SOFTWARE FRAMEWORK FOR HYPERSPECTRAL DATA EXPLORATION AND PROCESSING IN MATLAB

Abstract. This paper presents a user introduction and a general overview of the MATLAB software package hsicube developed by the author for simplifying the data manipulation and visualization tasks often encountered in hyperspectral analysis work, and the design principles and software development methods used by the author. The framework implements methods for slicing, masking, visualization and application of existing functions to hyperspectral data cubes without the need to use explicit indexing or reshaping, as well as enabling expressive syntax for combining these operations on the command line for highly efficient data analysis workflows. It also includes utilities for interfacing with existing file reader scripts for easy access to files using the framework. The hsicube framework is released as open source to promote the free use and peer review of the code and enable collaborative development.