Abstract

We report the results of experimental tests of an optical-correlator-based automatic target recognition (ATR) system that uses the correlation-peak moment analysis technique of Caprari [Appl. Opt. 38, 1317 (1999)] to assist in discrimination between target and clutter peaks. The ATR system and its operation are briefly described with particular attention devoted to a practical scheme for enhancing the basic ATR system with correlation-peak moment detectors. We investigate the variation of detection and false-alarm rates of moment detectors with moment threshold values. For fixed moment thresholds, we present receiver operating characteristics of both basic and enhanced systems as the conventionally used correlation-peak energy threshold is varied. Results demonstrate that correlation-peak moment analysis materially improves ATR system target-detection performance.

© 1999 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. D. Gregory, J. Kirsch, J. Loudin, “Optical correlators: optical computing that really works,” in Advances in Optical Information Processing IV, D. Pape, ed., Proc. SPIE1296, 2–19 (1990).
    [CrossRef]
  2. D. Casasent, “General-purpose optical pattern recognition image processors,” Proc. IEEE 82, 1724–1734 (1994).
    [CrossRef]
  3. Y. Ichioka, T. Iwaki, K. Matsuoka, “Optical information processing and beyond,” Proc. IEEE 84, 694–719 (1996).
    [CrossRef]
  4. F. Yu, D. Gregory, “Optical pattern recognition: architectures and techniques,” Proc. IEEE 84, 733–752 (1996).
    [CrossRef]
  5. P. Miller, M. Royce, P. Virgo, M. Fiebig, G. Hamlyn are preparing the following paper for publication: “Evaluation of an optical correlator automatic target recognition system for acquisition and tracking in densely cluttered imagery of natural scenes.”
  6. S. Lindell, “Summary of the transfer of optical processing to systems: optical pattern recognition program,” in Transition of Optical Processors into Systems 1995, D. Casasent, ed., Proc. SPIE2489, 20–34 (1995).
    [CrossRef]
  7. R. Caprari, “Method of target detection in images by moment analysis of correlation peaks,” Appl. Opt. 38, 1317–1324 (1999).
    [CrossRef]
  8. D. Jared, D. Ennis, “Inclusion of filter modulation in synthetic-discriminant-function construction,” Appl. Opt. 28, 232–239 (1989).
    [CrossRef] [PubMed]
  9. J. Egan, Signal Detection Theory and ROC Analysis (Academic, New York, 1975).

1999

1996

Y. Ichioka, T. Iwaki, K. Matsuoka, “Optical information processing and beyond,” Proc. IEEE 84, 694–719 (1996).
[CrossRef]

F. Yu, D. Gregory, “Optical pattern recognition: architectures and techniques,” Proc. IEEE 84, 733–752 (1996).
[CrossRef]

1994

D. Casasent, “General-purpose optical pattern recognition image processors,” Proc. IEEE 82, 1724–1734 (1994).
[CrossRef]

1989

Caprari, R.

Casasent, D.

D. Casasent, “General-purpose optical pattern recognition image processors,” Proc. IEEE 82, 1724–1734 (1994).
[CrossRef]

Egan, J.

J. Egan, Signal Detection Theory and ROC Analysis (Academic, New York, 1975).

Ennis, D.

Fiebig, M.

P. Miller, M. Royce, P. Virgo, M. Fiebig, G. Hamlyn are preparing the following paper for publication: “Evaluation of an optical correlator automatic target recognition system for acquisition and tracking in densely cluttered imagery of natural scenes.”

Gregory, D.

F. Yu, D. Gregory, “Optical pattern recognition: architectures and techniques,” Proc. IEEE 84, 733–752 (1996).
[CrossRef]

D. Gregory, J. Kirsch, J. Loudin, “Optical correlators: optical computing that really works,” in Advances in Optical Information Processing IV, D. Pape, ed., Proc. SPIE1296, 2–19 (1990).
[CrossRef]

Hamlyn, G.

P. Miller, M. Royce, P. Virgo, M. Fiebig, G. Hamlyn are preparing the following paper for publication: “Evaluation of an optical correlator automatic target recognition system for acquisition and tracking in densely cluttered imagery of natural scenes.”

Ichioka, Y.

Y. Ichioka, T. Iwaki, K. Matsuoka, “Optical information processing and beyond,” Proc. IEEE 84, 694–719 (1996).
[CrossRef]

Iwaki, T.

