Multispectral mapping across thousands of frames
Flying and processing multispectral surveys of more than 3,000 images into calibrated reflectance maps.
A multispectral survey is only as good as its processing. Past a few thousand frames, the usual photogrammetry settings either run out of memory or take days, and small calibration errors spread across the whole map.
In the field
We fly a DJI Mavic 3 Multispectral with RTK positioning, capturing green, red, red-edge and near-infrared bands alongside the colour camera. Flights are planned around the sun angle, and irradiance is logged for every frame so that images taken an hour apart can be compared.
In the office
Datasets of more than 3,000 images are processed in Pix4D with a workflow we tuned for this aircraft: how the project is split, how tie points are matched, and where calibration is applied. The outputs are orthomosaics, surface models and vegetation index maps that open directly in GIS software.
What clients use it for
- Crop condition and stress mapping
- Wetland and vegetation extent
- Baselines before construction or restoration, so change can be measured later