| Summary: | The projection of two-dimensional video images diffractively using computer-generated holograms displayed on a spatial light modulator (SLM) offers compelling advantages over conventional approaches to video projection. However, the images that result are generally of poor quality, exhibiting low contrast and a significant amount of reproduction noise which leads to graininess and non-uniformity. In this thesis, we begin by modelling this noise, showing how its presence is inevitable in conventional approaches to holographic projection. By studying its perceptual effect, we present a novel paradigm for holographic projection, whereby each video frame is formed not by a single hologram but by a set of holograms, optimised collectively to address the problem of reconstruction noise by exploiting the temporal frequency response characteristics of the eye, and present the one-step phase retrieval (OSPR) algorithms to take advantage of this principle, going on to demonstrate how holographic projection can formhigh-quality images at a sub-diffraction-limited resolution. In a real-world system, beam profile non-uniformities, optical aberrations and microdisplay defects can degrade system performance. We model these effects and show how they can be compensated for in a manner not achievable with conventional approaches to projection. We go on to demonstrate for the first time a full-colour high-quality two-dimensional holographic video projection system, and we then extend the algorithms and approaches discussed to three-dimensional video display and projection using the OSPR principle. We conclude by presenting an ultra-miniature (~ 57 cm<sup>3</sup>) holographic video projection system.
|
|---|
