Abstract:In this article, an effective superpixel‐guided locality preserving projection and spatial–spectral classification method is proposed for hyperspectral image classification. In this method, a superpixel‐guided locality preserving projection technique is applied to extract low‐dimensional features. Compared with the fixed‐window neighbourhood of a pixel, the spatial–adaptive superpixel is beneficial to reasonably excavate spatial–spectral information and generate more distinguishing features. In addition, in the proposed framework, a superpixel‐guided postprocessing is designed, which adopts the superpixel structure as spatial constraint to optimize the initial classification result and realizes the great improvement of classification performance. Experiments conducted on two datasets demonstrate that the proposed method has significant advantages in hyperspectral image classification performance compared with several well‐known approaches. Locality preserving projection (LPP) is a typical feature extraction method based on spectral information for hyperspectral image (HSI) classification. Recently, to improve the classification performance, the spatial information of HSI has been applied in the LPP method. However, for most of spatial–spectral‐based LPP methods, they explore the spatial–spectral information within a fixed local window, which cannot be appropriate to the irregular‐shape ground objects in HSI. To over this issue, an effective superpixel‐guided LPP and spatial–spectral classification method are proposed, in which the spatial–adaptive structure information is fully excavated for HSI classification. Specifically, superpixel segmentation is first conducted on the HSI to generate shape‐adaptive homogeneous subregions. Then, to learn more discriminative projection, the neighbourhood graph for LPP is constructed based on spatial–spectral similarity, in which pixels within the same superpixel are connected. Finally, the obtained projection feature is input a classifier to yield the initial classification result, and the edge information of ground objects captured by superpixels is utilized to optimize the initial classification result. Experiments on two real hyperspectral datasets demonstrate that the proposed superpixel‐guided and spatial–spectral classification method significantly outperforms the other well‐known techniques for HSI classification.

Convex Formulation For Hyperspectral Image Classification With Superpixels

Convex Formulation for Multiband Image Classification with Superpixel-Based Spatial Regularization.

Spectral-spatial hyperspectral image classification via SVM and superpixel segmentation

Hyperspectral Classification Based on Coupling Multiscale Super-Pixels and Spatial Spectral Features

Hyperspectral image superresolution: An edge-preserving convex formulation

Hyperspectral image classification based on superpixel merging and broad learning system

Hyperspectral Image Classification Via Joint Sparse Representation of Multi-layer Superpixles.

SUBPIXEL MAPPING OF HYPERSPECTRAL IMAGE BASED ON LINEAR SUBPIXEL FEATURE DETECTION AND OBJECT OPTIMIZATION

Image Classification via Object-Aware Holistic Superpixel Selection

A Spectral-Texture Kernel-Based Classification Method for Hyperspectral Images.

Superpixel-Based Hsi Classification Via Semi-Supervised K-Svd And Multi-Scale Sparse Representation

Spectral-spatial Hyperspectral Image Classification Using Super-Pixel-based Spatial Pyramid Representation

An improved spectral-spatial classification framework for hyperspectral remote sensing images

Deep Convolutional Networks with Superpixel Segmentation for Hyperspectral Image Classification

Superpixel‐guided locality preserving projection and spatial–spectral classification for hyperspectral image

Adaptive hypergraph superpixels

Spectral-Spatial Constraint Hyperspectral Image Classification

A 3-stage Spectral-spatial Method for Hyperspectral Image Classification

Local Low-Rank Approximation With Superpixel-Guided Locality Preserving Graph for Hyperspectral Image Classification

Superpixel Segmentation for Hyperspectral Images Based on False Colour Composition with Colour Histogram Driving

Joint learning framework of superpixel generation and fuzzy sparse subspace clustering for color image segmentation