Browse Prior Art Database

Solid Light for Solid Modelling

IP.com Disclosure Number: IPCOM000101399D
Original Publication Date: 1990-Aug-01
Included in the Prior Art Database: 2005-Mar-16
Document File: 3 page(s) / 125K

Publishing Venue

IBM

Related People

Walter, AGN: AUTHOR

Abstract

"Conventional" computer graphic solid modellers mimic reality using numerous algorithms when calculating lighting. This disclosure shows ways to illuminate and display visual interactions of objects that are impossible by conventional means.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 49% of the total text.

Solid Light for Solid Modelling

       "Conventional" computer graphic solid modellers mimic
reality using numerous algorithms when calculating lighting. This
disclosure shows ways to illuminate and display visual interactions
of objects that are impossible by conventional means.

      "Non-realistic" lighting can highlight a selected portion of
the model, display visual object-overlap, simulate structured
lighting or compute shadows most economically in computer time.
Light is constrained to exist only within an arbitarily shaped
volume.

      Conventional Solid Modellers take two different approaches with
light when rendering an image.  The first approach assumes that for
the sake of computing speed, light may be assumed to pass freely
through solid objects on its way to illuminate any point.  For each
pixel to be drawn on the screen, the modeller determines which object
in that scene is closest to the eyepoint;  that point is then "lit"
assuming that all light sources may contribute.  This method does not
calculate whether any other point in the scene obscures the light
sources, so shadows are not computed. Control over the lighting is
restricted to such parameters as colour of the light,
intensity/distance variability, and sometimes to the "spread" of
light, by using a cone-shaped spread to mimic a spotlight effect.  It
is difficult to control exactly how the final scene appears.  This is
the most usual approach chosen by users of typical systems, such as
the UKSC Winchester Solid-Modeller (WINSOM).  The second approach
attempts to mimic reality at the cost of slow rendering times,
including "ray-tracing" and newer "radiosity" methods.  It results in
such features as hard and soft shadows and reflections appearing.
WINSOM is capable of performing ray-tracing.  But to mimic reality is
often least required when using graphics to display scientific data.

      The disclosure is to imagine the intersection of a light with a
solid object as defining the boundaries of existence of that light.
WINSOM can build complex shapes (objects) by adding, subtracting,
intersecting and exclusive-ORing simpler shapes together via
operations known as Union, Diff, and Intersect.  By allowing a shape
to define the volume in which a light is considered to exist, we can
selectively illuminate portions of a scene in a way impossible in
reality.  This intersection of 'Light' and 'Solid' object is now
abbreviated to Lit_Sol for convenience.  A Lit_Sol is not a solid
object in its own right;  it can only illuminate, or even darken if
given a negative-value, a real solid object.  Consider the problem of
high lighting selected atoms in the model of a molecule, where the
atoms are represented as coloured spheres.  WINSOM cannot provide
extra lighting of such selected atoms without some of the extra light
highlighting other unwanted atoms. Ray-tracing and radiosity methods
are useless because what is wanted is an unrealistic lighting....