"Norbert Kern" <nor### [at] t-onlinede> wrote:
>
>...
> Much better than my own approach to depict mri scans.

I really like the way you have arranged the glass sheets and how the light
escapes at their edges.

BTW: Where did you find the MRI-data for this ?


> How did you remove bones, eyes etc.?

Luckily I did not have to deal with that. It had already been done, so the
DF3-file only contains the brain.

A while ago I made some (not very successful) attempts at filtering the data (in
Python) for a CT-scan according to a range of Hounsfield units for bone:
https://en.wikipedia.org/wiki/Hounsfield_scale

IIRC the problem was that the range for the bones overlapped the range for some
other tissue. (It might be that it wasn't a high quality scan.) I suspect that
they now use quite advanced techniques to isolate the different organs. I would
guess that some techniques find "gradient surfaces" in the data.

--
Tor Olav
http://subcube.com

Post a reply to this message

"Tor Olav Kristensen" <tor### [at] TOBEREMOVEDgmailcom> wrote:

> I had forgotten about the supershapes, so I searched for them
> - and ended up at one of Paul Bourke's web pages again:
> http://paulbourke.net/geometry/supershape/
>
> I see now that there's two files there that I've made; ss_macro1.inc &
> example1.pov
> - So apparently I have been experimenting with them once...
>
> It seems that they are commonly made with parametric meshes. Perhaps it is
> possible to model some of them as isosurfaces. But that would be very slow, so I
> think I'll wait some years (for more computing power) before looking into that.

I did it in 8/2017 with isosurfaces, so I would imagine you could do an
approximation of some sort to speed it up.

Here's my see-them-all scene if you wanted some code to to run a quick test
scene:


#version 3.71;
global_settings {
 assumed_gamma 1.0
}

#include "colors.inc"





#declare m  =  7;  // number of lobes
#declare n1 =  5;
#declare n2 =  5;
#declare n3 =  5;
#declare a  =  1;
#declare b  =  1;

#declare M = 10;  // iterations

#declare AL = 20;
#declare L = 0.05;
cylinder {<-AL*4, 0, 0>, <AL*4, 0, 0>, L pigment {Red} no_shadow}
cylinder {<0, -AL*4, 0>, <0, AL*4, 0>, L pigment {Green} no_shadow}
cylinder {<0, 0, -AL*4>, <0, 0, AL*4>, L pigment {Blue} no_shadow}

camera {
  location <M*4, 0, -75>
  look_at  <0, 0, 0>
  right x*image_width/image_height
  up 1*y
}

#declare LS = <M*4, M*2, -100>;

light_source {LS color White}
cylinder {LS, <0, 0, 0>, 0.01 pigment {White} no_shadow}

#declare CB = 4;

#for (n1, 1, M)
 #for (n2, 1, M)
  #for (m, 1, M)
   isosurface {
    function {
     pow (
      pow (abs (cos (0.25 * m * acos (y/sqrt(x*x+y*y)) * abs(y)/y) / a), n2)  +
      pow (abs (sin (0.25 * m * acos (y/sqrt(x*x+y*y)) * abs(y)/y) / b), n3)
      , (1/n1)
     )
     //* sqrt (x*x+y*y) - 1 }
     * sqrt ((x*x)+(y*y)+(z*z)) - 1
    }

        contained_by { box { <-CB, -CB, -CB>, <CB, CB, CB> } }
        accuracy 0.01
        max_gradient 12
        scale <1, 1, 1>
        #declare Hue = m * (360/M);
        texture {pigment {color CHSV2RGB (rgb <Hue, 1, 1>)} finish {specular
0.2}}
        translate -x*M
        translate  x*m*3
        translate  y*M
        translate -y*n2*3
        translate -z*M
        translate  z*n1*3
   }

  #end // end for m
 #end // end for n2
#end // end for n1

Post a reply to this message