Hyper-Graph Regularized Kernel Subspace Clustering for Band Selection of Hyperspectral Image

Meng Zeng,Chunyang Hu,Yaoming Cai,Qiong Gu,Shuijia Li,Bin Ning

doi:10.1109/access.2020.3010519

Abstract

Band selection is an effective way to deal with the problem of the Hughes phenomenon and high computation complexity in hyperspectral image (HSI) processing. Based on the hypothesis that all the pixels are sampled from the union of subspaces, many robust band selection algorithms based on subspace clustering were introduced in recent works, achieving significant performances. However, these methods focus on linear subspaces, which are not suitable for the typical nonlinear structure of HSIs. In this paper, to deal with these obstacles, a new hyper-graph regularized kernel subspace clustering (HRKSC) is presented for band selection of hyperspectral image. The proposed approach extends subspace clustering to nonlinear manifold by utilizing the kernel trick, which can better fit the nonlinear structure of HSIs. The hyper-graph regularized is introduced to consider the manifold structure reflecting geometric information and accurately describe the multivariate relationship between data points, which makes the modeling of HSIs more accurate. The results of the proposed algorithm are compared with existing band selection methods on three well-known hyperspectral data sets, showing that the HRKSC algorithm can accurately select an informative band subset and outperforming the current state-of-the-art band selection methods.

Highlights

Hyperspectral image consists of hundreds of continuous narrow spectral bands [1], [2], which acquired by remote sensors and offer the ability to accurately recognize the area of interest
In recent works, the hyper-graph learning has been used to describe the multivariate relationship between data points of HSI and achieve significant performances, such as SAHDA [27], SSHGDA [28], and EHGDA [29]. To address these issues above, we introduce a novel hyper-graph regularized kernel subspace clustering (HRKSC) algorithm for band selection of the hyperspectral image
The main contributions of this paper are summarized as follows: 1) The kernel subspace clustering is introduced for HSI band selection, which maps the feature points to a much higher dimensional space with the kernel trick to ensure the subspace clustering can be applied to the nonlinear structure of HSIs

Summary

INTRODUCTION

Hyperspectral image consists of hundreds of continuous narrow spectral bands [1], [2], which acquired by remote sensors and offer the ability to accurately recognize the area of interest. To address these issues above, we introduce a novel hyper-graph regularized kernel subspace clustering (HRKSC) algorithm for band selection of the hyperspectral image. The kernel subspace clustering method is applied to solve the nonlinear subspaces problem for band selection of HSIs. Besides, the hyper-graph regularized is added in kernel subspace clustering to make the algorithm. The main contributions of this paper are summarized as follows: 1) The kernel subspace clustering is introduced for HSI band selection, which maps the feature points to a much higher dimensional space with the kernel trick to ensure the subspace clustering can be applied to the nonlinear structure of HSIs. 2) The hyper-graph regularized is added to make full consideration of the manifold structure reflecting geometric information and accurately describe the multivariate relationship between data points, guaranteeing an informative band subset selected accurately. The HRKSC method adds the hyper-graph regularized in kernel subspace clustering for band selection of HSI.

We can multiply by

Findings

CONCLUSIONS