Eng During last 365 days Approved articles: 2075,   Articles in work: 296 Declined articles: 803 
Library

Back to contents

Software systems and computational methods
Reference:

Optimization of information signs recognition methods for the problem of indoor navigation
Lebedev Leonid Ivanovich

PhD in Physics and Mathematics

Leading Researcher, Lobachevsky National Research Nizhny Novgorod State University

603950, Russia, Nizhegorodskaya oblast', g. Nizhnii Novgorod, pr. Gagarina, 23

lebedev@pmk.unn.ru
Osipov Mikhail Pavlovich

PhD in Technical Science

Leading Researcher, Lobachevsky National Research Nizhny Novgorod State University

603950, Russia, Nizhegorodskaya oblast', g. Nizhnii Novgorod, pr. Gagarina, 23

virtulab@mail.ru

Abstract.

The paper proposes the solution of a number of problems related to the optimization of recognition of information signs by the correlation-extremal contour method based on the evaluation of similarity invariant with respect to affine transformations.A description is given of a speed-efficient algorithm for obtaining regions of a certain color and its parallelization. It is shown that solving this problem allows reducing both the number of objects presented for recognition and the number of standards used, as well as obtaining a better contour description of objects and, therefore, significantly increasing the speed of the recognition algorithm itself. The solution of the problem of optimization of calculations in the recognition algorithm itself is given. It is shown that one of the ways to solve this problem for a particular chosen recognition algorithm is to parallelize the computation of the similarity estimate by the affine transformation parameters. Another presented way of solving the problem of optimization of computations was focused on obtaining estimates of the similarity of the current standard with objects from the local area of the image, which is determined on the basis of the parameters of affine transformation and dimensions of the standard. It is noted that the created algorithmic software allows to solve problems of recognition of information objects in real time.

Keywords: affine transformation, optimization, color segmentation, parallel algorithm, pattern recognition, image processing, computer graphics, standart image, adaptive description, correlation-extreme contour method

DOI:

10.7256/2454-0714.2018.4.28374

Article was received:

18-12-2018


Review date:

15-12-2018


Publish date:

10-01-2019


This article written in Russian. You can find full text of article in Russian here .

References
1.
Mautz R. Indoor Positioning Technologies. - Zurich: Department of Civil, Environmental and Geomatic Engineering, ETH, 2012. - 127 p.
2.
Computer vision and GIS for the navigation of blind persons in buildings / M. Serrão, S. Shahrabadi, M. Moreno et al. // Universal Access in the Information Society. 2014. Vol. 14. 1. P. 6780.
3.
Text detection and recognition for semantic mapping in indoor navigation / M. Sami, Y. Ayaz , M. Jamil et al. // International Conference on IT Convergence and Security. - IEEE. - 2015. - P. 1−4.
4.
Werner M., Kessel M., Marouane C., Indoor positioning using smartphone camera // Proceedings of the International Conference on Indoor Positioning and Indoor Navigation (IPIN 11). Portugal : IEEE, 2011. R. 16.
5.
Osipov M.P., Patrushev A.O. Metody korrektirovki mestopolozheniya v zadache avtonomnoi navigatsii v zakrytykh pomeshcheniyakh // Trudy 26-oi Mezhdunarodnoi konferentsii po komp'yuternoi grafike i zreniyu. - Nizhnii Novgorod, 2016. - S. 417-419.
6.
Osipov M. P., Andreev V. S. Method monitoring of movement in the task of indoor navigation // Proceedings of the International Conference Information Technology and Nanotechnology. Session Data Science. - Samara, Russia, 2018. - Vol. 2212. - P. 222-227.
7.
Linev A.V., Bogomolov D.K., Bastrakov S.I. Tekhnologiya parallel'nogo programmirovaniya dlya protsessorov novykh arkhitektur : uchebnoe posobie.M.: Izd-vo Mosk. un-ta, 2010. - 160 s.
8.
Fain V.S. Opoznavanie izobrazhenii.- M.: Nauka, 1970. - 299 s.
9.
Lebedev L.I. Korrelyatsionno ekstremal'nye konturnye metody raspoznavaniya. Teoreticheskie osnovy: uchebnoe posobie. - Nizhnii Novgorod: Izd-vo Nizhegorodskogo gosudarstvennogo universiteta, 2013. - 113 s.
10.
Lebedev L.I., Vasin Yu.G., Osipov M.P. Avtonomnaya navigatsiya v zakrytykh pomeshcheniyakh s ispol'zovaniem adaptivnogo opisaniya etalonnykh izobrazhenii // Trudy 27-oi Mezhdunarodnoi konferentsii po komp'yuternoi grafike i zreniyu GraphiCon 2017. - Perm': PGNIU,2017. - S. 212-216.