No Products in the Cart
Raw multispectral imagery is only the starting point. Turning thousands of individual band captures from a MicaSense sensor into a trustworthy NDVI map or a variable-rate prescription depends on a disciplined processing chain: radiometric calibration, accurate stitching, index generation, and clean export to your farm management system. This guide walks agronomists and GIS teams through each stage, with the practices that keep your outputs scientifically defensible across flights and seasons.
Why processing discipline determines data quality
A multispectral dataset is only as good as the calibration and georeferencing behind it. Sensors such as the MicaSense RedEdge-P and the MicaSense Altum-PT capture discrete narrow bands - blue (475 nm), green (560 nm), red (668 nm), red edge (717 nm) and near-infrared (842 nm) - that must be converted from raw digital numbers into calibrated reflectance before any index is meaningful. Skip that step and your NDVI values will drift with cloud cover, sun angle and time of day, making field-to-field or week-to-week comparison impossible. For a broader overview of these platforms, see our precision agriculture and forestry guide.
Step 1: Radiometric calibration with the panel and DLS
Calibrated reflectance is the foundation of every downstream product. Both MicaSense sensors ship with a Calibration Reflectance Panel (CRP) and an integrated Downwelling Light Sensor (DLS 2). Use them together for the most reliable results.
- Calibration panel: Capture an image of the CRP on the ground immediately before takeoff and again after landing, holding the panel at the correct angle to the sun and avoiding your own shadow. The panel's known reflectance values anchor the absolute calibration.
- DLS 2: The downwelling light sensor records incident light and sun angle continuously during flight, allowing processing software to correct for changing illumination - the difference between a clear pass and a passing cloud.
- Best practice: Fly within a few hours of solar noon, re-shoot the panel between batteries on long missions, and keep the DLS mounted level and unobstructed on the airframe.
This dual approach - panel plus DLS - is what makes results from the same field repeatable week over week, which matters for any operation tracking change over a season.
Step 2: Stitching and orthomosaic generation
Once images are calibrated, photogrammetry software aligns and stitches them into a single georeferenced orthomosaic per band. Several common processing packages handle MicaSense data natively and read the per-image calibration metadata automatically.
- Plan flights with high overlap - typically around 75 percent front and side lap - so the software has enough tie points for a clean alignment.
- If your sensor and drone support RTK or PPK, use it. The RedEdge-P and Altum-PT integrate with platforms such as the DJI M300/M350 RTK series for centimetre-grade positioning, which sharpens mosaic geometry and reduces ground control requirements.
- Leverage the panchromatic band for pan-sharpening. The Altum-PT's 12.4 MP panchromatic sensor delivers pan-sharpened output down to roughly 1.25 cm per pixel at 60 m, giving you crisp boundaries on plot trials and individual plants.
The output of this stage is a stack of aligned, calibrated reflectance rasters - the raw material for index generation.
Step 3: Generating NDVI and other vegetation indices
With calibrated band rasters in hand, index maps are straightforward band-math operations. The right index depends on the agronomic question.
| Index | Bands used | Typical use |
|---|---|---|
| NDVI | NIR, Red | General canopy vigour and biomass |
| NDRE | NIR, Red Edge | Mid- to late-season nitrogen status in dense canopy |
| GNDVI | NIR, Green | Chlorophyll and nitrogen sensitivity |
| Thermal map | LWIR (Altum-PT) | Crop water stress and irrigation uniformity |
NDRE is often more informative than NDVI in a closed canopy because the red edge band (717 nm) penetrates further before saturating - a key reason the red edge sensors are valued in agronomy. The Altum-PT adds a radiometric thermal layer from its 320 x 256 FLIR Boson sensor, letting you overlay canopy temperature against vegetation indices to separate water stress from nutrient stress. For applied examples, see maximizing crop yields with the Altum-PT and environmental monitoring with the RedEdge-P Dual.
Step 4: From index maps to prescriptions
An index map is a diagnostic; a prescription is the action. To convert one to the other, classify the index raster into management zones - by quantile, by natural breaks, or against scouting ground truth - then assign an input rate to each zone. Most processing and farm-management platforms export these zones as shapefiles or as controller-ready formats such as ISOXML or RX shapefiles for variable-rate seeding, fertilizer or crop protection.
- Validate zones against field knowledge before exporting - low NDVI from a drowned-out low spot calls for a different response than low NDVI from nitrogen deficiency.
- Match raster resolution to your applicator's working width and minimum cell size; an overly granular prescription a controller cannot resolve adds no value.
- Keep calibrated source rasters archived so prescriptions can be re-derived as new ground truth arrives.
For a representative end-to-end workflow, our case study on fertilizer efficiency in Canadian farming shows how calibrated multispectral data feeds variable-rate nitrogen decisions.
Hardware and integration considerations
Processing quality starts at capture. If you are weighing sensors, our Altum-PT vs. RedEdge-P Dual comparison and our guide to integrating MicaSense sensors with DJI Enterprise drones cover the trade-offs in band count, thermal capability and airframe fit. You can browse the full lineup in our MicaSense collection, and our team can help you spec a complete capture-to-prescription pipeline. Reach out for a quote to get started.
Key Takeaways
- Always calibrate with both the reflectance panel and the DLS 2 before trusting any index.
- Fly within a few hours of solar noon and re-shoot the panel on long missions.
- High overlap and RTK/PPK positioning produce cleaner, more accurate orthomosaics.
- NDRE often outperforms NDVI for nitrogen status in dense, closed canopies.
- The Altum-PT's radiometric thermal band separates water stress from nutrient stress.
- Classify index rasters into validated management zones before building prescriptions.
- Export prescriptions as ISOXML or RX shapefiles matched to your applicator's resolution.
