Improving Hyperspectral Image Classification based on Graphs using Spatial Preprocessing

Author

Velasco-Forero, Santiago ; Manian, Vidya

Author_Institution

Lab. for Appl. Remote Sensing & Image Process., Univ. of Puerto Rico, Mayaguez

Volume

3

fYear

2008

fDate

7-11 July 2008

Abstract

Spatial smoothing over the original hyperspectral data based on wavelet and partial differential equations (PDEs) are incorporated in the classifiers using composite kernel with kNN graphs. The kernels combine spectral-spatial relationships using the smoothed and original images. Experiments with real hyperspectral scenarios are presented. Comparison with recent graph based methods show that the proposed scheme improves existing methods.

Keywords

geophysical techniques; geophysics computing; image classification; learning (artificial intelligence); partial differential equations; pattern recognition; remote sensing; AVIRIS; Semi-Supervised Learning; airbone visible/infrared imaging spectrometer; composite kernel; hyperspectral image classification; kNN graphs; original images; partial differential equations; pattern recognition algorithms; smoothed images; spatial preprocessing; spectral-spatial relationships; wavelet transform; Hyperspectral imaging; Hyperspectral sensors; Image classification; Image processing; Kernel; Partial differential equations; Pixel; Remote sensing; Semisupervised learning; Smoothing methods; Hyperspectral Images; PDE; Semi-supervised Learning; Wavelet;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. IEEE International

Conference_Location

Boston, MA

Print_ISBN

978-1-4244-2807-6

Electronic_ISBN

978-1-4244-2808-3

Type

conf

DOI

10.1109/IGARSS.2008.4779433

Filename

4779433