Advances in Parallel Techniques for Hyperspectral Image Processing

Abstract

The recent advancements in the field of computer science that have led to the development of systems with the massive computing power of 1015 floating-point operations per second. High-performance computing (HPC) systems use a large number of computing nodes to attain that performance practically. Parallel computing is a branch of HPC that focuses on reducing the execution time of any application. This work focuses on embedding parallelism in three primary Hyperspectral Image (HSI) processing techniques that are widely used in many applications like object identification, military operations, food security, monitoring natural disasters, and many more. Parallelism in these techniques is required as they work on complex mathematical operations and learning algorithms having high runtime. Some applications of these techniques work on real-time processing having energy and time constraint. It provides a comparative analysis of some recent and trending research works in the field of HSI segmentation, HSI compression, and HSI classification. Important evaluation metrics for parallel algorithms used in subsequent works have been described in detail. The purpose of this chapter is to provide a survey along with future research directions in the field of parallel image processing techniques to get the benefit of parallel computing in HSI processing.