Dear Thayse,
You are right, for now sampling of training/validation set is tied to
training inside TrainImagesClassifier application, it is therefore
impossible to know exactly which pixels have been selected to form the
training set. Please note that we have ongoing developments on
separating the sampling step from the training step, and therefore
giving full insight on the sampled training set [1].
Right now this is only a prototype written by Paul Gely, but we intend
to integrate this is the classification workflow of Orfeo ToolBox.
Regards,
Julien
[1] https://github.com/PaulGely/AppSampling
Le 11/09/2015 16:49, Hristo Nikolov a écrit :
One more suggestion - check
here https://www.orfeo-toolbox.org//packages/OTBCookBook.pdf page 278
On Friday, 11 September 2015 15:09:15 UTC+3, Hristo Nikolov wrote:
Dear Thayse ,
Could you please describe the way you use OTB?
My experience with OTB is as part of QGIS.
I could say that the output files generated by
training procedure are with confusion matrix and
the model file. Also in the log window of the GUI
have the conf. matrix as well.
Good luck
Hristo
On Thursday, 27 August 2015 09:48:25 UTC+3, Thayse Nery wrote:
Hi all,
I am doing an object based Land Use changes classification and
I have my sample training with almost 20000 pixels. I used the
TrainImagesClassifier and selected the ratio between training
and validation samples to be 70% training and 30% validation.
As we know, this tool generates 2 files, one is the transition
matrix for the validation set and the other one is .txt file
for the training set. I would like to have a transition matrix
also for my training set. But since its generates only a text
file, I am not able to Compute a Confusion Matrix for it. I
can use this text file only for the Image Classification but I
do not know how to access the performance for the training
sample based on the F-score, precision, recall and overall
accuracy.
I would like to know if it is possible to*calculate a
confusion matrix for this training samples*. I could have used
my original shapefile to do that, but since the
TrainImagesClassifier uses some calculation such as the
maximal size of the available validation sample list per class
will be equal to the surface area of the smallest class
multiplied by the validation sample ratio, I do not know
exactly which samples were selected for the training and
validation samples.
Thanks for any help!
Cheers,
Thayse Nery
--
--
Check the OTB FAQ at
http://www.orfeo-toolbox.org/FAQ.html
You received this message because you are subscribed to the Google
Groups "otb-users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/otb-users?hl=en
---
You received this message because you are subscribed to the Google
Groups "otb-users" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected]
<mailto:[email protected]>.
For more options, visit https://groups.google.com/d/optout.
--
Julien MICHEL
CNES - DCT/SI/AP - BPI 1219
18, avenue Edouard Belin
31401 Toulouse Cedex 09 - France
Tel: +33 561 282 894 - Fax: +33 561 283 109
--
--
Check the OTB FAQ at
http://www.orfeo-toolbox.org/FAQ.html
You received this message because you are subscribed to the Google
Groups "otb-users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/otb-users?hl=en
---
You received this message because you are subscribed to the Google Groups "otb-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
