Hi. We have been working on the RE problem for some time. We have two
cyberware laser scanners as well as point digitizer but we tend to use laser
digitizers since they are more accurate and frankly easier to do.
On the issue of RE (fitting surface patches) we have been taking a slightly
different approach than others. We are able to fit NURB surfaces to
digitized data - not only that we use trimmed nurbs to create exact boundary
patch. My student presented a paper on this at the Austin SFF conference in
august. You can retrieve a copy of the paper (which has examples) from
http://prism.asu.edu -> go to Archives and then publications - its the
first/second paper.
What the paper does not have are the examples for fitting full size figures.
We have applied the same algorithm to fit whole face surfaces and also full
figures. Now, ofcourse there is always a caveat when it comes to RE and the
question is what tolerance you were looking for. If you do a full figure
suface fitting which is quite difficult in one shot - you are not going to
get .00XX mm tolerance unless you use a large number of segments. To see two
examples of body/bust rev eng go to following:
http://prism.asu.edu/~razdan/Images/RevEng
and look at the first four images
Polygon_DanBust.jpg
Polygon_TwistDan.jpg
Shaded_DanBust.jpg
Shaded_TwistDan.jpg
The twisted figure was created when the person being digitized spun himself
on a lazy susan while being digitized to create the equivalent of 3D Xerox
blur or sculpture. The piece has been prototyped on FDM 1650 in wax and cast
in bronze with the lost wax process at the ASU Art foundary. There is always
the issue of how to ahndle fragmented hair - some parts get digitized and
some dont.
Thanks
Anshuman Razdan
> -----Original Message-----
> From: owner-rp-ml@ltk.hut.fi [mailto:owner-rp-ml@ltk.hut.fi]On Behalf Of
> lonepeak
> Sent: Thursday, December 03, 1998 8:41 AM
> To: RP-ML
> Subject: Digitizing / Laser Scanning
>
>
> We recently bid out a large job that required "reverse engineering" of a
> life size, complicated figure (Star Wars kind of detail). This figure
> had clothes, earrings, head pieces, etc.
>
> We choose to digitize the figure rather than laser scan it but I am
> curious about what experiences others have had in this area and where the
> group as a whole feels the state-of-the-art is with respect to digitizing
> vs laser scanning is currently.
>
> Our requirements for the project were:
>
> 1. Produce both NURBS and Polygon files.
> 2. Create numerous small scale replicas using FDM and 3DP
> 3. Create a 20 foot replica
> 4. Produce CAD files suitable for ProE or I-DEAS
> 5. Create billboard size renderings.
> 6. Mirror features that could be mirrored (for example - half the face,
> the feet)
> 7. Not include "blemishes" in the figure that were pointed out by the
> artists.
> 8. Obtain maximum geometry details for all areas, including overhangs
> (such as in the ear areas).
> 9. Realistic (meaning as small as possible) file sizes.
>
> We are very happy with the files that were generated via digitizing but
> I would like to hear discussions comparing the two technologies.
>
> Thanks in advance!!!
>
> Alair Griffin
> Engineer
> Javelin, a Lone Peak Company
>
>
>
> For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/
>
For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/
This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:47:31 EEST