D.A. Aksenov1, O.A. Strizhenok2, O.D. Mironova3, E.A. Krapivina4, D.V. Zaharov5, N.P. Shivarev6

1-5 Financial University under the Government of the Russian Federation (Moscow, Russia)

6 College of Computer Science and Programming of the Financial University under the Government of the Russian Federation (Moscow, Russia)

The problem of navigation of blind and visually impaired people inside the building is relevant and socially significant. It is especially acute when a person with a visual impairment comes to a particular building for the first time and does not know where the object he needs is located.

Currently, to move around various buildings, a person with visual impairment uses either a white cane, the standard and most common means to help navigate and identify various obstacles, or a guide dog, or the help of other people. All of these options have their objective disadvantages. At the same time, most blind and visually impaired people also use mobile phones in everyday life. In this case, it is logical to create an application that would help the user in moving around the building and warn him about various obstacles that arise in his way. This would help make the environment more accessible to people with limited vision.

At the moment, there is no full-fledged navigator that could help a blind or visually impaired person move inside a building without the help of other people, so this idea is unique and especially useful. Existing solutions either simply recognize obstacle objects, or are focused on moving along the street, or use the help of volunteers.

The article deals with the problem of navigation of the blind and visually impaired inside the building.

The aim of the work is to develop a methodology for navigating visually impaired people inside the building without the use of special technical means and the constant help of volunteers.

As a result of our work, machine learning models have been developed that allow blind and visually impaired users to navigate the building with sufficient accuracy without the help of other people. The main advantages of this approach are the autonomy of a person with visual impairment, its relative cheapness and the absence of the need to use other technical means, the user only needs a mobile phone. As technical means of implementation, the programming language python (tensorflow library) and swift were chosen, since the application is implemented for smartphones on the IOS operating system.

Aksenov D.A., Strizhenok O.A., Mironova O.D., Krapivina E.A., Zaharov D.V., Shivarev N.P. Using machine learning technologies to create a navigation system inside the building. Neurocomputers. 2022. V. 24. № 3. Р. 50-60. DOI: https://doi.org/10.18127/j19998554-202203-05 (in Russian)

1. Mezhdunarodnyj den' slepyh. 2012. URL: https://aif.ru/health/life/37895 (in Russian).
2. «Dostupnaja sreda»: federal'naja programma i chastnaja zhizn'. 2012. URL: https://specialviewportal.ru/articles/post149 (in Russian).
3. Plikinas D., Zvironas A., Gudauskis M., Budrionis A., Daniusis P., i Slisorajtite I. Dostizhenija v oblasti navigacii vnutri pomeshhenij dlja slepyh ljudej: kratkij obzor tehnologicheskogo priborostroenija. IEEE Instrumentation and Measurement. 2020. № 23(4). S. 22-32 (in Russian).
4. Dovydenkov V.N. Obzor sovremennyh apparatnyh i programmnyh sredstv jekrannogo dostupa dlja nezrjachih pol'zovatelej komp'juternoj tehniki. Mezhdunar. nauch.-praktich. konf. «Realizacija sovremennyh podhodov k reabilitacii invalidov v processe obshhego professional'nogo obrazovanija». 2013 (in Russian).
5. Podderzhka navigacii na smartfonah dlja slepyh i slabovidjashhih ljudej - kompleksnyj analiz potencialov i vozmozhnostej. 2020 g. URL: https://www.scopus.com/login?qurl=https://www.scopus.com%2frecord%2fdisplay.uri%3feid%3d2-s2.0-85088539723% 26origin%3dresultslist%26sort%3dcp-f%26src%3ds%26sid%3dbf565101bccac280e7cf831b337a04e7%26sot%3db%26sdt%3db%26 sl%3d142%26s%3dTITLE-ABS-KEY%2528Smartphone%2bnavigation%2bsupport%2bfor%2bblind%2band%2bvisually%2bim-paired%2bpeople%2b-%2ba%2bcomprehensive%2banalysis%2bof%2bpotentials%2band%2bopportunities%2529%26relpos%3d0 %26citeCnt%3d2%26searchTerm%3d (in Russian).
6. Povetina V., Serkov K., Obabkov I., Letavin D. Sravnitel'nyj analiz tehnologij vnutrennej navigacii. Materialy Mezhdunar. mul'tikonf. po inzhenernym, komp'juternym i informacionnym naukam «SIBIRKON 2019». 2019. S. 73-78 (in Russian).
7. iBeacon: razrabotka mobil'nyh prilozhenij. 2015. URL: https://wnfx.ru/ibeacon-razrabotka-mobilnyih-prilozheniy-dlya-ios-po-tehnologii-ibeacon/ (in Russian).
8. Ahmetovich D., Masketti S., Bernaredzhi S., Gerrejro Dzh., O. U., Asakava S. Kompensacija oshibok povorota pri glubokom obuchenii pri pomoshhi navigacii dlja ljudej s narushenijami zrenija ili slepotoj. ACM Transactions on Accessible Computing. 2019. № 12(4) (in Russian).
9. Dosajja V., Varshni S., Parameshvarappa V.K., Benival A., Tak S. Nedorogaja sistema dopolnennoj real'nosti dlja shirokopolosnoj navigacii vnutri pomeshhenij 2020. Materialy Mezhdunar. konf. po nauke i primeneniju pomoshhi v prinjatii reshenij. DASA. 2020. S. 190-195 (in Russian).
10. Navigacija po ViaOpta. 2020 g. URL: https://apps.apple.com/ru/app/nav-by-viaopta/id908435532 (in Russian).
11. Microsoft Soundscape. 2021. URL:https://www.microsoft.com/en-us/research/product/soundscape/
12. GetThere. 2021. URL: https://www.getthere.com/
13. Obzor programmy DotWalker Pro. 2020. URL: http://tiflo-pro.net/category/audio-fajly/obzor-programmy-dotwalker/ (in Russian).
14. AriadneGPS. 2021. URL: https://www.ariadnegps.eu/en/un-app-innovativa-per-la-mobilitaaninnovative-app-for-your-mobility-english/
15. Right-Here Blog. 2021. URL: https://www.right-hear.com/blog/ 
16. Osmand. 2021. URL: https://osmand.net/
17. Определитель предметов. 2020. URL: http://sensor-tech.ru/object_app
18. TatTapSee. 2019. URL: https://taptapseeapp.com/ 
19. Chen Z., Zou H., Jiang H., Zhu Q., Soh Y.C., Xie L. Ob’edinenie Wi-Fi, datchikov smartfonov i orientirov s ispol'zovaniem fil'tra Kalmana dlja lokalizacii vnutri pomeshhenij. Sensors (Shvejcarija). 2015. № 15(1). С. 715-732 (in Russian).