<br><font size=2 face="sans-serif">Zachary,</font>
<br>
<br><font size=2 face="sans-serif">thanks for an extremely comprehensive reply. I will work through your suggestions and let you know how I get on.</font>
<br>
<br><font size=2 face="sans-serif">Best regards,</font>
<br>
<br><font size=2 face="sans-serif">Nick</font>
<br>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td>
<td><font size=1 face="sans-serif"><b>Zachary Pincus <zpincus@stanford.edu></b></font>
<br><font size=1 face="sans-serif">Sent by: insight-users-bounces@itk.org</font>
<p><font size=1 face="sans-serif">16/06/2004 19:52</font>
<br>
<td><font size=1 face="Arial"> </font>
<br><font size=1 face="sans-serif"> To: "Nick Arini" <nick.arini@amersham.com></font>
<br><font size=1 face="sans-serif"> cc: insight-users@itk.org</font>
<br><font size=1 face="sans-serif"> Subject: Re: [Insight-users] Measuring texture of objects</font></table>
<br>
<br>
<br><font size=2 face="Courier New">Nick,<br>
<br>
If you want to perform spatial autocorrelations, probably the easiest <br>
thing to do is to perform the operation in the fourier domain. The <br>
basic idea is that the fourier transform of the autocorrelation image <br>
is the component-wise modulus of the fourier transform of the original <br>
image. So generate the fourier coefficients of an image, then mulitply <br>
each coefficient by its complex conjugate, and perform the inverse <br>
transform.<br>
<br>
I think that the ITK FFT filters might work, or you could use a <br>
different FFT package.<br>
<br>
However, there are other texture measurements that might be useful, <br>
too. I recently submitted a class <br>
(itkScalarImageToGreyLevelCoocurrenceMatrixGenerator) that computes a <br>
Grey Level Co-occurrence Matrix for an image. Hopefully it will show up <br>
in the CVS soon. GLCM's are used frequently in texture analysis (see: <br>
http://www.ucalgary.ca/~mhallbey/texture/texture_tutorial.html )<br>
<br>
I plan in the near future to extend this class to work on masked images <br>
-- currently it just works on entire requested regions. Additionally, I <br>
hope to contribute a "texture coefficient calculator" class that will <br>
calculate various texture parameters from an image, as described on the <br>
above web page, and here: <br>
http://www.cssip.uq.edu.au/meastex/www/algs/algs/algs.html )<br>
<br>
That said, I don't yet have these methods written, so if you're in a <br>
hurry, other methods might prove more useful. There's a lot of <br>
information about texture (and everything else) here: <br>
http://homepages.inf.ed.ac.uk/rbf/CVonline/ -- use the search.<br>
<br>
Zach Pincus<br>
<br>
Department of Biochemistry and Program in Biomedical Informatics<br>
Stanford University School of Medicine<br>
<br>
<br>
<br>
On Jun 16, 2004, at 8:17 AM, Nick Arini wrote:<br>
<br>
><br>
> Dear ITK users,<br>
><br>
> I would like to measure the texture (say spatial autocorrelation) of <br>
> objects or regions (irregular in shape) defined by a segmentation.<br>
><br>
> For example, say I have a 2D image which I binary threshold to give me <br>
> a collection of region masks and I want the texture of each of these <br>
> regions by measuring the spatial relationships of pixels in the <br>
> original image which form a connected set defined by the region mask <br>
> (I hope that was a clarification!).<br>
><br>
> Can someone point me in the right direction with regard to ITK <br>
> classes. I just need a kick start.<br>
><br>
> Many thanks,<br>
><br>
> Nick<br>
><br>
> _______________________________________________<br>
> Insight-users mailing list<br>
> Insight-users@itk.org<br>
> http://www.itk.org/mailman/listinfo/insight-users<br>
<br>
_______________________________________________<br>
Insight-users mailing list<br>
Insight-users@itk.org<br>
http://www.itk.org/mailman/listinfo/insight-users<br>
</font>
<br>
<br>