Loading [MathJax]/extensions/MathZoom.js
Region-Kernel-Based Support Vector Machines for Hyperspectral Image Classification | IEEE Journals & Magazine | IEEE Xplore

Region-Kernel-Based Support Vector Machines for Hyperspectral Image Classification


Abstract:

This paper proposes a region kernel to measure the region-to-region distance similarity for hyperspectral image (HSI) classification. The region kernel is designed to be ...Show More

Abstract:

This paper proposes a region kernel to measure the region-to-region distance similarity for hyperspectral image (HSI) classification. The region kernel is designed to be a linear combination of multiscale box kernels, which can handle the HSI regions with arbitrary shape and size. Integrating labeled pixels and labeled regions, we further propose a region-kernel-based support vector machine (RKSVM) classification framework. In RKSVM, three different composite kernels are constructed to describe the joint spatial-spectral similarity. Particularly, we design a desirable stack composite kernel that consists of the point-based kernel, the region-based kernel, and the cross point-to-region kernel. The effectiveness of the proposed RKSVM is validated on three benchmark hyperspectral data sets. Experimental results show the superiority of our region kernel method over the classical point kernel methods.
Published in: IEEE Transactions on Geoscience and Remote Sensing ( Volume: 53, Issue: 9, September 2015)
Page(s): 4810 - 4824
Date of Publication: 06 April 2015

ISSN Information:

Funding Agency:


I. Introduction

HYPERSPECTRAL data contain a set of images with the same geographic scene. These images correspond to different spectral bands of electromagnetic radiation. Fixed a band, the hyperspectral data reduce to a single image containing the scene structure information of different materials. Fixed an image coordinate, it obtains a spectral curve vector, which is called a pixel [here, “pixel” refers to “sample” in a hyperspectral image (HSI)]. Different materials have different absorptions or reflections at a certain spectral band. Thus, it can identify and classify the materials based on their spectral curves. Traditional classifiers, such as the Bayesian classifier, the -nearest neighbor classifier, and neural networks, use the spectral signatures in the HSI classification.

References

References is not available for this document.