<div>
	Classification of objects in images with distortions based on a two-stage topological analysis</div>

Sergey V. Eremeev, Artyom V. Abakumov

2022 , VOLUME 22, NUMBER 1 ( january-february )

ISSN 2226-1494 (print), ISSN 2500-0373 (online)

Publications

Editor-in-Chief

Nikiforov
Vladimir O.
D.Sc., Prof.

Partners

doi: 10.17586/2226-1494-2022-22-1-82-92

Classification of objects in images with distortions based on a two-stage topological analysis

S. V. Eremeev, A. V. Abakumov

Read the full article

Article in Russian

For citation:

Eremeev S.V., Abakumov A.V. Classification of objects in images with distortions based on a two-stage topological analysis. Scientific and Technical Journal of Information Technologies, Mechanics and Optics, 2022, vol. 22, no. 1, pp. 82–92 (in Russian). doi: 10.17586/2226-1494-2022-22-1-82-92

Abstract

The authors propose a method for automatic classification of spatial objects in images under conditions of a limited data set. The stability of the method to distortions appearing in images due to natural phenomena and partial overlap of urban infrastructure objects is investigated. High classification accuracy, when using existing approaches, requires a large training sample, including data sets with distortions, which significantly increases computational complexity. The paper proposes a method for a two-step topological analysis of images. Topological features are initially extracted by analyzing the image in the brightness range from 0 to 255, and then from 255 to 0. These features complement each other and reflect the topological structure of the object. Under certain deformations and distortions, the object preserves its structure in the form of extracted features. The advantage of the method is a small number of patterns, which reduces the computational complexity of training compared to neural networks. The proposed method is investigated and compared with the modern neural network approach. The study was performed on a DOTA dataset (Dataset for Object deTection in Aerial images) containing images of spatial objects of several classes. In the absence of distortion in the image, the neural network approach showed a classification accuracy of over 98 %, while the proposed method achieved about 82 %. Further distortions such as 90 degree rotation, 50 % narrowing and 50 % edge truncation and their combinations were applied in the experiment. The proposed method showed its robustness and outperformed the neural network approach. In the most difficult combination of the test, the decrease in classification accuracy of the neural network was 46 %, while the described method showed 12 %. The proposed method can be applied in cases with a high probability of distortion in the images. Such distortions arise in the field of geoinformatics when analyzing objects of various scales, under different weather conditions, partial overlap of one object with another, in the presence of shadows, etc. It is possible to use the proposed method in vision systems of industrial enterprises for automatic classification of the parts that belong to superimposed objects.

Keywords: topological analysis, persistent homology, image distortion, object classification, neural networks

Acknowledgements. The reported study was funded by the YSU Programme (the research project No. П2-К-1-Г-3/2021).

References

1. Yarashevich P.V., Bohush R.P. Classification algorithm of parking space images based on a histogram of oriented gradients and support vector machines. Computer Optics, 2017, vol. 41, no 1, pp. 110–117. https://doi.org/10.18287/2412-6179-2017-41-1-110-117. (in Russian)

2. Krasnov F.V., Butorin A.V., Sitnikov A.N. Review of approaches to the analysis of high-resolution spatial images for Geophysics. Cloud of Science, 2019, vol. 6, no. 1, pp. 127–143. (in Russian)

3. Fu J., Rui Y. Advances in deep learning approaches for image tagging. APSIPA Transactions on Signal and Information Processing, 2017, vol. 6, pp. e11. https://doi.org/10.1017/ATSIP.2017.12

4. Yashchenko A.V., Belikov A.V., Peterson M.V., Potapov A.S. Distillation of neural network models for detection and description of image key points. Scientific and Technical Journal of Information Technologies, Mechanics and Optics, 2020, vol. 20, no. 3, pp. 402–409. (in Russian). https://doi.org/10.17586/2226-1494-2020-20-3-402-409

5. Blyumin S., Pogodaev A., Khabibullina E. Graph-structural modeling of some special organizational systems. Proc. 2nd International Conference on Control Systems, Mathematical Modeling, Automation and Energy Efficiency (SUMMA), 2020, pp. 279–283. https://doi.org/10.1109/SUMMA50634.2020.9280724

6. Handrich S. Al-Hamadi A. Localizing body joints from single depth images using geodetic distances and random tree walk. Proc. 24th IEEE International Conference on Image Processing (ICIP), 2017, pp. 146–150. https://doi.org/10.1109/ICIP.2017.8296260

