Abstract

Information about objects in a scene can sometimes be extracted directly from a spatial autocorrelation function (or equivalently, a power spectrum). Such information may be the quantity of ultimate interest, or it may constrain or provide trial solutions for an iterative image reconstruction. Previous research has described extraction of the most basic of such information, consisting of the number of objects present and their relative positions. From there it is possible to unfold certain properties of individual extended objects (that is, two-dimensional image-plane brightness distributions). I show that when several objects are present, their defining parameter values can be extracted from the corresponding properties of subsidiary features in a set of spatial autocorrelation data by use of a least-squares approach. Furthermore, many potential ambiguities in the single-object case do not arise with multiple objects.

© 1993 Optical Society of America

Extracting number and relative positions of multiple extended objects from autocorrelation data

George E. Cipperly
Appl. Opt. 32(5) 699-706 (1993)

Object detection from a real scene using the correlation peak coordinates of multiple circular harmonic filters

Yunlong Sheng and Henri H. Arsenault
Appl. Opt. 28(2) 245-249 (1989)

Multiple-object input in nonlinear correlation

Tuvia Kotzer, Joseph Rosen, and Joseph Shamir
Appl. Opt. 32(11) 1919-1932 (1993)

Triple and autocorrelation functions from factorial moments measured

M. P. Cagigal
Appl. Opt. 32(17) 3059-3063 (1993)

Multiple-object detection with a chirp-encoded joint transform correlator

Qing Tang and Bahram Javidi
Appl. Opt. 32(26) 5079-5088 (1993)