A Machine learning approach for Object Parameter Estimation and Discrimination Using Hyperspectral Data

This textbook course answers all these questions and more! The textbook book course presents not only the basic theoretical principles of spectroscopy, spectral matching, labeling and discrimination, but also a new novel method, the k-step methodology, that automates the entire process. Both for object parameter estimation and spectral discrimination! A machine learning approach is incorporated to achieve the full automation; the simple genetic algorithm. For all these topics, extensive measurements were collected and experiments were performed in order to prove the concept. Spectral measurements of different varieties of plants (vetch and lentil) were used to showcase the subtle spectral discrimination concept. Regarding the parameter estimation, soil spectral measurements were taken along with chemical analysis to quantify the soil organic matter.