[CST-2] AdvGraph: CSG

[Kevin Chan]

PS - he did go through it quite thoroughly in lectures so examinable it is
=)

----- Original Message -----
From: "Kevin Chan" <kkfc2@cam.ac.uk>
To: <cst-2@srcf.ucam.org>
Sent: Thursday, May 31, 2001 1:36 AM
Subject: Re: [CST-2] AdvGraph: CSG


> Leaves are primative objects (defined by the first (entry) and last (exit)
> intersection points with the ray) e.g Sphere[t1,t2] where t1 gives the
first
> intersection point and t2 the last
>
> Nodes are CSG operators
>
> The tree therefore looks like:
>
>                                 U
>                               /     \
>                             /         \
>                        Box        Sphere
>                     [b1, b2]     [s1,s2]
>
> You then work from the bottom up combining the leaves using the operators
to
> get the intersection points of the higher nodes.
>
> This is better explained with a diagram.  Look at diagram 1 (sorry about
its
> crudity - it is afterall 1.30am and I can't be bothered to use anything
more
> sophisticated than paint =)
>
> S1, S2, B1 & B2 represent the intesections with the two objects.  From
this
> it is obvious that the first intersection points of the following objects
> are:
>
>
>          S                     ->  S1
>          B                     ->  B1
>       S U  B                ->  S1
>       S intersection B   -> B1
>       S - B                   -> S1
>       B - S                   -> S2
>
> The last intersection point is just as easy to work out.  You can
therefore
> work upwards from the leaves and work out the intersection points for each
> node until you get to the top where you will end up with the intersection
> points for the whole CSG object.
>
> Hope that helps - if not I'll try to give a more coherent explanation in
the
> morning =)
>
> Kev
>
>
>
>
>
>
>
> ----- Original Message -----
> From: "M.Y.W.Y. Becker" <mywyb2@cam.ac.uk>
> To: <cst-2@srcf.ucam.org>
> Sent: Thursday, May 31, 2001 1:09 AM
> Subject: [CST-2] AdvGraph: CSG
>
>
> > Can anyone remember how a binary tree representation of a CSG object can
> be
> > used to compute intersection points of a ray with the entire object
> > (1999/9/4)?
> >
> > It's not in the notes but it seems to be examinable...
> >
> > Moritz
> >
> >
> >
> >
> > _______________________________________________
> > CST-2 mailing list
> > CST-2@srcf.ucam.org
> > http://www.srcf.ucam.org/mailman/listinfo/cst-2
> >
>