Hyperspectral Image Classification—Traditional to Deep Models: A Survey for Future Prospects

Abstract

Hyperspectral Imaging (HSI) has been extensively utilized in many real-life applications because it benefits from the detailed spectral information contained in each pixel. Notably, the complex characteristics i.e., the nonlinear relation among the captured spectral information and the corresponding object of HSI data make accurate classification challenging for traditional methods. In the last few years, Deep Learning (DL) has been substantiated as a powerful feature extractor that effectively addresses the nonlinear problems that appeared in a number of computer vision tasks. This prompts the deployment of DL for HSI classification (HSIC) which revealed good performance. This survey enlists a systematic overview of DL for HSIC and compared state-of-the-art strategies of the said topic. Primarily, we will encapsulate the main challenges of traditional machine learning for HSIC and then we will acquaint the superiority of DL to address these problems. This survey breakdown the stateof-the-art DL frameworks into spectral-features, spatial-features, and together spatial-spectral features to systematically analyze the achievements (future research directions as well) of these frameworks for HSIC. Moreover, we will consider the fact that

A Triple-Path Spectral–Spatial Network With Interleave-Attention for Hyperspectral Image Classification

“…Recently, GCN and transformer have been introduced to HSI classification field which can further cooperate with CNN to model the relations between samples and process the sequential spectral data. Still, they are computationally expensive and have massive parameters [47]. So, in this paper, we aim to construct a triple path network which can achieve spectral-spatial feature utilization, meanwhile it owns fewer parameters and lower computational complexity.…”

Hyperspectral Unmixing Using Transformer Network

“…In particular, hyperspectral imaging (HSI) has attracted much attention in recent years. Its tasks include but are not limited to land used and land cover classification [1]- [4], forest applications [5], [6] and target detection [7] etc. In hyperspectral remote sensing, each spectral pixel might cover several pure materials on the ground due to its limited spatial resolution.…”

Hyperspectral Image Labeling and Classification Using an Ensemble Semi-Supervised Machine Learning Approach

“…A graph-based semisupervised learning or ensemble label propagation method using spectral-spatial similarity measurements from a graph representation is proposed in [8]. Recently, Deep Learning (DL) methods are being developed and used for HSI classification [9]. Autoencoders have been used for hyperspectral unmixing and extended to the classification of HSI [10][11][12].…”