Raw light field images need to be resampled according to the number of desired virtual pixels, with the grid aligned to each microlens. Currently this is done separately through old software from the Computational Imaging group at Stanford (see: http://graphics.stanford.edu/software/LFDisplay/ )
A pure Julia implementation ought to be possible via Interpolations.jl. Would also be nice to improve the process by making it automated (automated peak finding to align the microlenses instead of manual alignment as done in LFDisplay)
Raw light field images need to be resampled according to the number of desired virtual pixels, with the grid aligned to each microlens. Currently this is done separately through old software from the Computational Imaging group at Stanford (see: http://graphics.stanford.edu/software/LFDisplay/ )
A pure Julia implementation ought to be possible via
Interpolations.jl. Would also be nice to improve the process by making it automated (automated peak finding to align the microlenses instead of manual alignment as done in LFDisplay)