Significant Classes of Operations Towards Classification of Hyperspectral Images

Abstract

With the advent of advanced technological penetration towards remote sensing applications, hyperspectral image (HSI) has been a pivotal point of focus in the development of spectroscopic investigation toward further modernizing remote sensing applications. The outcome is only the identification and classification of the sensed object leading to the generation of a classification map. However, various approaches are available with a diverse focus on different problems. It is, therefore, essential to understanding the best-suited process. Therefore, this manuscript reviews different significant approaches to understanding the strength and weaknesses associated with processing HSI. The study's outcome highlights that machine learning is the most suitable approach to be considered for classification in the future direction of work, which can address various weaknesses and bridge the research gap explored in existing approaches of HSI.