Browse Prior Art Database

'Winning Primitive' in Point-Membership Tests

IP.com Disclosure Number: IPCOM000101397D
Original Publication Date: 1990-Aug-01
Included in the Prior Art Database: 2005-Mar-16
Document File: 2 page(s) / 70K

Publishing Venue

IBM

Related People

Quarendon, P: AUTHOR [+3]

Abstract

This article describes a programming implementation for Solid Modelling Systems to permit the attachment of a primitive identifier to the result of set-theoretic operations within the inside of a solid model. In particular the article relates to graphical display and to the IBM Winchester Solid Modeller System (WINSOM) but is applicable to graphical display and other systems employing similar structure and algorithms.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

'Winning Primitive' in Point-Membership Tests

       This article describes a programming implementation for
Solid Modelling Systems to permit the attachment of a primitive
identifier to the result of set-theoretic operations within the
inside of a solid model.  In particular the article relates to
graphical display and to the IBM Winchester Solid Modeller System
(WINSOM) but is applicable to graphical display and other systems
employing similar structure and algorithms.

      A point membership test determines whether a point in space is
inside or outside a set-theoretic solid model.  In the classical form
described by Tilove, logical values are combined to give a logical
result.  Replacing each primitive by a penalty function, that is,
zero at the surface of the primitive, and increasingly negative
within it but increasingly positive outside it, while also replacing
the set-theoretic operators by maximum and minimum operations, a
penalty function is obtained from the membership test that is an
approximate measure of distance from the surface of the object.  Such
tests are commonly used.

      The subject of this disclosure is a simple extension to the
above.  Notice that a maximum or minimum operation between two
penalty functions can be made to return, not only the appropriate
value of the penalty function, but also the name of the primitive
which contributed 'WINNING the 'winning' value (i.e., the one
responsible for the numerical result).  This name then becomes
associated with the penalty function for the new sub-model, and thus
a si...