(50-2) 12 * << * >> * Russian * English * Contents * All issues

On the possibilities of manipulating the shape of the maximum correlation function in the synthesis of filters to minimize the average correlation energy
D.V. Pavlenko1, R.S. Starikov1

1National Research Nuclear University "MEPhI", 31 Kashirskoe shosse, Moscow, 115409, Russia

  Full text (PDF)

DOI: 10.18287/COJ1720

Article ID: 1720

Abstract:
This article presents and discusses a modification of the MACE correlation filters that generates a cross-correlation function of a given shape. Numerical simulations of using these filters for binary classification of grayscale images with out-of-plane rotated "object of interest" are done. Five variants of the forming functions with different sizes were used: 32×32, 64×64, 128×2128, and 256×256 pixels. The size of the recognized images is 256×256 pixels. The simulation results demonstrate the possibility of achieving an invariant recognition quality of 99% using the studied filter modification and a special correlation metric.

Keywords:
image recognition, invariant correlation filter, minimization of average correlation energy, MACE.

Citation:
Pavlenko DV, Starikov RS. On the possibilities of manipulating the shape of the maximum correlation function in the synthesis of filters to minimize the average correlation energy. Computer Optics 2026; 50(2): 1720. DOI: 10.18287/COJ1720.

Acknowledgements: The work was funded by the Russian Science Foundation (RGNF) under grant No. 23-12-00336.

References:

  1. Vijaya Kumar BVK, Mahalanobis A, Juday R. Correlation pattern recognition. Cambridge, UK: Cambridge University Press; 2005. 390 p. ISBN 9780511541087. DOI: 10.1017/CBO9780511541087.
  2. Vijaya Kumar BVK. Tutorial survey of composite filter designs for optical correlators. Appl Opt 1992; 31: 4773-4790. DOI: 10.1364/AO.31.004773.
  3. Perie JN, ed. Advances in Digital Image Correlation (DIC). Basel: MDPI; 2020. 212 p. DOI: 10.3390/books978-3-03928-515-0.
  4. Starikov RS. Optical diffractive image correlators: A brief history and the newest achievements. Review. J Opt Technol 2025; 92(3): 144-161. DOI: 10.1364/JOT.92.000144.
  5. Kerekes R, Vijaya Kumar BVK. Selecting a composite correlation filter design: A survey and comparative study. Opt Eng 2008; 47: 067202. DOI: 10.1117/1.2943217.
  6. Vijaya Kumar BVK, Fernandez J, Rodriguez A, et al. Recent advances in correlation filter theory and application. Proc SPIE 2014; 9094: 909404. DOI: 10.1117/12.2051719.
  7. Mahalanobis A, Vijaya Kumar BVK, Casasent D. Minimum average correlation energy filters. Appl Opt 1987; 26: 3633-3640. DOI: 10.1364/AO.26.003633.
  8. Vijaya Kumar BVK, Mahalanobis A, Song S, Sims S, Epperson J. Minimum squared error synthetic discriminant functions. Opt Eng 1992; 31: 915-922. DOI: 10.1117/12.56169.
  9. Casasent D, Ravichandran G, Bollapragada S. Gaussian-minimum average correlation energy filters. Appl Opt 1991; 30: 5176-5185. DOI: 10.1364/AO.30.005176.
  10. Casasent D, Iyer A, Ravichandran G. Circular-harmonic function, minimum average correlation energy filters. Appl Opt 1991; 30: 5169-5175. DOI: 10.1364/AO.30.005169.
  11. Evtikhiev N, Shaulskiy D, Zlokazov E, Starikov R. Variants of minimum correlation energy filters: Comparative study. Proc SPIE 2012; 8398: 83980G. DOI: 10.1117/12.919644.
  12. Goyal S, Nishchal N, Beri V, Gupta A. Wavelet-modified maximum average correlation height filter for rotation invariance that uses chirp encoding in a hybrid digital-optical correlator. Appl Opt 2006; 45: 4850-4858. DOI: 10.1364/AO.45.004850.
  13. Ravichandran G, Casasent D. Minimum noise and correlation energy optical correlation filter. Appl Opt 1992; 31: 1823-1833. DOI: 10.1364/AO.31.001823.
  14. Vijaya Kumar BVK, Carlson D, Mahalanobis A. Optimal trade-off synthetic discriminant function filters for arbitrary devices. Opt Lett 1994; 19: 1556-1558. DOI: 10.1364/OL.19.001556.
  15. Patnaik R, Casasent D. Automated distortion-invariant filter synthesis and training set selection (auto-Minace). Proc SPIE 2006; 6245: 624507. DOI: 10.1117/12.673443.
  16. Casasent D, Patnaik R. Automated synthesis of distortion-invariant filters: AutoMinace. Proc SPIE 2006; 6384: 638401. DOI: 10.1117/12.693228.
  17. Figue J, Refregier P, Rajbenbach H, et al. Neural optoelectronic correlator for pattern recognition. Proc SPIE 1991; 1564: 550-560. DOI: 10.1117/12.49757.
  18. Dadeshidze V, Kompanets I, Lindunen I, et al. Adaptive neuromorphic optical correlator with binary phase-only filters. Proc SPIE 1994; 2051: 426-435. DOI: 10.1117/12.166047.
  19. Casasent D. An optical correlator feature extractor neural net system. Opt Eng 1992; 31: 971-980. DOI: 10.1117/12.57138.
  20. Lu T, Hughlett C, Zhou H, Chao TH, Hanan J. Neural network post-processing of grayscale optical correlator. Proc SPIE 2005; 5908: 590810. DOI: 10.1117/12.615573.
  21. Chao TH, Lu T. Grayscale optical correlator for CAD/CAC applications. Proc SPIE 2008; 6977: 697704. DOI: 10.1117/12.785873.
  22. Chao TH, Lu T. Automatic target recognition (ATR) performance improvement using integrated grayscale optical correlator and neural network. Proc SPIE 2009; 7340: 734003. DOI: 10.1117/12.820948.
  23. Chao TH, Lu T. Autonomous learning approach for automatic target recognition processor. Proc SPIE 2011; 8055: 805502. DOI: 10.1117/12.886145.
  24. Chen H, Lou S, Wang Q, Huang P, Duan H, Hu Y. Diffractive deep neural networks: Theories, optimization, and applications. Appl Phys Rev 2024; 11: 021332. DOI: 10.1063/5.0191977.
  25. Birch P, Akter H, Young R, Chatwin C. Coherent correlator design analysis for the implementation of deep learning networks. J Opt Soc Am A 2024; 41: 272-281. DOI: 10.1364/JOSAA.504956.
  26. Goncharov D, Starikov R. Improving correlation method with convolutional neural networks. arXiv preprint 2020; arXiv:2004.09430 [cs.CV].
  27. Goncharov D, Starikov R, Zlokazov E. Pattern recognition system based on a coherent diffractive correlator with deep learned processing of downsampled correlation responses. Appl Opt 2024; 63: 9196-9205. DOI: 10.1364/AO.541305.
  28. Goncharov DS, Zlokazov EYu, Petrova EK, Starikov RS. Processing the output signals of joint transform correlators using a pre-trained convolutional neural network. J Opt Technol 2025; 92(2): 120-128. DOI: 10.1364/JOT.92.000120.

151, Molodogvardeyskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20