Classification of hyperspectral images with very small training size using sparse unmixing

Vera Andrejchenko, Rob Heylen, Paul Scheunders, Wilfried Philips, Wenzhi Liao

Research output: Contribution to conferencePaper

1 Citation (Scopus)

Abstract

Hyperspectral images are high dimensional while the available number of training samples can be very low. For very small training sizes, classical supervised classification strategies may fail. In this work we propose an alternative, semi-supervised approach which is based on sparse unmixing. In this method, all training samples are gathered in a dictionary and serve as possible endmembers. Unmixing then reveals the relative contributions of the different training samples to an unlabeled sample. Since standard unmixing strategies as the Fully Constrained Linear Spectral Unmixing (FCLSU) typically assume only one endmember per class, we investigate the use of sparse unmixing. In this work, we apply the SunSAL algorithm. We show that this method outperforms SVM classification in the case of extremely small training sizes of only a few samples per class.
Original languageEnglish
Number of pages3
DOIs
Publication statusPublished - 3 Nov 2016
Event2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) - Beijing, China
Duration: 10 Jul 201615 Jul 2016

Conference

Conference2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS)
Abbreviated titleIGARSS 2016
CountryChina
CityBeijing
Period10/07/1615/07/16

Keywords

  • hyperspectral sparse unmixing
  • fully constrained unmixing
  • support vector machines ,
  • hyperspectral Imaging
  • hyperspectral image classification
  • learning (artificial intelligence)

Fingerprint Dive into the research topics of 'Classification of hyperspectral images with very small training size using sparse unmixing'. Together they form a unique fingerprint.

  • Cite this

    Andrejchenko, V., Heylen, R., Scheunders, P., Philips, W., & Liao, W. (2016). Classification of hyperspectral images with very small training size using sparse unmixing. Paper presented at 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Beijing, China. https://doi.org/10.1109/igarss.2016.7730333