<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Hi Cagatay,<br>
    <br>
    First, I'd like to say that it makes total sense for you to make use
    of the new framework.<br>
    One of the main goals of the new framework is to ease the
    combination of any kind of "terms", which is the case here.<br>
    <br>
    In your case, if I understand correctly, you'd need to combine:<br>
    <ul>
      <li>Chan and Vese internal term</li>
      <li>Chan and Vese external term</li>
      <li>Propagation term where you'd provide your own feature image <br>
      </li>
      <li>and / or Advection term where you'd provide your own vector
        field</li>
      <li>Curvature term (optional for regularization) <br>
      </li>
    </ul>
    The two first terms and the curvature terms are present and
    validated within the new framework.<br>
    The propagation term is present in the framework, but the validation
    is coming along and should definitively be part of the upcoming
    release in December.<br>
    The advection term is in its way in the framework. As Matt
    mentioned, the patch is in review and still need some validation. It
    will also be part of the upcoming release in December.<br>
    <br>
    While waiting for the upcoming release (which will happen within few
    weeks), I invite you to combine Chan and Vese (internal and external
    terms) with the propagation term.<br>
    It may work fine at this point. But your feedback would definitively
    help us to improve the framework and the toolkit !!<br>
    <br>
    To combine the terms, you can have a look at the following tests:<br>
    <ul>
      <li>1 level set with Chan and Vese:<br>
      </li>
      <ul>
        <li>With a dense representation:
ITK/Modules/Segmentation/LevelSetsv4/test/itkSingleLevelSetDenseImage2DTest.cxx</li>
        <li>With a sparse representation</li>
        <ul>
          <li>Whitaker's formulation:&nbsp;
            ITK/Modules/Segmentation/LevelSetsv4/test/itkSingleLevelSetWhitakerImage2DTest.cxx</li>
          <li>Shi's formulation:
ITK/Modules/Segmentation/LevelSetsv4/test/itkSingleLevelSetShiImage2DTest.cxx</li>
          <li>Malcolm's formulation:
ITK/Modules/Segmentation/LevelSetsv4/test/itkSingleLevelSetMalcolmImage2DTest.cxx</li>
        </ul>
      </ul>
      <li>1 level set with Propagation term:</li>
      <ul>
        <li>ITK/Modules/Segmentation/LevelSetsv4/test/itkSingleLevelSetWhitakerImage2DWithPropagationTest.cxx</li>
      </ul>
    </ul>
    <br>
    In the case if you want to see the evolution of your level set
    through the iterations, you may also want to have a look at the
    tests in ITKLevelSetsv4Visualization.<br>
    <br>
    If you have questions, or remarks, please let us know.<br>
    <br>
    Thanks,<br>
    Arnaud<br>
    <br>
    <br>
    On 12/09/2011 02:06 AM, cagatay bilgin wrote:
    <blockquote
cite="mid:%20%3CCAHoiV3b+W6ip4qhxpCRaTxfHtxWQ9stN_xcvZZcT==zaWPvdZQ@mail.gmail.com"
      type="cite">Thank you very much for the clarifications&nbsp;
      <div>and the pointers Matt. I checked out the&nbsp;</div>
      <div>latest release a couple of hours ago and&nbsp;</div>
      <div>see a lot of changes as&nbsp;you pointed out.</div>
      <div><br>
      </div>
      <div>There is a GeodesicActiveContour class&nbsp;in&nbsp;</div>
      <div>the (old) LevelSet directory, which makes me</div>
      <div>believe that it wouldn't be too hard to come up&nbsp;</div>
      <div>with the corresponding term class for the new&nbsp;</div>
      <div>LevelSet framework. It is still not clear to me&nbsp;</div>
      <div>how to accomplish this though, I am completely&nbsp;</div>
      <div>puzzled&nbsp;with how it was done before.&nbsp;</div>
      <div><br>
      </div>
      <div>Cagatay</div>
      <div><br>
        <div class="gmail_quote">On Thu, Dec 8, 2011 at 6:49 PM, Matt
          McCormick <span dir="ltr">&lt;<a moz-do-not-send="true"
              href="mailto:matt.mccormick@kitware.com" target="_blank">matt.mccormick@kitware.com</a>&gt;</span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi
            Cagatay,<br>
            <div><br>
              On Thu, Dec 8, 2011 at 8:55 PM, cagatay bilgin &lt;<a
                moz-do-not-send="true" href="mailto:bilgincc@gmail.com"
                target="_blank">bilgincc@gmail.com</a>&gt; wrote:<br>
              &gt; I am trying to understand the new level set framework<br>
              &gt; and willing to use it in our project. Being able to
              add<br>
              &gt; terms to the level set function would be really
              helpful<br>
              &gt; I think. Is this the direction that ITK will be
              taking ?<br>
              <br>
            </div>
            Yes, the new level sets frame takes that direction, and it
