<br>Hi Martin,<br><br>Please do the following:<br><br>1) reduce this to a minimal case,<br>2) make an entry in the bug tracker, and<br>3) upload the source code that illustrates<br> the problem the bug entry<br><br>This behavior may well be a bug,... and the<br>
way to verify it is by doing first (1,2,3).<br><br><br> Thanks<br><br><br> Luis<br><div class="gmail_quote"><br>-------------------------------------------------------------<br>On Mon, Sep 13, 2010 at 5:26 AM, <span dir="ltr"><<a href="mailto:M.Baiker@lumc.nl">M.Baiker@lumc.nl</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;"><div link="blue" vlink="purple" lang="EN-US"><div><p class="MsoNormal">Dear all,</p><p class="MsoNormal">
</p><p class="MsoNormal">I have two corresponding 3D point sets and tried to map one domain towards the other using </p><p class="MsoNormal">“itkThinPlateSplineKernelTransform”. Since I want to allow small landmark localization errors, I set</p>
<p class="MsoNormal">the stiffness parameter > 0. However, the results do not make sense at all (I also have a Matlab</p><p class="MsoNormal">implementation where I can compare it with).</p><p class="MsoNormal">Having a look at the code I noticed, that the G matrix is computed as r i.e. the Euclidean distance</p>
<p class="MsoNormal">between landmarks. Shouldn’t that be –r instead? While for the interpolation case it shouldn’t </p><p class="MsoNormal">matter (since u(x) = A*x + B + I* sum_i=1_N(wi * ri) and if ri is –ri, the warping coefficients wi </p>
<p class="MsoNormal">become –wi as well), it does matter for the approximation case.</p><p class="MsoNormal">To give it a try, I changed r to –r and the results are already much better, however still not </p><p class="MsoNormal">
correct. Finally I tried m_Stiffness = 0 for –r and I saw that the result is different from the result</p><p class="MsoNormal">using r, which shouldn’t be the case (see above).</p><p class="MsoNormal">Could anyone comment on this? Did anyone use the TPS mapping with setting m_Stiffness > 0</p>
<p class="MsoNormal"> in the past? I tried to step through “itkKernelTransform” but couldn’t find an error so far. However</p><p class="MsoNormal">I also have to admit that I had quite some trouble getting through the code…</p>
<p class="MsoNormal"> </p><p class="MsoNormal">One further remark: What is in the current implementation is a smoothing TPS rather than a TPS</p><p class="MsoNormal">approximation since m_Stiffness has to be specified as being the same for all landmarks. For the</p>
<p class="MsoNormal">approximation case, as demonstrated in the Rohr et al. paper, it should be possible to set </p><p class="MsoNormal">m_Stiffness for all landmarks individually.</p><p class="MsoNormal"> </p><p class="MsoNormal">
Thanks in advance.</p><p class="MsoNormal"> </p><p class="MsoNormal">Greetings,</p><p class="MsoNormal">Martin</p></div></div><br>_____________________________________<br>
Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a><br>
<br>
Visit other Kitware open-source projects at<br>
<a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
<br>
Kitware offers ITK Training Courses, for more information visit:<br>
<a href="http://www.kitware.com/products/protraining.html" target="_blank">http://www.kitware.com/products/protraining.html</a><br>
<br>
Please keep messages on-topic and check the ITK FAQ at:<br>
<a href="http://www.itk.org/Wiki/ITK_FAQ" target="_blank">http://www.itk.org/Wiki/ITK_FAQ</a><br>
<br>
Follow this link to subscribe/unsubscribe:<br>
<a href="http://www.itk.org/mailman/listinfo/insight-users" target="_blank">http://www.itk.org/mailman/listinfo/insight-users</a><br>
<br></blockquote></div><br>