Abstract. The
article proposes a method for recognizing objects in the labyrinth domain
structures of magneto-optical materials and classification using a neural
network. Domain structure objects come in a variety of forms. The authors have
compiled a classification of types of domain structures on the basis of
practical results and a literature review. The features of domain
classification that significantly affect the choice of methods and algorithms
for solving problems are considered.
The classification is based on the
analysis of six coefficients of object shape by two neural networks. The
structures of neural networks and learning results are presented.
This method allows recogning nine
types of domains with simple and complex shapes and can be used for analyzing
images of objects in electronics, physics, chemistry, biology, medicine similar
in form with domains.
An example of object image
recognition in magneto-optical materials is presented. This method of
recognition is implemented in the software.
The developed software is used at
the Department of radio engineering of the Moscow state University. N. P.
Ogareva", for the study of magneto-optical and semiconductor materials.
Keywords:
neural network, object recognition, classification, domains, magneto-optical
materials.
References
1. Zvezdin
A.K., Kotov V.A. Modern magneto-optics and magneto-optical materials.
Bristol and Phyladelphia, IOP Publishing, 1997. 386 p.
2. Letiuk L. M.,
Kostishin V. G., Gonchar A. V. Tekhnologiya materialov magnitielektroniki
[Technology of magnetoelectronics materials]. Moscow, MISIS Publ., 2005. 352
p. (In Russian)
3. Bragin A. V.,
Logunov M. V., Pyanzin D. V., Navletov R. R., Ilyin S. V. Recognition of images
of micro-objects based on the neural network of direct distribution. 3
Vserossijskaya nauchno-tekhnicheskaya konferenciya molodyh konstruktorov i
inzhenerov «Mincevskie chteniya» - Third all-Russian scientific-technical
conference of young designers and engineers "Mints readings", Moscow,
Bauman MGTU Publ., 2016, pp. 140-148. (In Russian)
4. Barybin A.A.,
Tomilin V. I., Shapovalov V. I. Fiziko-tekhnicheskie osnovy makro-, mikro- I
nanoelektroniki [Physics and technological bases of macro-, micro - and
nanoelectronics. Moscow, FIZMATLIT Publ., 2011, p. 784. (In Russian)
5. Logunov M., Neverov
V., Mamin B., Skvortsov D., Sidorov R. Study of Nanoscale Inhomogeneities in
Silicon Carbide Crystals via Small-Angle X-Ray Scattering. Materials Science
Forum, 2016, Vol. 858, pp. 349-352. DOI
10.4028/www.scientific.net/MSF.858.349.
6. Logunov M. V., Gerasimov M. V.
Formation and evolution of giant dynamic domains in an ac
magnetic field.
Phys.
Solid State, 2003,
Vol. 45, Issue 6, pp. 1081-1086. DOI
https://doi.org/10.1134/1.1583794
7. Agashkov A.V.
Resonant domain photorefractivity in the structure of a liquid crystal-a photoconducting
orienting layer. Technical physics,
2010. Vol. 55, Issue 7, pp. 1009-1017. DOI
10.1134/S1063784210070157
8.
Maltsev V. N., Kandaurova G. S., Kartagulov L. N.
Dynamic stability of a spiral domain in an ac magnetic
field.
Phys.
Solid State, 2003,
Vol. 45, Issue 4, pp. 691-695. DOI
https://doi.org/10.1134/1.1569008
9. Kandaurova G. S.
New phenomena in the low-frequency
dynamics of magnetic domain ensembles.
Physics
– Uspekhi,
2002, Vol. 45, No. 10, pp. 1051-1072. DOI 10.1070/PU2002v045n10ABEH001064
10. Bragin, A. V.,
Pyanzin D. V. Selection of informative features for recognition of object
images of labyrinth domain structure. Vestnik of RGRTU - Bulletin of
Ryazan State Radio and Technical University, 2014, No. 1, pp. 21-25. (In Russian)
11. Bragin A.V.,
Gerasimov M. V., Logunov M. V., Pyanzin D. V., Navlatov N. R., Spirin A.V.
Software of the automated magneto-optical setup for the formation, registration
and image processing an ordered domain structures. Prikladnaya informatika -
Applied Informatics, 2016, Vol. 11, No. 6 (66), pp. 129-136. (In Russian)
12.
Haykin S. Neural
Networks: A Comprehensive Foundation, 2nd
edition, Prentice Hall PTR Upper Saddle
River, NJ, USA, 1998, ISBN:0132733501
13.
Gonzalez R.C., Woods R.E.. Digital image processing. 2nd ed. published by
Pearson Education, Inc, publishing as Prentice Hall. 2002. ISBN:
978-0-2011-8075-6.
14.
Bragin A.V., Logunov M. V., Nikitov S. A., Pyanzin D. V., Trifonov A.V.
Recognition of objects in labyrinthine domain structures. Komp'yuternaya
optika - Computer optics. 2013, Vvol. 37, No. 2, pp. 263-268. DOI:
10.18287/0134-2452-2013-37-2-263-268.
(In Russian)
15. Bragin A. V.,
Logunov M. V., Pyanzin D. V. The algorithm of recognition of objects in the
labyrinth structures. Trudy RTI im. Academician Mints A. L. - RTI works for
them. Academician. Mints A. L, 2012, Vol. 4, No. 48, pp. 46-53 (In Russian)