[Insight-users] Model-based Segmentation (this time not html)

[Gunther Sudra]

Hi,

My goal is getting orientation in a dataset, first in 2D and later in 3D image data.  For example to approx (thresholded) head and neck with two ellipses.

 

Problem case:

For testing I modified the example in the SoftwareGuide in the following way:

- Input is now a picture with one single white rectangle on black background

- I'm using a group (which contains only one ellipse spatial object) to detect this rectangle

- metric, optimizer, and all other I didn't modify

 

Questions:

- How can I invoke the optimizer (of the example) not only to change orientation and position of the group, but also the radius of the embedded ellipse?

- How can I define a range of values for the parameter to optimize? (For example, the searched radius must be in the interval [x..y]?

- Should I use total other components to solve this problem?

Thanks in advance

Gunther

 
 

______________________________________________________________________________
Erdbeben im Iran: Zehntausende Kinder brauchen Hilfe. UNICEF hilft den
Kindern - helfen Sie mit! https://www.unicef.de/spe/spe_03.php