Abstract

Real-time holovideo machines consist of three primary components: 3-D data generation (or sensing) of the data to be viewed, the processing to reconstruct the appropriate holographic wavefront, and the holographic display device itself. Numerous technologies for 3-D sensors have fully matured (stereo, active ranging, etc) as have constructive solid geometry modeling techniques for synthetic 3-D scene generation. Advances in high-resolution holographic display devices utilizing acousto optic devices¹ or liquid crystal on silicon SLMs² have proven challenging but feasible. In this paper we discuss the final remaining bottleneck facing holovideo displays due to the nearly intractable computational burden required to generate the holographic fringe pattern in real time. Several novel optical processing architectures for fringe computation and initial experimental results are presented.

© 1997 Optical Society of America