Content-Adaptive Parallax Barriers: Optimizing Dual-Layer 3D Displays using Low-Rank Light Field Factorization

摘要

We optimize automultiscopic displays built by stacking a pair ofrnmodified LCD panels. To date, such dual-stacked LCDs have usedrnheuristic parallax barriers for view-dependent imagery: the frontrnLCD shows a fixed array of slits or pinholes, independent of thernmulti-view content. While prior works adapt the spacing betweenrnslits or pinholes, depending on viewer position, we show both layersrncan also be adapted to the multi-view content, increasing brightnessrnand refresh rate. Unlike conventional barriers, both masks are allowedrnto exhibit non-binary opacities. It is shown that any 4D lightrnfield emitted by a dual-stacked LCD is the tensor product of tworn2D masks. Thus, any pair of 1D masks only achieves a rank-1 approximationrnof a 2D light field. Temporal multiplexing of masks isrnshown to achieve higher-rank approximations. Non-negative matrixrnfactorization (NMF) minimizes the weighted Euclidean distancernbetween a target light field and that emitted by the display. Simulationsrnand experiments characterize the resulting content-adaptivernparallax barriers for low-rank light field approximation.