Towards Non-Lambertian Scenes for Tensor Displays

Tensor displays are screens able to render a light field with correct depth perception without wearing glasses. Such devices have already been shown be accurately scene composed of Lambertian objects. This paper presents the model and prototyping tensor display three layers, using repurposed computer monitors, extends factorization method non-Lambertian Furthermore, we examine relation limitations between depth-of-field range scenes. Non-Lambertian scenes contain out-of-range disparities that can not properly rendered usual optimization method. We propose artificially compress disparity by two fields focused on different depths, effectively solving problem allowing clearly both simulated prototyped display.