Y. Ichioka, T. Iwaki, K. Matsuoka, “Optical information processing and beyond,” Proc. IEEE 84, 694–719 (1996).
[CrossRef]

Jared, D.

Kirsch, J.

D. Gregory, J. Kirsch, J. Loudin, “Optical correlators: optical computing that really works,” in Advances in Optical Information Processing IV, D. Pape, ed., Proc. SPIE1296, 2–19 (1990).
[CrossRef]

Lindell, S.

S. Lindell, “Summary of the transfer of optical processing to systems: optical pattern recognition program,” in Transition of Optical Processors into Systems 1995, D. Casasent, ed., Proc. SPIE2489, 20–34 (1995).
[CrossRef]

Loudin, J.

D. Gregory, J. Kirsch, J. Loudin, “Optical correlators: optical computing that really works,” in Advances in Optical Information Processing IV, D. Pape, ed., Proc. SPIE1296, 2–19 (1990).
[CrossRef]

Matsuoka, K.

Y. Ichioka, T. Iwaki, K. Matsuoka, “Optical information processing and beyond,” Proc. IEEE 84, 694–719 (1996).
[CrossRef]

Miller, P.

P. Miller, M. Royce, P. Virgo, M. Fiebig, G. Hamlyn are preparing the following paper for publication: “Evaluation of an optical correlator automatic target recognition system for acquisition and tracking in densely cluttered imagery of natural scenes.”

Royce, M.

P. Miller, M. Royce, P. Virgo, M. Fiebig, G. Hamlyn are preparing the following paper for publication: “Evaluation of an optical correlator automatic target recognition system for acquisition and tracking in densely cluttered imagery of natural scenes.”

Virgo, P.

P. Miller, M. Royce, P. Virgo, M. Fiebig, G. Hamlyn are preparing the following paper for publication: “Evaluation of an optical correlator automatic target recognition system for acquisition and tracking in densely cluttered imagery of natural scenes.”

Yu, F.

F. Yu, D. Gregory, “Optical pattern recognition: architectures and techniques,” Proc. IEEE 84, 733–752 (1996).
[CrossRef]

Appl. Opt.

Proc. IEEE

D. Casasent, “General-purpose optical pattern recognition image processors,” Proc. IEEE 82, 1724–1734 (1994).
[CrossRef]

Y. Ichioka, T. Iwaki, K. Matsuoka, “Optical information processing and beyond,” Proc. IEEE 84, 694–719 (1996).
[CrossRef]

F. Yu, D. Gregory, “Optical pattern recognition: architectures and techniques,” Proc. IEEE 84, 733–752 (1996).
[CrossRef]

Other

P. Miller, M. Royce, P. Virgo, M. Fiebig, G. Hamlyn are preparing the following paper for publication: “Evaluation of an optical correlator automatic target recognition system for acquisition and tracking in densely cluttered imagery of natural scenes.”

S. Lindell, “Summary of the transfer of optical processing to systems: optical pattern recognition program,” in Transition of Optical Processors into Systems 1995, D. Casasent, ed., Proc. SPIE2489, 20–34 (1995).
[CrossRef]

J. Egan, Signal Detection Theory and ROC Analysis (Academic, New York, 1975).

D. Gregory, J. Kirsch, J. Loudin, “Optical correlators: optical computing that really works,” in Advances in Optical Information Processing IV, D. Pape, ed., Proc. SPIE1296, 2–19 (1990).
[CrossRef]

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 (4)

Fig. 1
Fig. 1

Block diagram for the basic ATR system (within the dashed box) and for the enhanced ATR system with a moment detector.

Fig. 2
Fig. 2

(a) Sample image of the test sequence. (b) Output of the preprocessor with the bottom right-hand corner of (a) as the input. (c) Correlation-plane output with the image of (b) as the input.

Fig. 3
Fig. 3

Plots of P cc and P ic versus T for (a) angle, (b) eccentricity, (c) trace, and (d) composite moment detectors. P cc is plotted for when the system is in search mode or is tracking the target; P ic is plotted for when the system is tracking false alarms. T labels each setting of the moment-detector thresholds.

Fig. 4
Fig. 4

Comparison of the ROC of the basic system with that of the enhanced system by use of (a) angle, (b) eccentricity, (c) trace, and (d) composite moment detectors. Only the PE threshold varies along the ROC’s.

Equations (2)

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

Pcc=NccN,  Pic=NicN,
Pr=NTNV,  Pfa=NFAN-NT,

Metrics