[Insight-users] How to recognize similar images with ITK?

[Dženan Zukić]

This is a wrong approach for what you are trying to accomplish.

http://en.wikipedia.org/wiki/Content-based_image_retrieval
http://scholar.google.com/scholar?q=image+similarity+retrieval&hl=en&as_sdt=0&as_vis=1&oi=scholart

This is an active research area, and the problem is inherently hard.

Regards,
Dženan

On Tue, Aug 16, 2011 at 23:21, Agustin Conde Martin <
a.conde409 at uniandes.edu.co> wrote:

> Hello everyone
>
> I'm new to the registration topic with ITK and I would like to know if it
> is possible with ITK to be able to conclude if 2 images are similar or not?
> In the attached zip file you will find a short code and the images I want to
> compare. The code is taken from the registration section of "ITK Software
> Guide" and it compares 2 input images using CenteredSimilarity2DTransform
> (because I want it to be scale, translation and rotation invariant)
> and MeanSquaresImageToImageMetric. At the end of the code, the metrics and
> the other transformations values are printed on the screen. I have compared
> square.png with square1, square2 and square3 (which are very similar), and I
> have also compared square.png with other (NOT similar) images such as
> circle, lady, triangle, etc and from the values (metric value) printed in
> both cases I'm not sure if I can conclude that an image is or is not similar
> to another one. I made some research and I found out that ITK only offers
> intensity-based recognition algorithms rather than feature-based. As you can
> see, my images are like gestures and I think the comparison won't work using
> intensity-based algorithms. So what do you recommend me to do for comparing
> my images? Is there any feature-based implementation in ITK or VTK? Or do I
> have to write from scratch an algorithm for this?  Thanks in advance.
>
> Best regards,
>
> Agustín Conde Martín
>
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.html
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20110817/c83fba50/attachment.htm>