[Insight-users] How to use LevenbergMarquardt optimizer?

[Lars Friedrich Lars]

The IterativeClosestPoint1/2/3.cxx examples of ITK perfectly show how to use LM-optimization for finding a point-based transform. As you seem to have a calibration problem you might need to implement a custum transform (derived from e.g. MatrixOffsetTransformBase) that realizes the 2D-to-3D-mapping of the US pixel coordinates.

regards,
lars

-------- Original-Nachricht --------
> Datum: Wed, 7 Jul 2010 11:32:03 +0200
> Von: Michael Kirches <michael.kirches at googlemail.com>
> An: insight-users at itk.org
> Betreff: [Insight-users] How to use LevenbergMarquardt optimizer?

> Dear list,
> 
> in my current project, I have to solve a numeric optimization problem to
> get
> parameters for coordinate transformations. Some paper I read suggests to
> use
> Levenberg-Marquardt optimization algorithm to solve the optimization
> problem.
> 
> Generally speaking, I want to transform pixel coordinates from an
> ultrasound
> image into a world coordinate system defined by some tracking system. To
> do
> so, I need to get eight parameters from a given equation.
> 
> My question is, if it is possible to use ITK implementation of
> Levenberg-Marquardt to solve numeric optimization problems.
> 
> Best Regards,
> Michael

-- 
GMX DSL: Internet-, Telefon- und Handy-Flat ab 19,99 EUR/mtl.  
Bis zu 150 EUR Startguthaben inklusive! http://portal.gmx.net/de/go/dsl