Weeds Detection by Ground-level Hyperspectral Data

Weeds are a severe pest in agriculture, causing extensive yield loss. Weed control of grass and broadleaf weeds is commonly performed by applying selective herbicides homogeneously all over the field. As presented in several studies, applying the herbicide only where needed has economical as well as environmental benefits. Combining remote sensing tools and techniques with the concept of precision agriculture has the potential to automatically locate and identify weeds in order to allow precise control. The objective of this work is to detect annual grasses and broadleaf weeds among cereal as well as broadleaf crops, with the aid of field spectroscopy tools. Leaf and canopy spectral relative reflectance values of three targets: crop (wheat and chickpea), grass as well as broadleaf weeds, with soil background, were obtained by Analytical Spectral Devices (ASD) FieldSpec Pro FR spectrometer in the range of 400-2400 nm. Leaf spectral classifications for botanical genera as well as category were almost perfect (99%). Canopy spectral classification for targets was accurate (total of 95%) when the field of view (FOV) contained the same target. Within the critical period for weeds control (i.e. 25-40 days after emergence), classification of 87% was achieved for canopy spectra of target in heterogeneous FOV, providing an applicative herbicide implementation. The Spectral Camera HS (Specim) with 1600 pixel per line and 849 bands in the range of 400-1000 nm was selected to continue this study. The properties of the camera should improve the ability to separate spectrally between targets by applying spatial factor. The data obtained by the camera will also be resampled to bands of the superspectral future satellite Vegetation and Environmental New micro Spacecraft (VENμS). Thereafter, VENμS spectral and spatial resolution should potentially provide precision agricultural applications including weeds control.