Open AccessIndoor positioning technology based on map information perception
Author(s) -
Gan Xingli,
Huang Lu,
Wang Boyuan,
Li Shuang
Publication year - 2018
Publication title -
the journal of engineering
Language(s) - English
Resource type - Journals
ISSN - 2051-3305
DOI - 10.1049/joe.2018.8285
Subject(s) - inflection point , dead reckoning , computer science , constraint (computer aided design) , positioning technology , point (geometry) , artificial intelligence , path (computing) , computer vision , algorithm , real time computing , global positioning system , mathematics , programming language , telecommunications , geometry
With the increasing demand for location‐based services, indoor positioning technology has become one of the most attractive areas of research. Microelectromechanical systems sensors in smart terminals are used to realise a pedestrian dead reckoning algorithm. Owing to the accumulated error increased with time, the results of positioning will produce a large error, and an indoor positioning method based on the perception and constraint of map information is designed including straight path constraint and inflection point constraint. In the method of inflection point constraint, several common machine learning algorithms are compared through the experiments, and the secondary discriminant method is utilised to detect the inflection point with a detection accuracy of 97.62%. Finally, the performances of the improved algorithm and the traditional dead reckoning algorithm are compared in the experiments. The results show that the average positioning accuracy of the improved algorithm is 0.073 m, the positioning accuracy within 1 m reaches 100%, it is obviously higher than that of the traditional positioning algorithm and the effectiveness of the algorithm is verified.