Lightfield photography and the tomographic analysis of optical phase space

Current trends in optical imaging technology are characterized by the integrated design of optical hardware and signal processing. This synergy is also at the core of the recent resurrection of Lippmann's integral imaging technique in the form of lightfield cameras. Using a Shack-Hartman sensor in the image plane of an imaging system a four-dimensional data cube, called the "lightfield", can be recorded, which is identified as the phase space of geometrical optics. The lightfield can be used to correct numerically for image aberrations, and motion blur after the image data have been recorded. The recording of lightfields shares various features with phase-space tomography, a phase retrieval method based on multiple intensity measurements. From this observation phase-space tomographycan be generalized to include integral photography as a special case. The phase space interpretation further allows us to recognize optical sensing and imaging as a single discipline. Topics like feature specific imaging and compressed sensing are interpreted as tomographic reconstructions of phase space. While the presentation emphasizes heuristic arguments, the discussion highlights numerical tools for a practical implementation.