Multi-type spectral spatial feature for hyperspectral image classification

In recent years, many methods have been proposed to capture intra-spectrum features for the hyperspectral image classification task. However, most of these methods ignore inter-spectra information. In consideration of this, we propose a novel 3-D Inter-Spectra Difference Feature (ISDF) descriptor, which models the relationship between adjacent spectra using the difference between a center pixel and each of its spectral-adjacent spatial-neighbor pixels. Moreover, to increase the completeness of ISDF, the Neighbor Spectral Difference Feature (NSDF) guided by local spatial information is proposed as a supplement to the insufficient description of intra-spectrum information. At last, the Multi-type Spectral Spatial Feature (MSSF) is constructed by fusing ISDF, NSDF, and a global spatial texture feature. Experimental results on three public hyperspectral image datasets demonstrate that our proposed MSSF is effective and can outperform eight representative hyperspectral image classification methods. (C) 2021 Elsevier B.V. All rights reserved.

Automated summary

In recent years, various methods have been proposed for capturing intra-spectrum features in hyperspectral image classification. However, many of these methods fail to consider inter-spectra information. To address this issue, this paper introduces a novel 3-D Inter-Spectra Difference Feature (ISDF) descriptor, which models the relationship between adjacent spectra by computing the difference between a center pixel and its spatial-neighbor pixels. Additionally, a Neighbor Spectral Difference Feature (NSDF) is proposed to supplement the incomplete description of intra-spectrum information through local spatial information. The Multi-type Spectral Spatial Feature (MSSF) is then constructed by fusing ISDF, NSDF, and a global spatial texture feature. Experimental results on three public hyperspectral image datasets demonstrate the effectiveness of MSSF, outperforming eight representative hyperspectral image classification methods.