Classification of wheat using near-infrared spectral data

he utility of near-infrared spectra] data in classification of wheat is investigated in this thesis. A linear discriminant is used to test the usefulness of Spectral wavebands presently employed in classification, in addition to a near-infrared channel in the 1.55 micron to 1.75 micron region of the electromagnetic spectrum, Discrimination of wheat from its confusion crops is accomplished using a variety of channel combinations including the near-infrared channel. The results are then tested to determine if the best channels for discriminating wheat from its confusion crops barley and rye can distinguish early wheat from late planted wheat in the original data set.