This is my output of trainverctorclassifier : model.rf

%YAML:1.0
my_random_trees: !!opencv-ml-random-trees
   nclasses: 1
   nsamples: 41521
   nactive_vars: 2
   oob_error: 0.
   ntrees: 1
   is_classifier: 1
   var_all: 5
   var_count: 5
   ord_var_count: 5
   cat_var_count: 0
   training_params:
      use_surrogates: 0
      max_categories: 10
      max_depth: 5
      min_sample_count: 10
      cross_validation_folds: 0
      priors: !!opencv-matrix
         rows: 1
         cols: 1
         dt: d
         data: [ 1. ]
   var_type: [ 0, 0, 0, 0, 0 ]
   cat_count: !!opencv-matrix
      rows: 1
      cols: 1
      dt: i
      data: [ 1 ]
   cat_map: !!opencv-matrix
      rows: 1
      cols: 1
      dt: i
      data: [ 0 ]
   trees:
      -
         best_tree_idx: -1
         nodes:
            -
               depth: 0
               sample_count: 41521
               value: 0.
               norm_class_idx: 0
               Tn: 0
               complexity: 0
               alpha: 0.
               node_risk: 0.
               tree_risk: 0.
               tree_error: 0.


Why does it detect only one class? my training vector has 8 classes.

On Friday, February 10, 2017 at 2:40:46 PM UTC-3:30, [email protected] 
wrote:
>
> Hi all,
>
> I have been working on OTB for a while, and now, I want to do some 
> classification. I have followed the steps mentioned on the page below, but 
> the result that I get is a large black image that has no useful data and 
> apparently somewhere in the steps something went wrong, I assume.
>
> https://www.orfeo-toolbox.org/CookBook/recipes/pbclassif.html
>
> In the steps explained above, I should mention that no where in the 
> workflow I was told to import my shape file from the segmentation which I 
> got it before doing classification. I only worked on the input *.tif file 
> and the training vector (shape file) which has some classes defined inside 
> it(=ground truth). This is one of the things that I think I am missing. And 
> if we shouldn't use the output(shape file) of our segmentation, what was 
> the use of the segmentation being done earlier.
>
> To recap:
> I have the following files ready to start my classification:
> 1) an input image( a 5 band GeoTiff file)
> 2) a shape file (which I got from the segmentation application of OTB) = 
> this file is not used the above steps.
> 3) a training shape file (which is telling me about the existing classes 
> on the main input image)
>
> I was wondering if you guys could help me. Is there any other materials 
> available that explains the workflow or a video tutorial for instance?
>
> I am also working on a Windows 64 bit.
>

-- 
-- 
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.

Reply via email to