Fresnel transform phase retrieval from magnitude

Todd A. Pitts, James F. Greenleaf

Research output: Contribution to journalArticlepeer-review

22 Scopus citations

Abstract

This report presents a generalized projection method for recovering the phase of a finite support, two-dimensional signal from knowledge of its magnitude in the spatial position and Fresnel transform domains. We establish the uniqueness of sampled monochromatic scalar field phase given Fresnel transform magnitude and finite region of support constraints for complex signals. We derive an optimally relaxed version of the algorithm resulting in a significant reduction in the number of iterations needed to obtain useful results. An advantage of using the Fresnel transform (as opposed to Fourier) for measurement is that the shift-invariance of the transform operator implies retention of object location information in the transformed image magnitude. As a practical application in the context of ultrasound beam measurement we discuss the determination of small optical phase shifts from near field optical intensity distributions. Experimental data are used to reconstruct the phase shape of an optical field immediately after propagating through a wide bandwidth ultrasonic pulse. The phase of each point on the optical wavefront is proportional to the ray sum of pressure through the ultrasound pulse (assuming low ultrasonic intensity). An entire pressure field was reconstructed in three dimensions and compared with a calibrated hydrophone measurement. The comparison is excellent, demonstrating that the phase retrieval is quantitative.

Original languageEnglish (US)
Pages (from-to)1035-1045
Number of pages11
JournalIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
Volume50
Issue number8
DOIs
StatePublished - Aug 1 2003

ASJC Scopus subject areas

  • Instrumentation
  • Acoustics and Ultrasonics
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Fresnel transform phase retrieval from magnitude'. Together they form a unique fingerprint.

Cite this