<html>
<body>
I have had very good results using the AffineTransform with both
ViolaWells and Mattes MI. I did have to scale the translate parameters a
lot smaller than I used for the Rigid Quaternion.<br><br>
Bill<br><br>
At 02:01 PM 5/24/2004, raghu venkatram wrote:<br>
<blockquote type=cite class=cite cite>Hello Luis,<br>
<br>
I have been experimenting with the Affine transform along with
viola-wells imagetoimagemutualinformation and mattesmutual information
metrics.<br>
<br>
The affine transform which consists of 9(rotation, scaling, shearing) + 3
translation is a 12 parameter transform. Adopting rigid constraints and 3
parameters 3(rotation) + 3(translation) to define the transformation,
<br>
i am not sure of how to reduce the 12 parameters to 6, Ashburner and
Friston (SPM people) suggest a method in ther Spatial registration paper.
<br>
I am right now, weighting the rotation optimizer scales way higher than
the translation optimizer scales. <br>
The alternative would be to first determine the 3 translation paramters
and then the rotation parameters.<br>
<br>
I would be glad if you had any suggestions in defining and optimizing a 6
parameter transform under the rigit constraints<br>
<br>
<br>
Warm regards,<br>
Raghu <br><br>
<hr>
<a href="http://g.msn.com/8HMAENUS/2752??PS=47575">Express yourself with
the new version of MSN Messenger! Download today - it's FREE!</a> <br>
_______________________________________________<br>
Insight-users mailing list<br>
Insight-users@itk.org<br>
<a href="http://www.itk.org/mailman/listinfo/insight-users" eudora="autourl">http://www.itk.org/mailman/listinfo/insight-users</a></blockquote></body>
</html>