Purpose. The goal of this work was to develop and study the methods of texture classification of SEM images of micro surfaces of objects based on the statistical and spectral characteristics of texture fragments, as well as a comparative analysis of segmentation methods of SEM images. Methods. The determination of the texture characteristics was based on statistical moments computed by the brightness histogram of a SEM- image or its region. The spectral measures of texture of SEM image were based on properties of the Fourier spectrum. To determine the spectral texture characteristics, the parameters of the amplitude and axial functions were chosen. SEM images were segmented using four methods, namely: the global thresholding; the region growing; the region splitting and merging; and the watershed using markers. Results. The experiments on texture classification of the SEM series of soils and metals images showed the best result of texture classification by the feature of homogeneity compared to other statistical characteristics. Calculation of the spectral characteristics was used to detect the directionality of periodic or almost periodic texture elements in the SEM images of metals. Classification results using spectral properties and homogeneity values made it possible to obtain generalized texture characteristics of SEM images of metals. A comparative analysis of the four segmentation methods showed that the best result of finding the boundaries of objects in the SEM image was obtained by the watershed method using markers. Software implementation of texture classification and image segmentation methods were performed in the MatLab system. Scientific novelty. The authors proposed a method for classifying SEM-images based on spectral texture characteristics using the parameters of the amplitude and axial functions. It is shown that the segmentation by the splitting and merging method allows you to set the conditions for selecting regions with certain texture characteristics in the SEM-image. The practical significance. A generalized characteristic of SEM-image texture, determined by statistical and spectral measurements, is that it would be useful for automatic texture recognition and SEM-images analysis. The selection of regions with certain texture characteristics is the preprocessing step for finding points of interest suitable for the SEM-image matching and objects recognition.
- 1. Al-Janabi Akil Bahr Tarkhan, Shuvalova, L. A. (2014). Image enhancement using the watershed algorithm], Information Processing Systems, 8 (124), 3-7.
2. Asatryan, D. G., Kurkchiyan, V. V., & Kharatyan, L. R. (2014). Method for texture classification using image structural features, Computer Optics, 38(3), 574-579.
https://doi.org/10.18287/0134-2452-2014-38-3-574-579
3. Bagalkote, I. S., & Vibhute, A. S. (2016). Review on: Texture Discrimination Feature Analysis for Visually Similar Texture of Different Fields. International Journal for Scientific Research & Development, 3(9), 851-856.
4. Bavrina, A. Ju., Ilyasova, N. Ju. (2002). Investigation of photogrammetric images using brightness distribution probality matrices, Computer Optics, 23, 62-65.
5. Bhosle, V. V., Pawar, V. P. (2013). Texture Segmentation: Different Methods. International Journal of Soft Computing and Engineering (IJSCE) ISSN: 2231-2307, 3(5), 69-74.
6. Bogucharsky, S. I. Mashtalir, S. V. (2014). Image sequences texture analysis based on vector quantization, Radio Electronics, Computer Science, Control, 2, 94-99.
https://doi.org/10.15588/1607-3274-2014-2-14
7. Cavalin, Paulo & Soares de Oliveira, Luiz. (2017). A Review of Texture Classification Methods and Databases. 1-8. 10.1109/SIBGRAPI-T.2017.10.
https://doi.org/10.1109/SIBGRAPI-T.2017.10
8. Cord, A., Bach, F., Jeulin, D. Texture classification by statistical learning from morphological image processing: application to metallic surfaces. J. Microsc. 239(2), 159-166 (2010).
https://doi.org/10.1111/j.1365-2818.2010.03365.x
9. Fisenko, V. T., Fisenko, T. Yu. (2013). Fractal methods of texture image segmentation], Instrument making, 56(5), 63-70.
10. Forsyth, D., Pons, J. Computer Vision. The Modern Approach, Moscow: Williams Publishing House, 2004, 928 p.
11. Golduyeva, D. A., Mokshanina, M. A. (2015). Trace transformation as the basis of a method for segmenting halftone textures. Models, systems, networks in economics, technology, nature and society, 3 (15), 128-136.
