NabhyaWhat NDVI is measuring
NDVI compares reflected near-infrared light against reflected red light. Healthy leaf tissue reflects NIR strongly and absorbs red for photosynthesis, so the ratio rises with canopy vigour and falls under stress. With a calibrated sensor and a reflectance panel, that ratio is an absolute number you can compare between flights, seasons and farms.
The cost of that absolute number is a dedicated payload, a calibration step before every flight, and a post-processing pipeline afterwards.
What a near-NDVI index from RGB gives you
Nabhya infers a per-pixel vigour field from visible light alone, trained on paired samples where an RGB frame and the matching NDVI reference cover identical pixels. The output ranks the field against itself: which blocks are strongest, which are weakest, and where the worst decile sits.
- Relative vigour within a single frame or orthomosaic
- Percentile management zones for targeted scouting
- Canopy cover and a non-crop pixel mask
- Roughly two seconds per frame, over one API call
Where RGB is enough
Almost every in-season decision is a ranking decision. Which corner of the block do I walk first? Where do I pull a tissue sample? Which zone gets the variable-rate pass? None of those require an absolute index value — they require knowing how this field compares to itself, today.
Where you still need multispectral
If you need index values that are comparable across flights, across seasons or between farms — for a research protocol, an insurance product or a longitudinal yield model — a calibrated multispectral capture is the right instrument. Near-NDVI from RGB does not replace it, and we will say so.
The practical trade
A multispectral payload costs several times an RGB rig and restricts you to the drones that can carry it. The overwhelming majority of drones flying Indian fields today carry an RGB camera. If a relative vigour map answers the question, the flight you already ran can produce it.