Case study: Mapping agricultural uses

by combining satellite and aerial imagery analysis

The Challenge

  • RMH was tasked by EKNA Services, Inc. to identify agricultural uses for an assessment of ditch irrigation systems throughout the state of Hawaii.

  • Several crops are spectrally independent, but positive identification is unclear without considerable ground assessment.

 

RMH Solution 

  • We ran semi-automated, object based classifications of multispectral satellite imagery from WorldView-2, acquired in 2011.

  • We conducted aerial surveys in 2014 to produce natural color ortho-mosaics at 4cm GSD, then used visual interpretation to identify crops and land use.

 

Results

  • Final maps of crop types derived from satellite imagery were validated using aerial imagery as virtual groundtruthing

Corn
Corn

Aerial imagery at 4cm GSD

press to zoom
Corn classification
Corn classification

Object-based classification of satellite imagery

press to zoom
Miscellaneous produce
Miscellaneous produce

Aerial imagery at 4cm GSD

press to zoom
Miscellaneous produce classification
Miscellaneous produce classification

Object-based classification of satellite imagery

press to zoom
Fruit & nut trees
Fruit & nut trees

Aerial imagery at 4cm GSD

press to zoom
Fruit & nut trees classification
Fruit & nut trees classification

Object-based classification of satellite imagery

press to zoom