pyolaf is a Python port of the oLaF 3D reconstruction framework for light field microscopy (LFM).
The light field microscope (LFM) allows for 3D imaging of fluorescent specimens using an array of micro-lenses (MLA) that capture both spatial and directional light field information in a single shot. oLaF is a Matlab framework for 3D reconstruction of LFM data with a deconvolution algorithm that reduces aliasing artifacts.
pyolaf brings these same features to the Python ecosystem, using GPU acceleration and some further code optimizations to speed up deconvolution by 20x.
pyolaf only supports regular grids and single-focus conventional light-field microscopes. In particular Fourier LFM, hexagonal grids, and multi-focus lenslets are currently not supported. Pull requests to add these are welcome!
(Left) Raw image a fly from light field microscope, showing the 10k+ lenslets in the microlens array, acquired within Tuthill Lab
(Right) Volume reconstructed from the image, shown as successive slices in a gif. See also the higher resolution video.
Copyright (c) 2017-2020 Anca Stefanoiu, Josue Page, and Tobias Lasser -- original oLaF code
Copyright (c) 2023 Lili Karashchuk -- pyolaf
When using pyolaf in academic publications, please reference the following citation:
- A. Stefanoiu et. al., "Artifact-free deconvolution in light field microscopy", Opt. Express, 27(22):31644, (2019).