12. Gonzalez, R. C., & Woods, R. E. (2005). Book on "Digital image processing". Prentice-Hall of India Pvt. Ltd.
13. Gonzalez, R. C., Woods, R. E., & Eddins, S. L. (2006). Ruan Qiuqi (Translation). Digital Image Processing Using MATLAB. Publishing House of Electronics Industry, 76-77.
14. Gonzales-Barron, U. A. & Butler, F. (2006). Statistical and spectral texture analysis methods for discrimination of bread crumb images. pp. 749-759. DOI: 10.1051/IUFoST:20060164.
https://doi.org/10.1051/IUFoST:20060164
15. Gray, A, & Marshall S. & McKenzie, J. (2006). Modeling of evolving textures using granulometries. Chapter in Eurasip Book Series on Signal Processing and Communications, pp. 240-270.
16. Gulakov, V. K., Trubakov, A. O., S. N. Ogurtsov, S. N. (2011). Informative significance of texture characteristics based on the adjacency matrix of brightness levels of image pixels, Bulletin of the Bryansk State Technical University. 2(30), 81-85.
17. Haindl, M., Mikeš, S. (2016). Unsupervised Texture Segmentation. In Pattern Recognition Techniques, Technology and Applications by Peng-Yeng Yin (eds), Chap. 9. 227-248
18. Haralick, R. M. (1979). Statistical and structural approach to textures, Proceedings IEEE, 67, 786-804.
https://doi.org/10.1109/PROC.1979.11328
19. Hu, X. (2017). Frequency Based Texture Featurte Descriptors. 147 p.
20. Ivanchuk, O., Tumska, O. (2017). Comparative analysis of the statistical and scaling characteristics of SEM images, obtained on different types of SEM, Recent advances in geodetic science and industry, Lviv, II (34), 119-131.
21. Khokhlov, M., Fischer, A., Rittel, D. (2012). Multi-Scale Stereo-Photogrammetry System for Fractographic. Analysis Using Scanning Electron Microscopy. Experimental Mechanics. 52, 975-991. DOI 10.1007/s11340-011-9582-0.
https://doi.org/10.1007/s11340-011-9582-0
22. Kolodnikova, N. V. (2004). A review of texture features for pattern recognition tasks. Proceedings of the employees of TUSUR. Tomsk, 113-124.
23. Kupriyanov, A. V. (2008). Segmentation of texture images based on the evaluation of local statistical signs, Bulletin of the Samara State Aerospace University, 2, 245-251.
24. Lee, J. H., Yoo, S. I. An effective image segmentation technique for the SEM image. Conference Paper May 2008, pp. 1-6. DOI: 10.1109/ICIT.2008.4608647, Source: IEEE Xplore.
https://doi.org/10.1109/ICIT.2008.4608647
25. Liu, X. and Wang, D. L. Image and Texture Segmentation Using Local Spectral Histograms. IEEE transactions on image processing, 15(10), 3066-3077.
https://doi.org/10.1109/TIP.2006.877511
26. Lu, D. A., & Weng, Q. (2007). A survey of image classification methods and techniques for improving classification performance. International Journal of Remote Sensing, 28(5), 823-870.
https://doi.org/10.1080/01431160600746456
27. Madasu, V. K., Yarlagadda, P. (2007). An in depth comparison of four texture segmentation methods. Digital Image Computing Techniques and Applications. IEEE. pp. 366-372. DOI 10.1109/DICTA.2007.83
https://doi.org/10.1109/DICTA.2007.4426820
28. Manjunath, B. S., Haley, G. M., Ma Wei-Ying, Newsam, S. D. (2005). Multiband Techniques for Texture Classification and Segmentation. Chap. 4.9 in Handbook of Image and Video Processing. (Second Edition) by Bovik Al. (eds), pp. 455-470. Academic Press., 2005.
https://doi.org/10.1016/B978-012119792-6/50091-7
29. Materka, A., Strzelecki, M. (1998). Texture Analysis Methods - A Review. Technical University of Lodz, Institute of Electronics, COST B11 report, Brussels, 10(1.97), 4968.
