shapeworks-users

Text archives Help


Re: [shapeworks-users] correspondence points error help


Chronological Thread 
  • From: "kmill2803 ." <kmilligan28@gmail.com>
  • To: "shapeworks-use." <shapeworks-users@sci.utah.edu>, Shireen Elhabian <shireen@sci.utah.edu>
  • Subject: Re: [shapeworks-users] correspondence points error help
  • Date: Wed, 14 Jan 2015 17:27:14 -0700
So I should use the .lpts files in the correspondence.xml file? Right now I haven't specified anything. attached is my correcpondence.xml file 

On Wed, Jan 14, 2015 at 5:19 PM, Shireen Elhabian <shireen@sci.utah.edu> wrote:

the binary shape and the distance transform are having different centers ...

I assume that these points are output of shapeworks from previous run, right? if so, use the lpts in your param file not the wpts ...

Shireen


On 01/14/2015 05:03 PM, kmill2803 . wrote:
did I send the proper pts file?

On Wed, Jan 14, 2015 at 5:02 PM, kmill2803 . <kmilligan28@gmail.com> wrote:


On Wed, Jan 14, 2015 at 4:54 PM, Shireen Elhabian <shireen@sci.utah.edu> wrote:

What are those points? would you mind sharing a sample DT with its pts just to have an idea ...

thanks
Shireen


On 01/14/2015 04:52 PM, kmill2803 . wrote:
Thanks,

I dont have paraview to be able to visualize isosurface and corresponding points but I am working on getting access to be able install it on the computer i am working on. 

I was wondering why it works for 256 correspondence points. Isnt the points provided for initialization the same for 256 as 2048? Also if this is the case how do I change the initialization poiints.

Thanks again
Kenneth Milligan

On Wed, Jan 14, 2015 at 4:37 PM, Shireen Elhabian <shireen@sci.utah.edu> wrote:

Hi Kenneth,

This usually happens if the points you are providing for initialization
doesn't lie on the isosurface of the respective distance transforms, as
a visual QC, you might want to load a sample nrrd file in paraview then
load its corresponding points and make sure that the points lie on the
isosurface ...

hope this would help

regards
Shireen

On 01/14/2015 04:31 PM, kmill2803 . wrote:
> I have got the shapeworks software to run with a set of proximal
> femurs with 256 correspondence points. I used the TorusExample files
> as a template.
>
> However when i bump up the correspondence points to 512 or 1024 or
> 2048 I get an itk::Error PowerOfTwoPointTree Point is not contained
> within tree domain.
>
> Attached is the correspondence.xml file. I only changed the number of
> points in this file.
>
>
> Thanks,
> Kenneth Milligan







<?xml version="1.0" ?>

<!-- List of files containing the set of shape surfaces. -->
<inputs>        
testDT.00.nrrd  
testDT.01.nrrd  
testDT.02.nrrd
testDT.03.nrrd
testDT.04.nrrd
testDT.05.nrrd
testDT.06.nrrd
testDT.07.nrrd
testDT.08.nrrd
testDT.09.nrrd
 </inputs>		
 
<!-- 
OPTIONALLY we could specify a set of point files to initialize the optimization like this: 
<point_files> 
testDT.00.lpts 
testDT.01.lpts 
:
.
testDT.19.lpts
</point_files>
-->

<!-- If point files are not specified, then the application will initialize particles 
     by splitting until each shape has this total number of particles. -->
<number_of_particles> 512 </number_of_particles> 

<!-- Iterations between splitting during initialization phase. -->
<iterations_per_split> 200 </iterations_per_split> 

<!-- Starting regularization for the entropy-based correspondence optimization. -->
<starting_regularization> 10.0 </starting_regularization> 

<!-- Final regularization for the entropy-based correspondence. -->
<ending_regularization> 0.1 </ending_regularization> 

<!-- Number of iterations for the entropy-based correspondence. -->
<optimization_iterations> 200 </optimization_iterations> 

<!-- Number of iterations between checkpoints (iterations at which results are saved) -->
<checkpointing_interval> 20 </checkpointing_interval> 

<!-- Prefix for the output files. -->
<output_points_prefix> test </output_points_prefix>

Archive powered by MHonArc 2.6.18.

Top of page