing of hyperspectral data. The approximation formula of its normalizing factor is also derived.

UR - http://www.scopus.com/inward/record.url?scp=72049083494&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=72049083494&partnerID=8YFLogxK

U2 - 10.1109/WHISPERS.2009.5289004

DO - 10.1109/WHISPERS.2009.5289004

M3 - Conference contribution

AN - SCOPUS:72049083494

SN - 9781424446872

T3 - WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing

BT - WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing

ER -