Re: Skulls

From: Chase D. Garfinkle (
Date: Thu Apr 09 1998 - 00:47:41 EEST

Jon Wild <> writes:

> Can you help? I am trying to get hold of an accurate scan of a human skull
> with at least one half in perfect condition? My end aim is to generate a
> number of horizontal sections through the eye socket area. I am told that I
> will NOT be able to generate the outlines I need from an STL file? Ideally
> I need to generate my sections in the Frankfort plane (Anthropometric
> reference plane which passes through the ear hole and the bottom of the eye
> sockets). CAT scans are generally taken at an arbitrary angle that is why I
> need the whole 3D file of the skull so I can orientate it how I please.

There's a 200^3 volume data set of a human skull at,
in /pub/morph/extra_volumes/human.den.gz . This data set was extracted from a
CT scan of a female cadaver, so it should be accurate, but we were only using
it for research in volumetric morphing techniques, so I can't give you any
iron-clad guarantee about its precision. (There's also a data set from an
orangutan skull if you're interested, but it's been touched up to make it look
older.) The volumes are in a file format which is basically just a header and
a series of bytes for density values; the format is described in a file called
FORMAT in the same directory, and there's source code for reading and writing
these files in /pub/morph/code_and_samples/morph.tar.gz in denfile.{c,h} .


Chase D. Garfinkle                650-306-8917

For more information about the rp-ml, see

This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:45:17 EEST