Research data supporting "Multi-Depth Phase-Only Hologram Optimization using L-BFGS Algorithm with Sequential Slicing"

We implement limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) optimization of phase-only computer-generated hologram (CGH) for multi-depth three-dimensional (3D) target. Instead of computing the full 3D reconstruction of the hologram, we use a novel method using L-BFGS with sequential slicing (SS) for partial evaluation of the hologram during optimization, that only computes loss for a single slice of the reconstruction at every iteration. The dataset is generated to demonstrate that the ability to record curvature information makes L-BFGS a perfect match with the SS technique. The baseline methods for comparison are sum-of-holograms (SoH) and sum-of-loss (SoL). The SoH technique simply takes the sum of holograms individually computed for each slice. The SoL technique computes the full 3D reconstruction of the hologram at each iteration and calculates the sum of loss. Data is generated for SoH and SoL techinques to form the baselines of comparison. All files in the dataset are images in PNG format. The files with "holo" in the filename are the phase holograms, with each pixel value indicating the phase of the hologram mapped to the [0,255] range. The files with "recon" in the filename are therefore the reconstructions of the hologram, at different depths. The plots with "iter" in the filename are the raw plots for all combinations of the optimizaiton algorithms and loss functions involved.