Computation in Hyperspectral Imagery (HSI) Data analysis: Role and opportunities

Abstract

Successful quantitative information extraction and the generation of useful products from hyperspectral imagery (HSI) require the use of computers. Though HSI data sets are stacks of images and may be viewed as images by analysts, harnessing the full power of HSI requires working primarily in the spectral domain. Algorithms with a broad range of sophistication and complexity are required to sift through the immense quantity of spectral signatures comprising even a single modestly sized HSI data set. The discussion in this chapter will focus on the analysis process that generally applies to all HSI data and discuss the methods, approaches, and computational issues associated with analyzing hyperspectral imagery data.