[ITK-users] RegistrationParameterScalesFromPhysicalShif and itkVersorRigid3DTransform

[Bradley Lowekamp]

Hello,

> 
> *Optimizer stop condition: OnePlusOneEvolutionaryOptimizerv4: Fnorm
> (0.000777958) is less than Epsilon (1 at iteration #0*

This message may mean that the scales are not correctly set, and the the initial search area failed to yield a significant change. Things that can impact the scales include the metric used and the spacing of your image. This can some times be a difficult thing to track down. Would the spacing of you image happen to be very large or very small?

Also keep in mind that the Versor transforms are not Euclidean spaces and a best suited for use with the regular step gradient descent optimizers as only that optimizer takes into consideration the correct normalization of the versor space.

In short I think you have picked three challenging components that may just now work well together.

Brad

> 
> I found almost no documentation of ITK version 4, why ?In Manuals not found
> anything. 
> 
> This is the part of code to configure the estimation scale (based on
> examples found in the DoxyGen):
> 
>      typedef itk::RegistrationParameterScalesFromPhysicalShift<MetricType>
> ScalesEstimatorType;
>      ScalesEstimatorType::Pointer scalesEstimator =
> ScalesEstimatorType::New();
>      scalesEstimator->SetMetric( metric );
>      scalesEstimator->SetTransformForward( true );
>      scalesEstimator->SetSmallParameterVariation( 1.0 );
> 
>      optimizer->SetScalesEstimator( scalesEstimator );
>      optimizer->SetDoEstimateScales(true);  
> 
> Hopefully someone can help me with something, thanks.
> 
> 
> 
> 
> --
> View this message in context: http://itk-insight-users.2283740.n2.nabble.com/RegistrationParameterScalesFromPhysicalShif-and-itkVersorRigid3DTransform-tp7586348.html
> Sent from the ITK Insight Users mailing list archive at Nabble.com.
> _____________________________________
> Powered by www.kitware.com
> 
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
> 
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.php
> 
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
> 
> Follow this link to subscribe/unsubscribe:
> http://public.kitware.com/mailman/listinfo/insight-users