Multisource data classification using a hybrid semi-supervised learning scheme

Ranga Raju Vatsavai, Budhendra Badhuri, Shashi Shekhar, Thomas E. Burk

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

2 Scopus citations

Abstract

In many practical situations thematic classes can not be discriminated by spectral measurements alone. Often one needs additional features such as population density, road density, wetlands, elevation, soil types, etc. which are discrete attributes. On the other hand remote sensing image features are continuous attributes. Finding a suitable statistical model and estimation of parameters is a challenging task in multisource (e.g., discrete and continuous attributes) data classification. In this paper we present a semi-supervised learning method by assuming that the samples were generated by a mixture model, where each component could be either a continuous or discrete distribution. Overall classification accuracy of the proposed method is improved by 12% in our initial experiments.

Original languageEnglish (US)
Title of host publication2008 IEEE International Geoscience and Remote Sensing Symposium - Proceedings
PagesIII1016-III1019
Edition1
DOIs
StatePublished - 2008
Event2008 IEEE International Geoscience and Remote Sensing Symposium - Proceedings - Boston, MA, United States
Duration: Jul 6 2008Jul 11 2008

Publication series

NameInternational Geoscience and Remote Sensing Symposium (IGARSS)
Number1
Volume3

Other

Other2008 IEEE International Geoscience and Remote Sensing Symposium - Proceedings
Country/TerritoryUnited States
CityBoston, MA
Period7/6/087/11/08

Keywords

  • Expectation maximization
  • GMM
  • Multisource data
  • Semi-supervised learning

Fingerprint

Dive into the research topics of 'Multisource data classification using a hybrid semi-supervised learning scheme'. Together they form a unique fingerprint.

Cite this