Fedotov Nikolay Gavrilovich, Doctor of engineering sciences, professor, head of sub-department of economic cybernetics, Penza State University (40 Krasnaya street, Penza, Russia), firstname.lastname@example.org
Goldueva Dar'ya Alekseevna, Candidate of engineering sciences, associate professor, sub-department of economic cybernetics, Penza State University (40 Krasnaya street, Penza, Russia), email@example.com
Background. In many branches of human knowledge an essential part of information lies in images of compound structures, a lot of them contain textures. Along with its general-purpose significance, the task of classifying such images has also proved its applied relevance. Successful solution of the task would account for the efficiency of information processing in such fields as aerospace research, earth observation data analysis, as well as medical and technical diagnostics. The purpose of the present study is to elaborate a theory of image analysis and recognition, based on stochastic geometry and functional analysis, to analyze and classify coloured textures.
Materials and methods. In the course of the work the authors applied an innovative approach to coloured textures' analysis and recognition from the standpoint of stochastic geometry and functional analysis. To form triple features of coloured textures the approach, developed by Professor N. G. Fedotov's scientific school, was taken to help describe an image under investigation from the viewpoint of its geometrical characteristics, as well as from the aspect of its colour specific features.
Results. A new approach based on stochastic geometry and functional analysis has been offered to form features of coloured textures. Triple features theory extension helped analyze coloured textures directly, using no prior binarization. The features' invariance towards linear deformations of the coloured textures under investigation has been checked experimentally.
Conclusions. The built group of triple features provides for a more complete description of coloured textures. The trinary structure serves to generate a great number of triple features, which helps enhance recognition flexibility, versatility, and robustness. Moreover, the conducted experiments have shown that a specific choice of functionals within the structure of a triple feature form characteristics invariant to a group of motions and linear deformations.
1. Fedotov N. G. Metody stokhasticheskoy geometrii v raspoznavanii obrazov [Methods of stochastic geometry in image recognition]. Moscow: Radio i svyaz', 1990, 144 p.
2. Fedotov N. G. Teoriya priznakov raspoznavaniya obrazov na osnove stokhasticheskoy geometrii i funktsional'nogo analiza [Theory of image recognition features on the basis of stochastic geometry and functional analysis]. Moscow: FIZMAT-LIT, 2009, 304 p.
3. Kendall W. S., Molchanov I. New Perspectives in Stochastic Geometry. Oxford: UK: Oxford University Press, January, 2010.
4. Kadyrov A. A., Saveleva M. V., Fedotov N. G. Image scanning leads to alternative understanding of image. Third int. conf. on automation, robotics and computer vision (ICARCV’94). Singapore, 1994.
5. Fedotov N. G., Kadyrov A. A. In Proc. of 5th International Workshop on Digital In Processing and Computer Graphics, Proc. International Society for Optical Engineering (SPIE). 1995, vol. 2363, pp. 256–261.
6. Fedotov N. G., Shulga L. A. WSCG’2000 Conference Proceedings. – University of West Bohemia. 2000, vol. 1 (2), pp. 373–380.
7. Kadyrov A. A. and Fedotov N. G. Advances in Mathematical Theory and Applications. 1995, vol. 5, no. 4, pp. 546–556.
8. Fedotov N. G., Kol'chugin A. S., Smol'kin O. A., Moiseev A. V., Romanov S. V. Izmeritel'naya tekhnika [Measuring technology]. 2008, no. 2, pp. 56–61.
9. Vidal M., Amigo J. M. Chemometrics and Intelligent Laboratory Systems. 2012, vol. 117, no. 1, pp. 138–148.
10. Fedotov N. G., Nikiforova T. V. Izmeritel'naya tekhnika [Measuring technology]. 2002, no. 12, pp. 27–31.
11. Fedotov N. G., Shulga L. A., Roy A. V. Proc. Int. Conf. Pattern Recognition and Image Analysis. PRIA-7-2004. 2004, vol. 2, pp. 473–475.
12. Shin B.-S., Cha E.-Y., Kim K.-B., Cho K.-W., Klette R., Young W. W. 3rd International Conference on Bio-Inspired Computing: Theories and Applications, BIC-TA 2008 sponsors: IEEE South Australia Section. Adelaide, NT, 2008, pp. 97–102.
13. Fooprateepsiri R., Kurutach W. Proc. of the IADIS Int. Conf. Intelligent Systems and Agents 2010, Proc. of the IADIS European Conference on Data Mining 2010, Part of the MCCSIS–2010. 2010, pp. 83–90.
14. Fedotov N. G., Shulga L. A., Kol’chugin A. S., Smol’kin O. A., Romanov S. V. Proc. Of the 8th Int. Conf. on Pattern Recognition and Image Analysis (PRIA-8-2007). Yoshkar-Ola, Russia, 2007, vol. 1, pp. 299–300.