Remote sensing & change detection
Multispectral indices and epoch-to-epoch comparison over the same ground, so change is measured, not eyeballed.
- Bands
- RGB + NIR / multispectral
- Index output
- NDVI and derivatives
- Format
- .TIF
- Comparison
- Multi-epoch differencing
Multispectral indices and epoch-to-epoch comparison over the same ground.
Multispectral index rasters (.TIF)
NDVI or other vegetation/moisture indices computed per pixel, not estimated from RGB alone.
Change-detection layers
Two or more epochs differenced over the same georeferenced ground, flagging where and how much has changed.
Zonal statistics
Index values summarised per field, parcel or management zone, ready to drop into a report.
No surprises between quote and handover.
- 01
Baseline capture
First epoch flown with a multispectral sensor and the same flight geometry planned for repeat visits.
Baseline dataset - 02
Repeat capture
Subsequent epochs flown to match the baseline's geometry and ground control, so differences reflect real change, not processing artefacts.
Time-series dataset - 03
Index calculation
Vegetation or moisture indices computed per epoch from the calibrated bands.
Index rasters per epoch - 04
Differencing & reporting
Epochs are differenced and summarised into zonal statistics and change maps.
Change-detection layers · summary report
Who this is for
- Agriculture
- Forestry
- Water resources
- Renewables
Questions worth answering upfront.
- What is NDVI actually measuring?
- Normalized Difference Vegetation Index compares near-infrared and red reflectance — healthy vegetation reflects strongly in near-infrared and absorbs red, so NDVI is a proxy for vegetation vigour, not a direct measurement of any single biological property.
- How do you make sure two epochs are actually comparable?
- Same sensor calibration, same ground control, and flight lines planned to match the baseline's geometry — differencing two datasets that weren't captured consistently produces noise that looks like change.
- Can this replace ground sampling entirely?
- No — remote sensing indices correlate with ground conditions but should be calibrated or spot-checked against ground truth for anything where the finding drives a costly decision.
