Abstract

A compound rotation-invariant filtering technique is developed which leads to sharp cross-correlation peaks for desired targets while suppressing random noise and unwanted target correlations. The basic compound filter structure is a filter bank consisting of several circular harmonic filters. The output image moduli from the filter bank are weighted and summed to form the final output image. In the design procedure, the weights are constrained so that the compound filter output has a peak value of 1 for the desired target. Measures of random noise and deterministic noise due to correlations with unwanted targets are minimized with respect to the weights. The SNR of the compound filter is, therefore, maximized. The weights constitute a real solution to a classical least-squares problem and are, therefore, unique. Some limits to the performance of the technique are discussed. Examples illustrating the filter performance are given. Optical, digital, and hybrid implementations of the compound filter are possible. Extensions of this work are discussed.

© 1987 Optical Society of America

Amplitude compensated matched filters using circular harmonic expansion and a Mellin transform

Ying Sun, Zhao-Qi Wang, and Guo-Guang Mu
Appl. Opt. 29(32) 4779-4783 (1990)

Optical generation of a circular harmonic filter for rotation and translation invariant optical pattern recognition

Yeou-Yen Cheng
Appl. Opt. 26(20) 4413-4415 (1987)

Synthetic circular-harmonic phase-only filter for shift, rotation, and scaling-invariant correlation

Ping Zi-Liang and Erik Dalsgaard
Appl. Opt. 34(32) 7527-7531 (1995)

Method for determining expansion centers and predicting sidelobe levels for circular-harmonic filters

Yunlong Sheng and Henri H. Arsenault
J. Opt. Soc. Am. A 4(9) 1793-1797 (1987)

Theory and optical implementation of the geometrical approach of multiple circular harmonic filters

Lixin Shen, Yunlong Sheng, and Guy Prémont
Appl. Opt. 34(20) 4004-4012 (1995)