Stochastic optimal phase retrieval algorithm for high contrast imaging

Amir Give'on, N. Jeremy Kasdin, Robert J. Vanderbei, David N. Spergel, Michael G. Littman, Pini Gurfil

Research output: Contribution to journalConference articlepeer-review

2 Scopus citations


The Princeton University Terrestrial Planet Finder (TPF) has been working on a novel method for direct imaging of extra solar planets using a shaped-pupil coronagraph. The entrance pupil of the coronagraph is optimized to have a point spread function (PSF) that provides the suppression level needed at the angular separation required for detection of extra solar planets. When integration time is to be minimized, the photon count at the planet location in the image plane is a Poisson distributed random process. The ultimate limitation of these high-dynamic-range imaging systems comes from scattering due to imperfections in the optical surfaces of the collecting system. The first step in correcting the wavefront errors is the estimation of the phase aberrations. The phase aberration caused by these imperfections is assumed to be a sum of two-dimensional sinusoidal functions. Its parameters are estimated using a global search with a genetic algorithm and a local optimization with the BFGS quasi-Newton method with a mixed quadratic and cubic line search procedure.

Original languageEnglish (US)
Pages (from-to)276-287
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
StatePublished - 2003
EventAstronomical Adaptive Optics Systems and Applications - San Diego, CA, United States
Duration: Aug 3 2003Aug 4 2003

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering


  • Adaptive optics
  • High contrast imaging
  • Phase retrieval


Dive into the research topics of 'Stochastic optimal phase retrieval algorithm for high contrast imaging'. Together they form a unique fingerprint.

Cite this