Re: stereo photography reverse engineering

From: Ron Jamieson (
Date: Thu Jul 17 1997 - 15:20:10 EEST

> Ron Jamieson wrote:
> >
> > A colleague and I have just finished reverse engineering a cast of a
> > human with data extracted from a stereo image. This was not new to me
> > in that I am aware of CT and MRI scans. But its been an interesting
> > exercise and one that I cant help feeling has more to offer. For
> Dear Mr. Jamieson,
> I wonder how you accomplished this? Would you share this information?
> Thanks,

A few people have asked this question so I,ll copy this to the mailing
list. There are several expensive all singing all dancing systems
around. However, the way we did it was as follows:

Using a Zeis UMK 10-1318 large format camera we took two photos 1
metre apart at a stand off of 1.7 metres from the object. This was
repeated at four seperate positions around the object. The idea is to
mimic the right and left eye and using these settings we have a
ratio 1/1.7 for depth measurement.

These photos were then placed in our Zeis P3 analytical plotter, the
photos were lined up by means of checking common points and any
adjustment of images is made.

Points are then picked off as required and appended to a data file
containing their co-ordinates.

These points are used to form splines in our CAD package and a
surface is wrapped around.

After that its up to you! STL, Direct slices or just DXF 2D drawings
for the fabricator.

Hope this helps, we plan to do more and I,m still looking for that
unique angle for this particular reverse engineering process.



Bruce Building
Newcastle University
Newcastle NE1 7RU

phone 0191-2225847
fax 0191-2225833

This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:39:53 EEST