Dear Colleagues,

Hyperspectral remote sensing is currently a fast-moving area of not only research but also industrial development, where captured hyperspectral cubes provide abundant information with great potential in many different applications. In this Special Issue, we aim to compile state-of-the-art research on how to tackle the “big data” problem of extracting the most useful information out of the hyperspectral paradigm for remote sensing applications.

This Special Issue (Volume 2) is open to any researcher working on hyperspectral remote sensing data mining and data classification. Specific topics include (but are not limited to) the following:

• Denoising and enhancement;
• Band selection and data reduction;
• Supervised and unsupervised feature extraction and feature selection;
• Compressive sensing and optimised data acquisition;
• Spatial–spectral data fusion;
• Spectral unmixing and super-resolution;
• Deep learning approaches for data mining and data classification;
• Visualisation of the data and features;
• Fast implementation of the algorithms using a GPU, etc.;
• Emerging new datasets and applications.

Dr. Jaime Zabalza
Prof. Dr. Jinchang Ren
Dr. Yijun Yan
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Remote Sensing is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2500 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• hyperspectral remote sensing
• feature extraction
• dimensionality reduction
• classification
• deep learning
• efficient computation

1. Title: CBANet: An End-to-end Cross Band 2D Attention Network for Hyperspectral Image Change Detection
    Authors: Yinhe Li, Yijun Yan, Jinchang Ren, Andrei Petrovski

2. Title: An unsupervised feature selection method based on endmember extraction for the dimension reduction of hyperspectral images

Abstract: Hyperspectral images (HSIs) include a wealth of spectral information that may be used for various purposes, including landcover classification. However, HSIs frequently suffer from "course of dimensionality," which degrades the efficiency of supervised classifiers due to insufficient training samples. A typical dimension-reduction strategy for dealing with this issue is feature selection (FS). This paper presents an unsupervised FS that uses endmember extraction methods. This method begins by extracting existing endmembers from the input Data. It creates a prototype space (PS) out of these endmembers, where a point represents each spectral band. Next, the PS identifies a set of pure spectral bands, from which the remaining bands are a linear combination. Lastly, it selects the identified bands while excluding others to reduce the HSI's high dimensionality. Experiment results validate the proposed method's superiority on three HSIs, outperforming seven state-of-the-art similar dimension reduction methods. Additionally, our technique performs better when just a small number of training samples (for example, five samples per class) are available for supervised classification.