2

I am teaching an ad hoc class in remote sensing, where students are bringing their own laptops

I wanted to go over how to perform a supervised classification in the open source software I had chosen - gvsig, but even the simplest of tasks seem to fail.

I loaded in a Landsat scene, subsetted it to about 10% of the original area, created an ROI of a single small region of about 10 pixels - and tried to perform a supervised classification on it. It sat on creating spectral signatures for about 10 minutes, and then gave me an error along the lines of the array out of bounds or something similar.

Does anyone know what I might be doing wrong, or alternatively have another suggestion for open source software that can do supervised classifications that will work on both PC AND Mac?

QGIS will with the SCP plug-in, but the plugin appears to be for PC only.

2
  • how about using GRASS GIS its available for both.
    – Maxima
    Sep 6, 2018 at 14:17
  • You could try the Orfeo Toolbox. It runs on its own but can also be integrated into QGIS.
    – user90323
    Sep 6, 2018 at 14:38

1 Answer 1

1

an error had been detected for unsupervised classification (https://redmine.gvsig.net/redmine/issues/3096), but not for supervised classification. We have been working on vector part in the latest versions of gvSIG, but we are working on the raster part now, where there will be a new raster architecture in gvSIG 3 version.

It would be interesting to know the whole problem. Could you send the gvSIG.log file (with the information error) to the gvSIG mailing list in order to create a ticket at bugtracker and take it into account for gvSIG 3?

Here you can see how to get that file and use gvSIG mailing list: https://blog.gvsig.org/2015/06/17/what-to-do-when-we-get-an-error-in-gvsig/ Thank you very much.

Best regards, Mario Carrera

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.