Computational 3D displays

Computational 3D display is an emerging class of display systems, which integrates spatial light modulators (SLMs) and optical components by computational methods. Those displays aim to reproduce a complete set of 3D visual cues. In this chapter, we first introduce different aspects of the human visual perception for display systems by interpreting spatial, spectral, temporal, and intensity parameters of the plenoptic function. Next, we provide a taxonomy of 3D display systems and categorize them based on whether headgear or visual separators are required for stereo cues (autostereoscopic vs stereoscopic displays), and how the depth cue is reproduced (varifocal vs multifocal displays). Then, we focus on a subclass of computational 3D displays, optical see-through displays, as a representative display type widely used in AR. Finally, we introduce perception-driven rendering algorithms, which can enhance the quality and performance of the applicable computational 3D displays.