ABSTRACT:
We propose a strategy for utilizing a focal stack, i.e., a set of contrastingly engaged pictures, as the contribution for a novel light field show called a “tensor show.” Although this show comprises of just a couple of light lessening layers situated before a backdrop illumination, it very well may be seen from numerous headings (points) all the while without the goals of each survey course being yielded. Ordinarily, a transmittance design is ascertained for each layer from a light field, specifically, an arrangement of thick multi-see pictures (commonly handfuls) that are to be seen from various bearings.
In any case, planning such a monstrous measure of pictures is frequently awkward for genuine articles. We built up a technique that does not require a total light field as the contribution, a central stack made out of just a couple of distinctively engaged pictures is straightforwardly changed into layer designs. Our technique significantly diminishes the cost of obtaining information while additionally keeping up the nature of the yield light field. We approved the strategy with tests utilizing engineered light field datasets and a central stack gained by a normal camera.
BASE PAPER: From Focal Stack to Tensor Light-Field Display
