High contrast computational imaging with vortex phase diversity

Abstract

Abstract Optical imaging systems employing spatially incoherent illumination are widely used in routine imaging applications like photography and bright-field microscopy. We describe an incoherent computational imaging system that uses an open aperture as well as a vortex phase aperture for recording the same scene. The two raw recorded images provide a diversity of information that can be effectively combined using the generalized Wiener filter. For the specific choice of aperture functions used here, the two corresponding generalized Wiener filters have nearly opposing polarity. This property leads to an effective computational PSF whose central lobe is $0.6$ times smaller compared to the diffraction-limited PSF and has a super-oscillatory character with side-lobes. The resultant computational imaging system provides images with significantly improved contrast. While our methodology requires two image records, the enhanced point spread function (PSF) with super-oscillatory character is obtained by employing bulk off-the-shelf optical elements instead of sub-wavelength structured masks. The vortex phase diversity concept along with computational image reconstructions are illustrated with both simulation and experimental data. The proposed imaging methodology may be used to improve imaging performance for wide ranging imaging systems without changing their form factor.