Groundtruth

GROUNDTRUTHbuyer room · privateManage buyer room

Kernel-level machine-vision imagery paired with NIR spectra and reference grading outcomes: every 1-2 kg sample poured into the benchtop analyser produces images of individual kernels classified into official grade factors (broken and damaged kernels, frost damage, sprouted kernels, diseased kernels including fusarium head blight and mildew, foreign material, admixture of other grains) alongside non-visual NIR-derived measurements (moisture, protein, test weight, oil content). The accumulated corpus is a labelled cereal-kernel defect image library tied to composite official grade outcomes across farms, seasons, geographies and crop types — with an on-combine variant in testing that adds in-field harvest-time grading.