|
|
Hi all,
to conclude this thread:
as Michael Schmid pointed out, the algorithm in the image-stacks-3D project program doesn't
work for stacks which are flat and wide (1000 pixels wide, only 100 slices deep) since this
algorithm assumes that you want to turn the stack by 90° or more. So it places the flat stack into
the center of a cube of 1000 pixel each length, and does the depth cueing along the 1000 pixesl
of this expanded stack. Then of course the difference between slice 1 and slice 100 (now being
counted as slice 450 and slice 550) is small relative to the depth of the cube (1000).
Here a workaround for generating a dephth-cued stereo projection of such a "flat" stack:
recalculate the stack (don't forget to make a copy because the data will be altered!), and "dimm"
the deeper slices using the following macro:
Process-Math-Macro: v=v*(1-N*z/h),
with N being a number you have to insert (the other variables are imageJ variables).
For me, N=11 [i.e. v=v*(1-11*z/h)] worked for a 30 slice stack, and N=3 worked for a 127 slice
stack.
Best,
Martin
|
Locked
