View Full Version : Anyone know about Computational Geometry?

09 September 2006, 07:40 AM
Hello, this is my first post on CGTalk

I've used Blender to create an 80-second animation showing how an invention would work (it can be downlud from my site ( The invention is a method for building walls from irregularly-shaped rocks where a computer fits the rocks together in the virtual world before outputting instructions to the builder.

It's just a concept at the moment as I don't yet have a computer program that could take a number of irregular polyhedra of no more than 10 faces each and fit them together. With Blender I can rotate and translate polyhedra and fit them together by eye but I need a program to do that automatically.

I've talked to lots of mathematically-minded computer scientists to find out if such a program already exists and they know of none. So I thought I'd ask those who create CG applications as I thought "they would know, or they might know someone who knows". I thought I'd see what's already out there before I attempt to write it myself or pay someone to do it.

Can anyone help me?


09 September 2006, 07:29 PM
The closest thing I'm aware of to what you seem to be looking for is software used in archeology to reconstruct pottery from found fragments. I think IBM was doing something similar, a while ago.

09 September 2006, 05:11 AM
try michael hofer's work on that:

(near the bottom of the page)
Reassembling Fractured Objects by Geometric Matching. (


09 September 2006, 08:05 PM
Please don't help this guy. If you check his site, he is applying for a patent on this, and there is clearly prior art in the field. If he wants to get a software patent on techniques which already exist, please let him do the work himself.

He gets a patent which can later be sold or used in predatory patent money grabs, when you did the hard work of finding the existing techniques for him, so he could claim them as covered by his patent.

Totally unethical.

09 September 2006, 02:07 AM
Prior art is documented work which has been done previously. Patent offices around the world are quite strict about what can and can't be claimed as novel in a new patent. Prior art can not be claimed as being novel in a new patent. So it's not unethical to enquire as to what prior art is out there. It is a normal and sensible part of innovation and also a neccessary part of what every inventor should be doing so as not to waste money trying to patent something someone else has already done. It would be unethical if an inventor tried to sectretly obtain unpublished prior art and claim it as part of the patent.

From my previous post: "I thought I'd see what's already out there before I attempt to write it myself or pay someone to do it" - If I write an original program myself then I can claim the intellectual property (IP) for myself. If I pay someone to do it then ownership of the IP will be agreed apon as part of the contract for the work to be done. If I find a suitable program already exists then I will have to get permission from the owners of the IP to use it. As far as I know that's normal business practice.

CGTalk Moderation
09 September 2006, 02:07 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.