            is a<br>
            flexible and powerful approach that allows for new avenues
            of<br>
            research.<br>
            <div><br>
              &gt; I see two directories LevelSets and LevelSetsv4, one<br>
              &gt; having term classes and the other not having them<br>
              &gt; which makes me wonder how the framework will be<br>
              &gt; working.&nbsp;It is a little hard to find directions and
              documentation<br>
              &gt; for me as the framework is still being developed.<br>
              <br>
            </div>
            Currently, the old levelsets framework is in the
            ITKLevelSets module,<br>
            and the new framework is in the ITKLevelSetsv4 module. &nbsp;They
            are not<br>
            compatible, and the term-based framework is ITKLevelSetsv4.<br>
            <br>
            Suggestions for disambiguation of the old and new framework
            are welcome.<br>
            <div><br>
              &gt;<br>
              &gt; I am trying to build an application that implements
              the<br>
              &gt; geodesic chan vese algorithm which uses the internal<br>
              &gt; and external terms from the chan vese algorithm and<br>
              &gt; would also include the gradient information from the<br>
              &gt; geodesic active contours to snap to the edges better.<br>
              <br>
            </div>
            The Chan and Vese internal and external terms have been
            implemented.<br>
            An advection term is in the works:<br>
            <br>
            &nbsp;<a moz-do-not-send="true"
              href="http://review.source.kitware.com/#change,2914"
              target="_blank">http://review.source.kitware.com/#change,2914</a><br>
            <br>
            Arnaud may have better suggstion on how to proceed with the<br>
            implementation you desire.<br>
            <br>
            You may find the MICCAI tutorial useful.<br>
            <br>
            &nbsp;<a moz-do-not-send="true"
              href="http://midas.kitware.com/community/view/65"
              target="_blank">http://midas.kitware.com/community/view/65</a><br>
            <br>
            There is a copy of the presentation, but also a VirtualBox
            virtual<br>
            machine containing the code and examples.<br>
            <br>
            The code has evolved since MICCAI. &nbsp;Most notably, the api
            has been<br>
            greatly simplified by not requiring different component
            class names to<br>
            correspond to each type of level set datatype (dense image,
            Shi,<br>
            Whitaker, Malcolm).<br>
            <br>
            Thanks,<br>
            Matt<br>
            <div><br>
              <br>
              &gt; Should I be working on the filters defined in
              LevelSets<br>
              &gt; or in LevelSetsv4? Is there a geodesic term class<br>
              &gt; defined somewhere that I can use with other terms in<br>
              &gt; LevelSetsv4 directory ?<br>
              &gt;<br>
              &gt; Any help would be&nbsp;appreciated<br>
              &gt;<br>
              &gt; Thank you,<br>
              &gt; Cagatay<br>
              &gt;<br>
            </div>
            &gt; _____________________________________<br>
            &gt; Powered by <a moz-do-not-send="true"
              href="http://www.kitware.com" target="_blank">www.kitware.com</a><br>
            &gt;<br>
            &gt; Visit other Kitware open-source projects at<br>
            &gt; <a moz-do-not-send="true"
              href="http://www.kitware.com/opensource/opensource.html"
              target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
            &gt;<br>
            &gt; Kitware offers ITK Training Courses, for more
            information visit:<br>
            &gt; <a moz-do-not-send="true"
              href="http://www.kitware.com/products/protraining.html"
              target="_blank">http://www.kitware.com/products/protraining.html</a><br>
            &gt;<br>
            &gt; Please keep messages on-topic and check the ITK FAQ at:<br>
            &gt; <a moz-do-not-send="true"
              href="http://www.itk.org/Wiki/ITK_FAQ" target="_blank">http://www.itk.org/Wiki/ITK_FAQ</a><br>
            &gt;<br>
            &gt; Follow this link to subscribe/unsubscribe:<br>
            &gt; <a moz-do-not-send="true"
              href="http://www.itk.org/mailman/listinfo/insight-users"
              target="_blank">http://www.itk.org/mailman/listinfo/insight-users</a><br>
            &gt;<br>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
  </body>
</html>