One-dimensional characterization of natural surfaces using n-dimensional spectral information

A numerical technique capable of completely compressing and quantifying the information contained in the entire reflectance spectrum, irrespective of the number of wavebands, was evolved. The new technique, based on the concept of information theory, was used to characterize and discriminate three types of natural surfaces, e.g., soils, crops and water bodies through total information content index. This index being a number, the natural surfaces characterized by it could now be represented geometrically on a number line; thus facilitating their one-dimensional separability and simplified clustering analysis.