[Insight-users] Lost in registration: NMI and 1+1

[Martin Kavec]

Hi Julien

On Mon, 29 Nov 2004, Julien Jomier wrote:

> Hi Martin,
>
> This type of exception means that the optimizer is going too far in the
> parameters space.
>
> One thing you can try to adjust for the 1+1 optimizer is the initial radius.
> Basically this optimizer is doing a random search of the parameters and
> start with an initial radius, meaning, for instance, that if your initial
> radius is 40 the first iteration will optimize each parameters in the range
> [-40,40] around th e initial parameter values.

I think I went through the SoftwareGuide and examples so many times that
everything look so familiar and I thought I understand it. However, I
haven't realize, until your email, what the initial radius does.

> I'm confident that this is
> your problem since you are optimizing a quaternion transform with such a
> range...

And you are right. The problem were initial radius, together with
translation scaling. I modified them so the ratio "came" closer to the
range and I got, at least, rid of the exception. Now, I can start
optimizing the whole registration loop.

> Try to decrease the initialRadius of optimizer (initialRadius =
> 1?).
>
> Also you might want to change the scaling such that initialRadius/scaling is
> within a "correct" range.
>
> Hope that helps,

It really helped a lot. I was becoming desperate. Thanks a lot.

>
> Julien

Martin