Posted by
Daniel James White on
Apr 22, 2011; 2:50pm
URL: http://imagej.273.s1.nabble.com/3D-using-Stack-Focusser-tp3684866p3684867.html
Hi Joel,
On Apr 22, 2011, at 6:03 AM, IMAGEJ automatic digest system wrote:
> Date: Thu, 21 Apr 2011 15:23:20 -0400
> From: "JOEL B. SHEFFIELD" <
[hidden email]>
> Subject: 3D using Stack Focusser?
>
> Colleagues,
>
> Here's an interesting challenge. As I understand it, both the 3D Viewer or
> the 3D rotation utility built into ImageJ work on some form of an algorithm
> in which the relative positions of the slices in a stack are shifted
> relative to each other, and then a Maximum projection is generated,
3D viewer is not using a max projection i think.... its more like an average or sum projection
through the volume in a given direction/line... but maybe thats not so important here...
> creating
> a new view through the shifted stack. We have been looking at the DIC
> stacks that are generated either with confocal or manual DIC systems, which
> also have optical slices. However, when we try a standard Max projection,
> the resultant is a mess, since the criteria for inclusion are not intensity,
> but contrast.
Yes, this is expected to be the case...
the intensity in a DIC image is a complicated function
of the local refractive index gradient, and the direction of that gradient
with respect to the angle the polarizers and prisms are set at.
I have seen no good way to convert this kind of image into
something that will give an intuitively sensible looking 3D volume rendered result...
but it would be neat.
> On the other hand, we have been able to use the Stack
> Focuser plugin to create "flattened" images of these stacks so that all
> components remain in focus. I am wondering if it is possible to use an
> image shift algorithm similar to the one for transparent fluorescent objects
> to generate a rotating, or at least a simple stereo view of such DIC
> samples.
>
the problem here is that, (unlike fluorescence) the same object looks very different in DIC depending on its orientation,
so when you move to a new view point in 3d space... well, things just dont add up or make sense...
Perhaps this might help:
Interesting features in DIC are usually where there is a steep intensity grdient in the image.
So, you can take a DIC image z stack, compute its gradient image per slice,
and use that for 3D rendering.
I dont have a DIC z stack handy... but i imagine the above trick might give something useful? Maybe?
?
D
Dr. Daniel James White BSc. (Hons.) PhD
Senior Microscopist / Image Processing and Analysis
Light Microscopy Facility
Max Planck Institute of Molecular Cell Biology and Genetics
Pfotenhauerstrasse 108
01307 DRESDEN
Germany
+49 (0)15114966933 (German Mobile)
+49 (0)351 210 2627 (Work phone at MPI-CBG)
+49 (0)351 210 1078 (Fax MPI-CBG LMF)
http://www.bioimagexd.net BioImageXD
http://pacific.mpi-cbg.de Fiji (is just ImageJ - batteries included)
http://www.chalkie.org.uk[hidden email]
(
[hidden email] )