Shift-Invariant-Subspace Discretization and Volume Reconstruction for Light Field Microscopy

Light Field Microscopy (LFM) is an imaging technique that captures 3D spatial information with a single 2D image. LFM attractive because of its relatively simple implementation and fast volume acquisition rate. Capturing time series at camera frame rate can enable the study behaviour many biological systems. For instance, it could provide insights into communication dynamics living neural networks. However, conventional reconstruction algorithms for typically suffer from high computational cost, low lateral resolution, artifacts. In this work, we origin these issues propose novel techniques to improve performance process. First, discretization approach uses shift-invariant subspaces generalize typical framework used in LFM. Then, shift-invariant-subspace assumption as prior under ideal conditions. Furthermore, present method reduce forward model by using singular value decomposition (SVD). Finally, use iterative approaches incorporate additional priors perform artifact-free real light field images. We experimentally show our performs better than Richardson-Lucy-based strategies time, image quality, artifact reduction.