[Insight-users] Using MultiResolutionImageRegistrationMethod

Ranjan, Sohan (GE, Research) sohan.ranjan at ge.com
Mon Mar 20 07:23:00 EST 2006

I would appreciate if I could get clarification on the following vis-a-vis 

1. Does the code uses copies of downsampled images for computation, or 
    image iterators are used smartly for computation?

2. Is it possible to skip levels in registration? 

For example, suppose I have a huge image which is read, but I want to do registration
at the four-times resolution. Can I use an instance of MultiResolutionImageRegistrationMethod 
with, say, 4 levels, and then I do registration only at the extreme coarse level and exit. By doing 
this, I would not need to make an extra copy of the downsampled image.

Sohan Ranjan

More information about the Insight-users mailing list