Manifold Learning-Based Semisupervised Neural Network for Hyperspectral Image Classification

Feature extraction (FE), an important preprocessing step in hyperspectral image (HSI) classification, has received growing attention in the remote sensing community. In recent years, the FE ability of deep learning (DL) methods has been widely recognized. However, most DL models focus on training networks with strong nonlinear mapping ability. They fail to explore the intrinsic manifold structure in HSI, and their performance depends on large size of the labeled training set. To address the above problems, a novel FE approach, termed manifold learning-based semisupervised neural network (MSSNet), was proposed in this article. By introducing the graph embedding (GE) framework, MSSNet develops a semisupervised graph model to explore the manifold structure in HSI with both labeled and unlabeled data. On the basis of this graph model, MSSNet constructs a combined loss function to take into account the metric of difference values and the exploration of manifold margins; thus, it reduces the difference between the predictive value and the actual value to enhance the separability of the features extracted by the network. Experiments conducted on real-world HSI datasets demonstrate that the performance of the proposed MSSNet outperforms some related state-of-the-art FE approaches.

I. Introduction

As a promising earth observation technology, hyperspectral remote sensing has developed rapidly in recent years [1]–[3]. Hyperspectral images (HSIs) are captured by passive spectrometers, which measure the solar radiation reflected by the observation areas and generates data cubes composed of hundreds of narrow and continuous spectral wavelengths [4], [5]. Due to the fine-spectral resolution, HSI has a wide variety of applications in many fields, such as urban planning, mineral exploration, and precision agriculture [6], [7]. However, due to the strong correlations among different spectral bands, HSI contains much redundant information that takes a lot of computing resources and weakens the performance of classifiers [8], [9]. Therefore, it is an urgent issue to achieve the fine classification of land covers by reducing the dimension of spectral features and extracting the intrinsic information in HSI [10]–[12].