Near Real-Time Estimation of Super-Resolved Depth and All-In-Focus Images from a Plenoptic Camera Using Graphics Processing Units

Algorithm 1

Input: Lightfield 𝐿_𝐹(𝐱,𝐮), 𝐱=(𝑥₁,𝑥₂)^′, 𝐮=(𝑢₁,𝑢₂)^′, |𝑥|≤𝑟|𝑢|≤𝑠.

Slope Δ𝑢/Δ𝑥 with Δ𝑢,Δ𝑥∈𝑍 as in the previous theorem

Output: Super-resolution focal stack image 𝐹_{Δ𝑢/Δ𝑥}(𝐤)

Compute g =g.c.d.(Δ𝑢,Δ𝑥), Δ𝑢_𝑔=Δ𝑢/𝑔 and Δ𝑥_𝑔=Δ𝑥/𝑔

For each microlens 𝐱=(𝑥₁,𝑥₂)^′

  For each pixel in that microlens 𝐮=(𝑢₁,𝑢₂)^′

    Compute 𝐤=(𝑘₁,𝑘₂)^′, 𝐤=Δ𝑢_𝑔𝐱−Δ𝑥_𝑔𝐮

    Update the mean value in 𝐹_{Δ𝑢/Δ𝑥}(𝐤) using 𝐿_𝐹(𝐱,𝐮)