I am trying to classify SPOT images with the support vector machine classification in Orfeo Toolbox with QGIS. I installed Orfeo Toolbox with OSGeo4W. I have made a ROI image with 600 training samples (2 classes), and used TrainImagesClassifier (svm). However when I use Image Classification to classify the image with the svm, it gives me an error. The log only says:

the following layers were not correctly generated: Output image, Confidence map

I have tried using different images, leaving it to save to temporary file, and changing the save directory but nothing works.

1 Answer 1


the confidence map parameter is not available for the SVM algorithm. You can output it if you use the random forests (rf) for instance.

For SVM you need to disable the option. I know that the interface is not clear for now, there should be improvments of OTB QGIS integration in 2018.

Don't hesitate if you've got any questions.

Thanks again for your feedback.

  • Thank you very much. I have disabled the confidence map now, but it still does not generate an output image? Commented Dec 28, 2017 at 16:21

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.