Case Study on Multispectral Land Cover Classification - Open Source Geospatial Tools: Applications in Earth Observation - page 312

ogrinfo -dialect SQLite -sql "ALTER TABLE valid ADD COLUMN assessment Text
ₒ " reference_fmap_supervised.sqlite
ogrinfo -dialect SQLite -sql "UPDATE valid SET assessment='omission' WHERE
ₒ label=1 AND class=2" reference_fmap_supervised.sqlite
ogrinfo -dialect SQLite -sql "UPDATE valid SET assessment='commission'
ₒ WHERE label=2 AND class=1" reference_fmap_supervised.sqlite
ogrinfo -dialect SQLite -sql "UPDATE valid SET assessment='forest correct'
ₒ WHERE label=1 AND class=1" reference_fmap_supervised.sqlite
ogrinfo -dialect SQLite -sql "UPDATE valid SET assessment='noforest
ₒ

correct' WHERE label=2 AND class=2" reference_fmap_supervised.

ₒ

sqlite

This case study has demonstrated an operational workflow to produce a landcover
classification using freely available satellite image data. We have sought to combine
a wide variety of the tools and skills that have been presented throughout this topic.
For clarity, we have presented the individual steps of the processing chain, but it is
conceivable that they could be integrated into a shell or Python script so that it could
be run in batch mode. This is one of the distinct advantages of the command line
tools. The script can be parameterized to provide more flexibility with respect to the
input parameters (e.g. file names, bounding boxes, etc.).