Field spectroscopy for weed detection in wheat and chickpea fields

Abstract

Weed control is commonly performed by applying selective herbicides homogeneously over entire agricultural fields. However, applying herbicide only where needed could have economical and environmental benefits. The objective of this study was to apply remote sensing to the detection of grasses and broadleaf weeds among cereal and broadleaf crops. Spectral relative reflectance values at both leaf and canopy scales were obtained by field spectroscopy for four plant categories: wheat, chickpea, grass weeds, and broadleaf weeds. Total reflectance spectra of leaf tissues for botanical genera were successfully classified by general discriminant analysis (GDA). The total canopy spectral classification by GDA for specific narrow bands was 95 ± 4.19% for wheat and 94 ± 5.13% for chickpea. The total canopy spectral classification by GDA for future Vegetation and Environmental Monitoring on a New Micro-Satellite (VENμS) bands was 77 ± 8.09% for wheat and 88 ± 6.94% for chickpea, and for the operative satellite Advanced Land Imager (ALI) bands was 78 ± 7.97% for wheat and 82 ± 8.22% for chickpea. Within the critical period for weed control, an overall classification accuracy of 87 ± 5.57% was achieved for >5% vegetation coverage in a wheat field, thereby providing potential for implementation of herbicide applications. Qualitative models based on wheat, chickpea, grass weed, and broadleaf weed spectral properties have high-quality classification and prediction potential that can be used for site-specific weed management.