Abstract

A new matched filter for pattern recognition is introduced. Previous researchers have introduced matched filters that are nonlinear functions of the spectrum, for which the classical matched filter is divided by some power m of the spectrum of interest. We further generalize this filter by making the power m a function of the spatial frequency; this permits the design of filters that combine the advantageous properties of matched filters, phase-only filters, and inverse filters without the corresponding disadvantages. The first computer experiments indicate that the new filter yields sharper correlation peaks and better discrimination than the matched filter and the phase-only filter and yields more robustness against noise.

© 1993 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. J. Horner, P. D. Gianino, Appl. Opt. 23, 812 (1984).
    [CrossRef] [PubMed]
  2. G. Ravichandran, D. Casasent, Appl. Opt. 31, 1823 (1992).
    [CrossRef] [PubMed]
  3. P. Refregier, Opt. Lett. 15, 854 (1990).
    [CrossRef] [PubMed]
  4. O. K. Ersoy, M. Zeng, J. Opt. Soc. Am. A 6, 636 (1989).
    [CrossRef]
  5. B. Javidi, Appl. Opt. 29, 1215 (1990).
    [CrossRef] [PubMed]
  6. B. V. K. Vijaya Kumar, L. Hassebrook, Appl. Opt. 28, 3112 (1989).
    [CrossRef]

1992 (1)

1990 (2)

1989 (2)

1984 (1)

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (3)

Fig. 1
Fig. 1

Correlation outputs obtained with the space shuttle image as input and (a) a matched filter, (b) a POF, and (c) a NLMF.

Fig. 2
Fig. 2

Correlation intensity outputs for (a) a POF and (b) a LNMF correlated with a noisy image of the space shuttle.

Fig. 3
Fig. 3

Input scene for the discrimination test; the space shuttle (top-left image) is the target. Images are numbered from left to right and top to bottom.

Tables (1)

Tables Icon

Table 1 Normalized Correlation Intensities for the Matched Filter, the POF, and the LNMF

Equations (3)

Equations on this page are rendered with MathJax. Learn more.

F ( u , υ ) | F ( u , υ ) | m ,
m = f ( ρ ) , ρ u 2 + υ 2 ;
f ( ρ ) = 2 exp ( k ρ 2 ) ,

Metrics