Apple powdery mildew infestation detection and mapping using high-resolution visible and multispectral aerial imaging technique

Powdery mildew (PM) in apple orchards is a critical fungal disease that considerably reduces yield, harvested fruit quality, and orchard health. Rapid detection mapping of resulting infestation at scale challenge with existing laborious manual scouting approaches. Therefore, this study explored the feasibility detecting PM an block using high-resolution visible (red-green-blue [RGB]) multispectral imaging technique. Imaging campaigns were conducted over experimental small unmanned aerial systems (UAS) integrated above optical sensors. K-means classifier trained on individual snapshots RGB imagery had mean accuracy 77%. Eight vegetation indices also showed significant differences (p < 0.001) between healthy (Mean: 0.25–0.84) infected 0.01–0.25) leaves. Modified Simple Ratio-Red (MSRR), Ratio-Blue (MSRB), Optimized Soil Adjusted Vegetation Index (OSAVI) highest contrast 0.46–0.79). Orchard block-scale orthomosaic layer, classified k-means spectral angle mapper techniques accuracies up to 73%. Overall, high resolution domain could sufficiently detect infection generate site-specific heat maps. These maps be useful growers directing prescriptive resources ( pruning labor or fungicide application) for management.