PDA

View Full Version : SurfDriver software/DXF-IGES conversion (summary)



Stephen Ferguson
11-02-1997, 07:08 PM
Here is a summary of the responses that I received to my original
query. First, the original question:

Is anybody familiar with the SurfDriver software? It's
a nice little program for generating 2D contours from
.bmp images (e.g. MRI, CT, histological sections) and then
creating 3D entities from those contours. The program
can export the objects in DXF format. However, most finite
element programs read in IGES data only.
First, I should admit my ignorance of the two file formats.
I know the DXF format from exporting 2D vector drawings
from AutoCAD. Is a DXF file of a 3D object truly 3D data
that would be useful for building an FE model? Does anybody know
a cheap solution for translating DXF files to IGES? I
would be generating just a few solid models, but the only
commercial translator packages that I have found, via the
www, are prohibitively expensive.

Thank you to all who responded. From the responses, I found out that:

1. DXF files now contain true 3D data, despite being originally
written as a method to transport 2D drafting files between
CAD (Computer Aided Drafting/Design) systems.

2. IGES translation capabilities are built into several CAD packages
such as AutoCAD and Cadkey, but often as an extra cost upgrade.
This extra cost can be very high!

3. The ease of conversion of this 3D data to a form that can be
used to generate a finite-element solid model depends on the
way that the 3D surface is represented by the original software (in
this case SurfDriver). Surfaces represented by tiles (small triangles)
are often easily transformed into a solid model. See, for example, this
nice paper from Marco Viceconti:

take a look at http://www.tecno.ior.it/people/papers/vicecmbbe97.pdf
It is full paper in Acrobat format which will be published in the
proceedeings of CMBBE97.
In that paper I describe a technique we have succesfully used to treat
tiled solids with solid modellers.

If the surfaces are represented by higher-order curves and splines, then
the process can be a little more difficult.

4. A lot of people asked about the SurfDriver software. The web page can
be found at:
http://surfdriver.ml.org/
and can also be found by following links from the Visible
Human web page if you want to read related material.
The software does a nice job of handling the bifurcations that have to
be modelled when generating 3D structures from anatomical data.
Apparently there is a fix on the way to allow the software to handle
shapes with cavities, such as the medullary canal in bones. I have tried
the demo version of the software on some coarse bitmaps of the
Visible Human data set and the resulting shape seems fairly good, although
there are some strange transitions from one contour to the next that
may be a result of the crude detail level that I tried. Try it
yourself and see!

regards,
Stephen Ferguson
ASIF Reseach Institute
Davos, Switzerland

-------------------------------------------------------------------
To unsubscribe send UNSUBSCRIBE BIOMCH-L to LISTSERV@nic.surfnet.nl
For more information: http://www.kin.ucalgary.ca/isb/biomch-l.html
-------------------------------------------------------------------