Abstract

Use of a solid modeler in the implementation of a non-sequential ray tracing engine has a number of advantages not easily achieved using sequential surface based optical system descriptions. These include a solid modeler’s ability to represent a physical part unambiguously as a bounded entity, with a well defined inside and outside. Reliable algorithms for transferring and intersecting rays in a solids environment with homogeneous optical media are compact and general purpose. By associating optical properties (e.g., refraction, reflection, diffraction, and scattering) to the surfaces of the solids, one can generate one or more scattered rays to track the geometrical exitance from the surface. Use of solids with optical surface properties provides a modeling architecture that is both powerful and extensible to handle new types of phenomena with no change to the basic ray trace algorithms.

© 1998 Optical Society of America