Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
194 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Discrete Morphological Neural Networks (2309.00588v2)

Published 1 Sep 2023 in cs.CV

Abstract: A classical approach to designing binary image operators is Mathematical Morphology (MM). We propose the Discrete Morphological Neural Networks (DMNN) for binary image analysis to represent W-operators and estimate them via machine learning. A DMNN architecture, which is represented by a Morphological Computational Graph, is designed as in the classical heuristic design of morphological operators, in which the designer should combine a set of MM operators and Boolean operations based on prior information and theoretical knowledge. Then, once the architecture is fixed, instead of adjusting its parameters (i.e., structural elements or maximal intervals) by hand, we propose a lattice descent algorithm (LDA) to train these parameters based on a sample of input and output images under the usual machine learning approach. We also propose a stochastic version of the LDA that is more efficient, is scalable and can obtain small error in practical problems. The class represented by a DMNN can be quite general or specialized according to expected properties of the target operator, i.e., prior information, and the semantic expressed by algebraic properties of classes of operators is a differential relative to other methods. The main contribution of this paper is the merger of the two main paradigms for designing morphological operators: classical heuristic design and automatic design via machine learning. As a proof-of-concept, we apply the DMNN to recognize the boundary of digits with noise, and we discuss many topics for future research.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (69)
  1. Jesus Angulo. Some open questions on morphological operators and representations in the deep learning era: A personal vision. In Discrete Geometry and Mathematical Morphology: First International Joint Conference, DGMM 2021, Uppsala, Sweden, May 24–27, 2021, Proceedings, pages 3–19. Springer, 2021.
  2. Mathematical morphology in color spaces applied to the analysis of cartographic images. Proceedings of GEOPRO, 3:59–66, 2003.
  3. Modular morphological neural network training via adaptive genetic algorithm for designing translation invariant operators. In 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, volume 2, pages II–II. IEEE, 2006.
  4. A morphological neural network for binary classification problems. Engineering Applications of Artificial Intelligence, 65:12–28, 2017.
  5. Differential evolution training algorithm for dendrite morphological neural networks. Applied Soft Computing, 68:303–313, 2018.
  6. A fast branch-and-bound algorithm for U-curve feature selection. Pattern Recognition, 73:172–188, 2018.
  7. Gerald Jean Francis Banon and Junior Barrera. Minimal representations for translation-invariant set mappings by mathematical morphology. SIAM Journal on Applied Mathematics, 51(6):1782–1798, 1991.
  8. Gerald Jean Francis Banon and Junior Barrera. Decomposition of mappings between complete lattices by mathematical morphology, Part I. general lattices. Signal Processing, 30(3):299–327, 1993.
  9. MMach: a mathematical morphology toolbox for the Khoros system. Journal of Electronic Imaging, 7(1):174–210, 1998.
  10. Automatic programming of binary morphological machines by design of statistically optimal operators in the context of computational learning theory. Journal of Electronic Imaging, 6(1):54–67, 1997.
  11. From mathematical morphology to machine learning of image operators. São Paulo Journal of Mathematical Sciences, 16(1):616–657, 2022.
  12. Set operations on closed intervals and their applications to the automatic programming of morphological machines. Journal of Electronic Imaging, 5(3):335–352, 1996.
  13. Automatic programming of morphological machines by PAC learning. Fundamenta Informaticae, 41(1-2):229–258, 2000.
  14. Serge Beucher. Watersheds of functions and picture segmentation. In ICASSP’82. IEEE International Conference on Acoustics, Speech, and Signal Processing, volume 7, pages 1928–1931. IEEE, 1982.
  15. Design of optimal binary filters under joint multiresolution–envelope constraint. Pattern recognition letters, 24(7):937–945, 2003.
  16. Nonlinear filter design using envelopes. Journal of Mathematical Imaging and Vision, 21:81–97, 2004.
  17. Jon Clayden. mmand: Mathematical Morphology in Any Number of Dimensions, 2023. R package version 1.6.3.
  18. Jennifer L Davidson. Simulated annealing and morphology neural networks. In Image Algebra and Morphological Image Processing III, volume 1769, pages 119–127. SPIE, 1992.
  19. Morphology neural networks: An introduction with applications. Circuits, Systems and Signal Processing, 12(2):177–210, 1993.
  20. Theory of morphological neural networks. In Digital Optical Computing II, volume 1215, pages 378–388. SPIE, 1990.
  21. Designing translation invariant operations via neural network training. In Proceedings 2000 International Conference on Image Processing (Cat. No. 00CH37101), volume 1, pages 908–911. IEEE, 2000.
  22. An exact algorithm for optimal MAE stack filter design. IEEE transactions on image processing, 16(2):453–462, 2007.
  23. Advances in morphological neural networks: training, pruning and enforcing shape constraints. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 3825–3829. IEEE, 2021.
  24. Hands-on morphological image processing, volume 59. SPIE press, 2003.
  25. Neural architecture search: A survey. The Journal of Machine Learning Research, 20(1):1997–2017, 2019.
  26. An efficient, parallelized algorithm for optimal conditional entropy-based feature selection. Entropy, 22(4):492, 2020.
  27. Deep morphological networks. Pattern Recognition, 102:107246, 2020.
  28. Some applications of morphological neural networks. In IJCNN’01. International Joint Conference on Neural Networks. Proceedings (Cat. No. 01CH37222), volume 4, pages 2518–2523. IEEE, 2001.
