METRIC GRANULATION METHODS FOR IMAGE DESCRIPTION IN THE IMAGE CLASSIFICATION PROBLEM
DOI:
https://doi.org/10.26906/SUNZ.2024.3.106Keywords:
computer vision, structural methods of image classification, set of descriptors, clustering, vector granulation, classification accuracyAbstract
The subject of the research of the article is the methods of image classification in computer vision systems. The aim is to improve the structural methods of classification to increase their speed by implementing a reduced classification features system based on the granulation of reference descriptions. Applied methods: BRISK and ORB key point detectors, set theory and vector spaces apparatus, metric models of vector clustering and granulation, “bad of words” and voting for classification models, software modeling. Results are obtained: models for the transformation of image description in the form of a set of cluster centers and the construction of a reduced description using granulation based on the similarity of vectors were developed, high-speed image classification methods based on the implementation of transformed descriptions were researched. The effectiveness of the developed modifications of the classifier depends on the data granulation method and the threshold parameters for the equivalence of vectors, the minimum metric value and the number of votes of the descriptors for the winner class in the voting scheme. The construction of a hierarchical system of features with several granulation levels was proposed. Based on the implementation of modifications, it was possible to reduce the computational costs by hundreds of times while ensuring the effectiveness of classification on the training sample of data. The practical significance of the work includes the software models for high-speed classification, confirming the functionality and accuracy of the proposed modifications for the applied image dataset, creating software application for implementing modifications of classifiers in computer vision.Downloads
References
Tymchyshyn R., Volkov O., et al. (2018), Modern Approaches to Computer Vision, Control systems and computers, 6, 46-73.
Daradkeh Y. I., Gorokhovatskyi V., Tvoroshenko I., Gadetska S., and Al-Dhaifallah M. (2023), Statistical data analysis models for determining the relevance of structural image descriptions, IEEE Access, vol. 11, pp. 126938–126949, Nov. 2023, doi: 10.1109/ACCESS.2023.3332291.
Gorokhovatskyi V. Vlasenko N. (2021), The image description reduction in the set of descriptors on informativeness metric criteria base. Advanced Information Systems, 5 (4), 10–16. DOI: https://doi.org/10.20998/2522-9052.2021.4.02
P. Flach. (2012), Machine learning. The Art and Science of Algorithms that Make Sense of Data. New York, NY, USA: Cambridge University Press.
Gorokhovatskyi V. A. (2018), Image classification methods in the space of descriptions in the form of a set of the key point descriptors, Telecommunications and Radio Engineering, 77(9), 787-797.
Kuchuk H., Kovalenko A., Ibrahim B.F., Ruban I. (2019), Adaptive compression method for video information. International Journal of Advanced Trends in Computer Science and Engineering. Vol. 8 (1). P. 66–69. DOI: http://dx.doi.org/10.30534/ijatcse/2019/1181.22019
Rublee E., Rabaud V., Konolige K., and Bradski G., ORB: An efficient alternative to SIFT or SURF, in Proc. Int. Conf. Comput. Vis., Nov. 2011, pp. 2564–2571, doi: 10.1109/ICCV.2011.6126544.
Leutenegger S., Chli M. and Siegwart R. Y. BRISK: Binary robust invariant scalable keypoints, in 2011 Int. Conf. on Computer Vision, Barcelona, Spain, pp. 2548–2555, 2011.
Gorokhovatskyi V., Tvoroshenko I. (2023), Identification of visual objects by the search request. International scientific symposium «Intelligent Solutions-S». Computational intelligence. Decision making theory: proc. of the international symposium, September 28, 2023, Kyiv-Uzhorod, Ukraine, pp. 25-27.
Gorokhovatskyi O., Gorokhovatskyi V. and Peredrii O. (2018), Analysis of application of cluster descriptions in space of characteristic image features, Data, vol. 3, no. 4, pp. 52.
Gorokhovatsky V., Putyatin Y. and Stolyarov V. (2017), Research of Effectiveness of Structural Image Classification Methods using Cluster Data Model, Radio Electronics, Computer Science, Control, vol. 3 (42), pp. 78–85.
Robnik-Sikonja,M., Kononenko, I. (2003), Theoretical and empirical analisis of ReliefF and RReliefF. Machine Learning 53 (1-2): 23-69.
Stańczyk U. (2015), Feature Evaluation by Filter, Wrapper, and Embedded Approaches. In: Stańczyk U., Jain L. (eds) Feature Selection for Data and Pattern Recognition. Studies in Computational Intelligence, Springer, Berlin, Heidelberg, vol. 584, 568 p.
Xiong, H. and Z. Li. (2014), Data Clustering: Algorithms and Application. 1st ed. Boca Raton: CRC Press.
Zhang X., Yu F. X., Karaman S.and Chang S.-F. (2017), Learning discriminative and transformation covariant local feature detectors,” in IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 4923–4931.
Lowe D. G. (2004), Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision. 60 (2). doi:10.1023/B:VISI.0000029664.99615.94
Crowley, J., Riff O. (2003), Fast computation of scale normalized Gaussian receptive fields, Proc. Scale-Space'03, Isle of Skye, Scotland, Springer Lecture Notes in Computer Science, 2695.
Putyanin, E.P., Averin, S.I. (1990), Image processing in robotics. Moscow, Mashinostroeniye, 320 p.
Gorokhovatskyi V., Tvoroshenko I., and Yakovleva O. (2024), Transforming image descriptions as a set of descriptors to construct classification features, Indonesian Journal of Electrical Engineering and Computer Science, vol. 33, no. 1, pp. 113-125, doi: 10.11591/ijeecs.v33.i1.
Martino A., De Santis E., Rizzi A. (2023), On Information Granulation via Data Filtering for Granular Computing-Based PatternRecognition: A Graph Embedding Case Study. SN Computer Science, 4:314, https://doi.org/10.1007/s42979-023-01716-1
WikiArt. Енциклопедія візуальних мистецтв. https://www.wikiart.org/uk/mariya-primachenko ( 01.04.2024)
Gorokhovatsky V.A. Putyatin Y. P. (2009), Image Likelihood Measures of the Basis of the Set of Conformities. Telecommunications and Radio Engineering, 68 (9), p. 763–778.
Tvoroshenko I., and Zarivchatskyi R. (2020), Analysis of existing methods for searching object in the video stream, Abstracts of VI International Scientific and Practical Conference «About the problems of science and practice, tasks and ways to solve them» (October 26-30, 2020). Milan, Italy, pp. 500–505.
Yakovleva, О., Kovtunenko, A., Liubchenko, V., Honcharenko, V., & Kobylin, O. (2023), Face Detection for Video Surveillance-based Security System (COLINS-2023). In CEUR Workshop Proceedings (Vol. 3403). pp. 69-86.
Kuchuk, H., Podorozhniak, A., Liubchenko, N., and Onishchenko, D. (2021), System of license plate recognition considering large camera shooting angles, Radioelectronic and Computer Systems, 4(100), 82–91.
OpenCV. URL: https://docs.opencv.org/
