Unless I misunderstand the problem, one possible approach is to use a CSG
representation for the objects, subdivide the space into boxes, eliminate
boxes where the implicit surface can't be, and use a marching-cubes algorithm
to recover the implicit surface from the remaining boxes.
I forsee a few difficulties [!?] such as where containing volumes touch.
Still, there does seem to be a fair amount of literature covering these topics.
The boxes themselves could be represented via Interval vectors. Hence questions
such as "does this box contain a portion of my implicit surface" may be answered
"yes", "no", "don't know so subdivide the box", and "don't know but run out of
floating point precision so can't refine further".
There seems to be a pretty good synergy between this problem and those of
representing convex hulls, pavings, subpavings, set computation, interval
arithmetic etc.
============================================================================
,-_|\ Richard Smith - SE Melbourne
/ \ Sun Microsystems Australia Phone : +61 3 9869 6200
[EMAIL PROTECTED] Direct : +61 3 9869 6224
\_,-._/ 476 St Kilda Road Fax : +61 3 9869 6290
v Melbourne Vic 3004 Australia
===========================================================================
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff JAVA3D-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
