NabhyaThe NDVI formula
NDVI = (NIR − Red) / (NIR + Red). Healthy leaves reflect near-infrared light strongly and absorb red light for photosynthesis, so vigorous canopy pushes the value up. Stressed, sparse or senescing vegetation pulls it down.
Values range from −1 to +1. Water and bare surfaces sit near or below zero, sparse or stressed vegetation sits low, and dense healthy canopy sits high. Exact thresholds depend on crop, growth stage and sensor.
What NDVI is a vegetation index of
A vegetation index combines spectral bands into a single number that tracks plant condition. NDVI is the best known, but others exist — some need near-infrared, others work from visible light only.
- NDVI, NDRE, SAVI — need near-infrared or red-edge bands
- VARI, ExG, GLI — computed from visible RGB bands
- Model-inferred near-NDVI — learned from paired RGB and NDVI samples
How drones capture NDVI
A classic NDVI workflow uses a multispectral camera, a reflectance panel for calibration, a planned mapping flight and photogrammetry software to stitch an orthomosaic. It produces calibrated values comparable across flights.
The alternative is inferring vigour from the RGB camera every drone already carries. Nabhya does this through one API call, returning a vigour map, management zones and canopy cover without a separate payload.
How to act on an NDVI map
- Look for patterns — irrigation lines, field edges, slopes — before single pixels
- Scout the weakest zones first and ground-truth the cause
- Compare zone shapes between flights to confirm an intervention worked
Further reading
See how NDVI from RGB works on our NDVI from RGB page, the full RGB vs multispectral comparison, and our crop health monitoring overview.