The effect of training data on hyperspectral classification algorithms

Author

Ozdemir, Okan Bilge ; Cetin, Y.Y.

Author_Institution

Enformatik Enstitusu, Orta Dogu Teknik Univ., Ankara, Turkey

fYear

2013

fDate

24-26 April 2013

Firstpage

1

Lastpage

4

Abstract

In this study, the performance of different hyperspectral classification algorithms with the same training set is investigated. In addition, the effect of the dimension and sampling strategy for the training set selection is demonstrated. Support Vector Machines (SVM), K- Nearest Neighbor (K-NN) and Maximum Likelihood (ML) methods are used. The contribution of using spatial information with spectral information is observed. Meanshift segmentation and window weighting methods are used for spatial information. High resolution Pavia University hyperspectral data and Indian Pines data are used in this study.

Keywords

geophysical image processing; hyperspectral imaging; maximum likelihood estimation; support vector machines; Indian Pines data; K-NN method; K-nearest neighbor method; ML method; SVM; high resolution Pavia University hyperspectral data; hyperspectral classification algorithms; maximum likelihood method; spatial information; support vector machines; training data effect; training set selection; Classification algorithms; Hyperspectral imaging; Kernel; Support vector machines; Training; Hyperspectral Classification; K-Nearest Neighbor; Maximum Likelihood; Support Vector Machines;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing and Communications Applications Conference (SIU), 2013 21st

Conference_Location

Haspolat

Print_ISBN

978-1-4673-5562-9

Electronic_ISBN

978-1-4673-5561-2

Type

conf

DOI

10.1109/SIU.2013.6531323

Filename

6531323