What Is Multispectral Drone Mapping?
A standard drone camera captures red, green, and blue — what a human sees. A multispectral sensor adds dedicated narrow bands, typically green, red, red edge (~717 nm), and near-infrared (~840 nm). Healthy vegetation reflects NIR strongly and absorbs red; stressed vegetation does the opposite. By flying a grid mission and photographing the field in all bands simultaneously, a multispectral drone records the physiological state of every square meter of crop.
Mapping software then does two jobs. First, photogrammetry: stitching hundreds or thousands of overlapping frames into a single georeferenced orthomosaic, plus a digital surface model (DSM). Second, radiometry: combining the bands into vegetation indices — NDVI, NDRE, GNDVI, EVI, VARI — each highlighting a different aspect of crop condition. If you're unsure which index answers which question, start with our guide: what indices can you extract?
Which Drones Capture Multispectral Imagery?
DroneField supports the two multispectral platforms that dominate agricultural work: the DJI Mavic 3 Multispectral (Mavic 3M) and the DJI Phantom 4 Multispectral (P4M). Both capture green, red, red edge, and NIR alongside RGB, and both carry sunlight sensors that DroneField uses with per-band gain correction — no calibration panel needed — so index values stay comparable between flights flown under different light.
For RGB-only work, DroneField also processes imagery from the DJI Mavic 3E, Phantom 4 RTK, Matrice 4E / 4T (wide camera), and in beta the Matrice 5TD and M300/M350 with Zenmuse P1 — producing orthomosaics, elevation models and the RGB indices VARI, VDVI, NGRDI and ExG, which estimate vegetation vigor from visible light alone. That means you can start drone field mapping with the drone you already own and add a multispectral sensor when the agronomy justifies it.
RTK positioning (available for the P4M, Mavic 3M and P4 RTK) is recommended: it gives the best positional accuracy, so this month's stress zones line up closely with last month's map.
How DroneField Processes a Multispectral Survey
Import the flight — DroneField reads the geotags and band structure automatically. One click runs the full pipeline locally: band alignment, bundle adjustment, DSM, orthomosaic and all five multispectral vegetation indices. Under 5 minutes for a 20 ha field at 5 cm/px on a Mac mini M4 is why operators process in the truck before leaving the field.
Because every index comes from the same processing run, comparing NDVI against NDRE costs one click — critical for closed-canopy nitrogen work where NDVI saturates. From any index layer, you create an application map in one click, adjust the zones, enter rates — or let Luna AI propose them — and export a prescription: the application map documentation covers zone classification and every export format in detail.
The result plugs straight into the spraying side of the operation: boundaries and obstacles marked on the fresh orthomosaic travel with the prescription to the spray pilot's flight plan. For the broader pipeline this page is part of, see the drone mapping software overview.
Local Processing vs Cloud Multispectral Platforms
Multispectral surveys are heavy — a P4M flight is thousands of files across five bands. Cloud platforms make you upload all of it before processing starts, which in rural connectivity terms often means overnight. DroneField's local pipeline removes the upload entirely: the images never leave your computer, results arrive in minutes, and the cost doesn't scale with hectares.
That architecture matters most exactly when multispectral matters most: in-season decisions. A fungicide or nitrogen decision made from a same-day NDRE map is worth more than the same decision made three days later. See how we compare against Pix4DFields on speed, price, and workflow — or the all features page for the complete list, from project sharing to the 3D terrain map.