7. Karimova L., Terekhov A., Makarenko N., Rybintsev A. Methods of computational topology and discrete Riemannian geometry for the analysis of arid territories. Cogent Engineering, 2020, vol. 7, no. 1, pp. 1808340. https://doi.org/10.1080/23311916.2020.1808340

8. Ojala T., Pietikäinen M., Mäenpää T. Multiresolution grayscale and rotation invariant texture classification with local binary patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, vol. 24, no. 7, pp. 971–987. https://doi.org/10.1109/TPAMI.2002.1017623

9. Myasnikov VV. Reconstruction of functions and digital images using sign representations. Computer Optics, 2019, vol. 43, no. 6, pp. 1041–1052. (in Russian). https://doi.org/10.18287/2412-6179-2019-43-6-1041-1052

10. Luo Z., Chen J., Takiguchi T., Ariki Y. Neutral-to-emotional voice conversion with cross-wavelet transform F0 using generative adversarial networks. APSIPA Transactions on Signal and Information Processing, 2019, vol. 8, pp. 1–11. https://doi.org/10.1017/ATSIP.2019.3

11. Ruckdeschel P., Kohl M. General Purpose Convolution Algorithm in S4 Classes by Means of FFT. Journal of Statistical Software, 2014, vol. 59, no. 4, pp. 1–25. https://doi.org/10.18637/jss.v059.i04

12. Kasimov D.R. Searching and describing objects in satellite images on the basis of modeling reasoning. Computer Optics, 2020, vol. 44, no. 5, pp. 772–781. (in Russian). https://doi.org/10.18287/2412-6179-CO-716

13. Sahu S.K., Pujari A.K., Kagita V.R., Kumar V., Padmanabhan V. GP-SVM: Tree structured multiclass SVM with greedy partitioning. International Conference on Information Technology (ICIT), 2015, pp. 142–147. https://doi.org/10.1109/ICIT.2015.24

14. Zhou Q., Lan W., Zhou Y., Mo G. Effectiveness evaluation of anti-bird devices based on random forest algorithm. Proc. of the 7th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS), 2020, pp. 743–748, https://doi.org/10.1109/ICCSS52145.2020.9336891

15. Xu H., Chen Y., Lin R., Kuo J. Understanding convolutional neural networks via discriminant feature analysis. APSIPA Transactions on Signal and Information Processing, 2018, vol. 7, pp. e20. https://doi.org/10.1017/ATSIP.2018.24

16. Ryumina E.V., Karpov A.A. Analytical review of methods for emotion recognition by human face expressions. Scientific and Technical Journal of Information Technologies, Mechanics and Optics, 2020, vol. 20, no. 2, pp. 163–176. (in Russian). https://doi.org/10.17586/2226-1494-2020-20-2-163-176

17. Edelsbrunner H., Letscher H., Zomorodian A. Topological persistence and simplification. Discrete and Computational Geometry, 2002, vol. 28, no. 4, pp. 511–533. https://doi.org/10.1007/s00454-002-2885-2

18. Carlsson E., Carlsson G., de Silva V. An algebraic topological method for feature identification. International Journal of Computational Geometry and Applications, 2006, vol. 16, no. 4, pp. 291–314. https://doi.org/10.1142/S021819590600204X

19. Makarenko N.D, Urtiev F.A, Knyazeva I.S, Malkova D., Pak I.T, Karimova L.M. Texture recognition in digital images by computational topology methods. Sovremennye Problemy Distantsionnogo Zondirovaniya Zemli iz Kosmosa, 2015, vol. 12, no. 1, pp. 131–144. (in Russian)

20. Gonzalez-Diaz R., Jimenez M.-J., Medrano B. Spatiotemporal barcodes for image sequence analysis. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2015, vol. 9448, pp. 61–70. https://doi.org/10.1007/978-3-319-26145-4_5

21. Eremeev S.V., Andrianov D.E., Titov V.S. An algorithm for matching spatial objects of different-scale maps based on topological data analysis. Computer Optics, 2019, vol. 43, no. 6, pp. 1021–1029. (in Russian). https://doi.org/10.18287/2412-6179-2019-43-6-1021-1029

22. Eremeev S.V., Abakumov A.V. Software complex for detection and classification of natural objects based on topological analysis. Software & Systems, 2021, vol. 34, no. 1, pp. 201–208. (in Russian). https://doi.org/10.15827/0236-235X.133.201-208

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License