Early Detection of Root-Knot Nematode (Meloidogyne Incognita) Infestation in Cotton Using Hyperspectral Data

Leaf reflectance data acquired using spectrometers contain rich spectral information that enables the classification of plants into subtly different classes. The southern root-knot nematode (RKN) (Meloidogyne incognita), a soilborne roundworm, threatens cotton and other crops. In this research, we study the effect of RKN on leaf reflectance using a spectrometer under controlled-environmental conditions. We applied statistical supervised learning algorithms to classify RKN infested cotton from the control group. We study the classification accuracy by selecting the visible and near-infrared spectra (350-2500nm) and partial spectra (350–1 000nm) as a supervised classifier data set. Our study also investigates temporal misalignments between training and testing conditions. This study will provide valuable insight into the use of hyperspectral data collected from the handheld spectrometer and small unmanned aerial systems for largescale mapping of RKN infestation. We found that the effect caused by the RKN on the root system of cotton can be non-invasively diagnosed using hyperspectral data at the early growth stage.