University of Rochester
Institute of OpticsPRAISE
Institute of Optics
Professor Fienup
Group members
Manuel Guizar-Sicairos' Research Page

As the desired resolution of an imaging system is increased, the aperture and depth of the optical system increases, similarly the cost and difficulty of creating the required aberration corrected surfaces becomes larger and unpractical at a fast rate. An imaging scheme that does not require imaging optics, permits the receiver to be wide and more than an order of magnitude thinner than its optical counterpart, as only a large intensity detector array is needed.

The proposed scheme consists of an object of interest being illuminated with coherent light having a shaped illumination pattern, and a simple array of intensity detectors to detect the speckle intensity pattern reflected from the object without imaging optics. If the shaped illumination pattern is known a priori, it provides an object support constraint for the reconstruction algorithm.

Since both the field intensity and phase are crucial for the reconstruction of the original object, the phase distribution must be retrieved.

A sample reconstruction of a complex valued object from the amplitude of its far field intensity and a support constraint on the object is shown; the hybrid input-output algorithm was used in this reconstruction.

Notice that since the support (or illumination shape) is symmetric, a twin image solution is allowed.

Current phase retrieval algorithms have very good performance if a sharp edge illumination is used, however, since the main advantage of this technique is the development of high resolution imaging with thin and possibly conformal devices, an illumination that is sharp to the order of the desired resolution and requires large-aperture projection optics would be disadvantageous.

Further research is required to find robust algorithms that can reconstruct objects with tapered edges, and illumination shapes that favor the reconstruction.

I am also in collaboration with the PMOG in research projects that deal with numerical methods for wave-front propagation and the study of propagation characteristics of novel beams. Helmholtz-Gauss beams are a solution of the paraxial wave equation with a field distribution at z = 0 that is given by the product of a Gaussian amplitude and a fundamental solution of the transverse Helmholtz equation.

These beams, a physical realization of non-diffracting wave-fields, are of interest because they preserve their transverse intensity distribution within a finite distance, and their propagation characteristics offer potential applications in optical trapping, telecommunications and patterned illumination.