The OTB team has put in a huge amount of effort behind the Monteverdi tool and it is impossible to play catch-up in the QGIS plugin with imperfect knowledge of the OTB Api for calling the functions. Not that is hard or anything , just unfamiliar. Monteverdi uses otbClassificationModule to setup and call the SVM classifier. Lots of tunable parameters are presented. This can lead very fast to window clutter.
I tried to address this from the Qgis-OTB plugin with a single dialog, which allows setting 1 raster layer for classification and 1 vector layer for training (possibly created using the easy to use Qgis digitization facilities). All that now remains to be done is to hook up all the parameters + input dataset and you get supervised classification in Qgis.
No comments:
Post a Comment