Abstract
Hyperspectral image (HSI) classification is a widely used application to provide important information of land covers. Each pixel of an HSI has hundreds of spectral bands, which are often considered as features. However, some features are highly correlated and nonlinear. To address these problems, we propose a new discrimination analysis framework for HSI classification based on the Two-stage Subspace Projection (TwoSP) in this paper. First, the proposed framework projects the original feature data into a higher-dimensional feature subspace by exploiting the kernel principal component analysis (KPCA). Then, a novel discrimination-information based locality preserving projection (DLPP) method is applied to the preceding KPCA feature data. Finally, an optimal low-dimensional feature space is constructed for the subsequent HSI classification. The main contributions of the proposed TwoSP method are twofold: (1) the discrimination information is utilized to minimize the within-class distance in a small neighborhood, and (2) the subspace found by TwoSP separates the samples more than they would be if DLPP was directly applied to the original HSI data. Experimental results on two real-world HSI datasets demonstrate the effectiveness of the proposed TwoSP method in terms of classification accuracy.
Highlights
Due to rapid development, hyperspectral images (HSIs) play a very significant role in various hyperspectral remote sensing applications, e.g., military [1], astronomy [2], and classification [3,4,5]
In order to practically address the dimensionality reduction problem, we propose a new discrimination-information based locality preserving projection (DLPP) method by computing the kernel distances in a k-nearest neighborhood for the foregoing kernel principal component analysis (KPCA) projected data when the training samples are belong to the same class
We conduct the experiments on two real-world HSI datasets, i.e., Indian Pines and Kennedy Space Center (KSC) datasets [40], to demonstrate the effectiveness of the proposed Two-stage Subspace Projection (TwoSP) method compared with the existing dimensionality reduction algorithms
Summary
Hyperspectral images (HSIs) play a very significant role in various hyperspectral remote sensing applications, e.g., military [1], astronomy [2], and classification [3,4,5]. The focus of feature extraction is to project the original high-dimensional feature data into an optimal low-dimensional subspace, which is able to construct valuable features in the projective transformation. Several popular feature extraction methods are principal component analysis (PCA) [19], independent component analysis (ICA) [20], linear discriminant analysis (LDA) [21], and locality preserving projection (LPP) [22]. PCA is a popular global dimensionality reduction method, while LPP is an effective local dimensionality reduction method Both global and local structures are important for projecting the original high-dimensional data into a low-dimensional subspace while preserving the valuable information. It is difficult to obtain better classification performance because the discrimination information is underutilized
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call