[Bf-committers] 2d Image to 3d?

Mal mal at candomultimedia.com
Wed Feb 16 02:10:21 CET 2005


Hi Austin,

> Would it be too much work to write a sub-routine that takes three 
> parts of an orthographic image (top, bottom, side) and turns them into 
> a mesh? Would it even be possible?
>
The best you could get from 3 ortho images  ( assuming that the outline 
of the shapes could be automatically detected ) would be a bit like this...

- Imagine creating 3 cookie cutter shapes from each of the 3 outlines
- Take a cube of clay
- Use the cookie cutters, one at a time, to cut through the clay from 
their respective directions, discarding everything outside the cutter ( 
effectively a Boolean AND )

You will end up with a rough estimate of the model, which might be 
enough to start modelling from for some users.

Something like that might be an interesting project to create in Python.

You would need...

a) routine to detect outline of each image
b) routine to convert each outline into a closed curve
c) routine extrude that closed curve into a solid, extruded shape ( 
deeper than the virtual cube of clay )
d) routine to Boolean: AND each of the shapes, so you end up with only 
the parts of the model that overlap.

You could then map the textures directly from the images onto the model, 
from the 3 viewports, with maybe a bit of tweaking for faces that are 
shared between views ( eg a face left over from the booleaning that is 
at 45% to both the front and side ortho view )

Hope this helps!  It would be a very interesting project, and one that 
I'm sure a lot of people would make use of.
Mal


More information about the Bf-committers mailing list