  29. Morphpool: Efficient non-linear pooling & unpooling in CNNs. arXiv preprint arXiv:2211.14037, 2022.
  30. Geometric back-propagation in morphological neural networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023.
  31. AutoML: A survey of the state-of-the-art. Knowledge-Based Systems, 212:106622, 2021.
  32. Nina ST Hirata. Multilevel training of binary morphological operators. IEEE Transactions on pattern analysis and machine intelligence, 31(4):707–720, 2008.
  33. Design of statistically optimal stack filters. In XII Brazilian Symposium on Computer Graphics and Image Processing (Cat. No. PR00481), pages 265–274. IEEE, 1999.
  34. On machine-learning morphological image operators. Mathematics, 9(16):1854, 2021.
  35. The incremental splitting of intervals algorithm for the design of binary image operators. Proceedings of the 6th ISMM, pages 219–228, 2002.
  36. Iterative design of morphological binary image operators. Optical Engineering, 39(12):3106–3123, 2000.
  37. Segmentation of microarray images by mathematical morphology. Real-Time Imaging, 8(6):491–505, 2002.
  38. Aperture filters. Signal processing, 80(4):697–721, 2000.
  39. Learning deep morphological networks with neural architecture search. Pattern Recognition, 131:108893, 2022.
  40. Image operator learning coupled with CNN classification and its application to staff line removal. In 2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR), volume 1, pages 53–58. IEEE, 2017.
  41. The lattice overparametrization paradigm for the machine learning of lattice operators. arXiv preprint arXiv:2310.06639, 2023.
  42. An algorithm to train unrestricted sequential discrete morphological neural networks. arXiv preprint arXiv:2310.04584, 2023.
  43. Georges Matheron. Random sets and integral geometry. John Wiley & Sons, 1974.
  44. Georges Matheron. Random sets and integral geometry. John Wiley & Sons, 1975.
  45. Morphological networks for image de-raining. In Discrete Geometry for Computer Imagery: 21st IAPR International Conference, DGCI 2019, Marne-la-Vallée, France, March 26–28, 2019, Proceedings 21, pages 262–275. Springer, 2019.
  46. Morphological network: How far can we go with morphological neurons? arXiv preprint arXiv:1901.00109, 2019.
  47. Staff removal using image operator learning. Pattern Recognition, 63:310–320, 2017.
  48. Kernel approximations for W-operator learning. In 2016 29th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), pages 386–393. IEEE, 2016.
  49. Alexandre Monteiro da Silva and Peter Sussner. A brief review and comparison of feedforward morphological neural networks with applications to classification. In Artificial Neural Networks-ICANN 2008: 18th International Conference, Prague, Czech Republic, September 3-6, 2008, Proceedings, Part II 18, pages 783–792. Springer, 2008.
  50. New windows based color morphological operators for biomedical image processing. In Journal of Physics: Conference Series, volume 705, page 012023. IOP Publishing, 2016.
  51. Morphological/rank neural networks and their adaptive optimal design for image processing. In 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings, volume 6, pages 3398–3401. IEEE, 1996.
  52. Neural networks with hybrid morphological/rank/linear nodes: a unifying framework with applications to handwritten character recognition. Pattern Recognition, 33(6):945–960, 2000.
  53. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2021.
  54. Optimal Boolean lattice-based algorithms for the U-curve optimization problem. Information Sciences, 2018.
  55. U-curve: A branch-and-bound optimization algorithm for U-shaped cost functions on Boolean lattices applied to the feature selection problem. Pattern Recognition, 43(3):557–568, 2010.
  56. Morphological perceptrons with dendritic structure. In The 12th IEEE International Conference on Fuzzy Systems, 2003. FUZZ’03., volume 2, pages 1296–1301. IEEE, 2003.
  57. An introduction to morphological neural networks. In Proceedings of 13th International Conference on Pattern Recognition, volume 4, pages 709–717. IEEE, 1996.
  58. Sebastian Ruder. An overview of gradient descent optimization algorithms. arXiv preprint arXiv:1609.04747, 2016.
  59. An information theory framework for two-stage binary image operator design. Pattern Recognition Letters, 31(4):297–306, 2010.
  60. M Schmitt. Mathematical morphology and artificial intelligence: an automatic programming system. Signal Processing, 16(4):389–401, 1989.
  61. Jean Serra. Image Analysis and Mathematical Morphology. Volume 1. Academic Press, Inc., 1984.
  62. Jean Serra. Image Analysis and Mathematical Morphology. Volume 2: Theoretical Advances. Academic Press, Inc., 1988.
  63. An overview of morphological filtering. Circuits, Systems and Signal Processing, 11:47–108, 1992.
  64. Efficient training for dendrite morphological neural networks. Neurocomputing, 131:132–142, 2014.
  65. Order space-based morphology for color image processing. Journal of Imaging, 9(7):139, 2023.
  66. Constructive morphological neural networks: some theoretical aspects and experimental results in classification. Constructive neural networks, pages 123–144, 2009.
  67. Fast computation of morphological operations with arbitrary structuring elements. Pattern recognition letters, 17(14):1451–1460, 1996.
  68. Vladimir Vapnik. Statistical learning theory. 1998, volume 3. Wiley, New York, 1998.
  69. Luc M Vincent. Efficient computation of various types of skeletons. In Medical Imaging V: Image Processing, volume 1445, pages 297–311. SPIE, 1991.
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now