CIBC:Project:ImageVis3D:Meetings:2008-03-06

From NCRR Biomedical Software Development, Engineering, and Dissemination Wiki

ImageVis3D Team Meeting

March 6, 2008

Link to main ImageVis3D Page

Dave, Jeroen, Rob, Allen, David

1. Still to check out
   • Paraview <a href="http://www.paraview.org"> http://www.paraview.org</a>
     • DOE ASCII product
     • parallel implementations
     • VISIT is the VACET counterpart with similar capabilities
     • ITK supported but with some of the VTK support replaced for higher efficiency
     • not clear who the target audience is
   • Metamorph

         <a href="http://www.moleculardevices.com/pages/software/metamorph.html">
      http://www.moleculardevices.com/pages/software/metamorph.html</a>
     

   • TU Wien

         <a href="http://www.cg.tuwien.ac.at/research/publications/2007/bruckner-2007-STF/">
           http://www.cg.tuwien.ac.at/research/publications/2007/bruckner-2007-STF/</a>
     

     • no software obviously available but some nice papers showing what they have done
2. Defining scope
   • goal: we still need to get more of our research into ImageVis3D
     • multi dimensional transfer function
     • ridge enhancement and occluding contours
     • sillouette lines
     • more non-photoralistic rendering
     • our goal is cover art
   • other features to look for
     • comparative vis
     • large data sets
     • time dependent data
     • multifield data
   • Charles had a list with JR; need to follow up
   • SCIRUn better than BioImage
     • add multiple elements to visualization
     • generate high res images
     • what is hard in SR is setting up X-fer function, rendering parameters
   • accepting masks to overlay the VR
   • illustrative volume rendering (SCI seminar)
3. Interface improvements:
   • to X-fer function
     • Dave has a pointer to another way of presenting plot of parameters in Xfer function
     • control of surface highlights also needs improvment
     • painting in image space linking to histogram space also great functionality
     • masking out regions of data to help tune the X-fer function based on local values
       • clipping planes are the simplest
       • ellipsoid is the next in complexity
       • polyline a la Osirix
   • create/remove wedges or other shapes from volume
   • make sets of settings in single button
   • time-dependent
     • one option is to read all the data into memory
     • other approaches exist but are complex

Link to main ImageVis3D Page