30. Melnik, V. M., Shostak, A. V. (2009). Raster electron stereomicrofraktografition, Luck, Vezha, 469 p.
31. Melnik, G. M. (2012). Information technology of analysis and synthesis of structural textures in automated systems for processing histological images, Thesis for a Candidate Degree in Engineering, Ternopil, 27 p.
32. Neogi, N., Mohanta, D. K., & Dutta, P. K. (2014). Review of vision-based steel surface inspection systems. EURASIP Journal on Image and Video Processing, 2014(1), 50. doi:10.1186/1687-5281-2014-50
https://doi.org/10.1186/1687-5281-2014-50
33. Noman, A. A., Khorkov, K. S., Shamin, P. Yu. (2014). Research Methods of Semiconductor Heterostructures: Textbook. Allowance, Vladimir: Publishing House of VlSU, 80 p.
34. Otsu, N. (1979). A threshold selection method from gray-level histograms. IEEE transactions on systems, man, and cybernetics, 9(1), 62-66.
https://doi.org/10.1109/TSMC.1979.4310076
35. Polyakova, M. V., Volkova, N. P., Ivanova, O. V. (2008). Segmentation of the image of stochastic textures by the amplitude-detector method in the vast wavelet-remake, Information-measuring systems AAEKS, 2, 81-88.
36. Potapov, A. A. (2003). New information technologies based on probabilistic texture and fractal features in the radar detection of low-contrast targets], Radio engineering and electronics, 48(9), 1101-1119.
37. Przybył, K., Gawałek, J., Koszela, K., Przybył, J., Rudzińska, M., Gierz, Ł., & Domian, E. (2019). Neural image analysis and electron microscopy to detect and describe selected quality factors of fruit and vegetable spray-dried powders-Case study: Chokeberry powder. Sensors, 19(20), 4413.
https://doi.org/10.3390/s19204413
38. Rangayyan, R. M. (2005). Chap. 7, Analysis of texture, pp. 1277-1375. In Biomedical Image Analysis CRC Press LLC, Boca Raton, FL, 2005.
39. Shapiro, L. G. & Stockman, G. (2001). Chap. 7, Texture, pp. 235-247. In Computer Vision, PrenticeHall, 609 p.
40. Sizov, P. V., Palamar, I. N. (2011). A method and system for analyzing images based on segmentation using methods of growing and merging areas [Electronic resource], III All-Russian Scientific Zvorykinsky Readings: Collection of articles. thesis. doc. III All-Russian Interuniversity Scientific Conference, Murom, 243-244.
41. Smelyakov, K. S. (2008). Correlation method for texture recognition such as a mixture of images based on the use of histograms, Management systems for navigation and communication, 4(8), 18-21.
42. Sparavigna, A. C. (2016). A method for the segmentation of images based on thresholding and applied to vesicular textures. Philica, (889). hal-01408383. pp. 1-10.
43. Szumilas, L., Mičušík, B. & Hanbury, A. (2006). Texture segmentation through salient texture patches. Computer Vision Winter Workshop, pp. 1-6.
44. Tsapaev, A. P., Kretinin, O. V. (2012). Image segmentation methods in surface defect detection problems. Computer Optics, 2012, 36(3), 448-452.
45. Tuceryan, M., Jain, A. K. (1998). Texture Analysis. Chapter 2.1 in The Handbook of Pattern Recognition and Computer Vision (2nd Edition), by C. H. Chen, L. F. Pau, P. S. P. Wang (eds.), pp. 207- 248, World Scientific Publishing Co.
https://doi.org/10.1142/9789812384737_0007
46. Vizilter, Yu.V., Zheltov, S.Yu. (2011). Problems of technical vision in modern aviation systems. p. 11-44 in Proceedings: Technical Vision in Mobile Object Management Systems 2010: Proceedings of the scientific and technical conference-seminar. Vol. 4. Ed. R. R. Nazirov. Moscow: KDU, 328 p.
47. Zavalishin, N. V., Muchnik, I. B., Sheinin, R. L. (1975). Automatic classification of texture images], Avtomat. and Telemech., 2, 95-103.