Geodetski list, Vol. 70 (93) No. 1, 2016.
Izvorni znanstveni članak
Pedestrian Indoor Positioning and Tracking using Smartphone Sensors, Step Detection and Map Matching Algorithm
Ľubica Ilkovičová
orcid.org/0000-0002-8061-4592
; Department of Surveying, Faculty of Civil Engineering, Slovak University of Technology in Bratislava, Bratislava, Slovakia
Pavol Kajánek
orcid.org/0000-0002-8006-317X
; Department of Surveying, Faculty of Civil Engineering, Slovak University of Technology in Bratislava, Bratislava, Slovakia
Alojz Kopáčik
orcid.org/0000-0002-3888-3940
; Department of Surveying, Faculty of Civil Engineering, Slovak University of Technology in Bratislava, Bratislava, Slovakia
Sažetak
The paper deals with indoor navigation using inertial sensors (accelerometers, gyroscopes, etc.) built in a smartphone. The main disadvantage of the use of inertial sensors is the accuracy, which rapidly decreases with the increasing time of the measurement. The reason of the deteriorating accuracy is the presence of errors in inertial measurements, which are accumulated in the integration process. The paper describes the determination of a pedestrian trajectory using step detection method, which is improved with utilization of the adaptive step length estimation algorithm. This algorithm reflects the change of the step length with different types of movement. The proposal of the data processing uses information from floormap, what allows the verification of the pedestrian position and detects the collision of the trajectory with the floormap. The proposed algorithm significantly increases the accuracy of the resulting trajectory. Another extension of the proposed algorithm is the implementation of the barometer measurements for determination of the height differences. This fact allows change the floor in a multi-storey buildings. The experimental measurement was realized with a smartphone Samsung Galaxy S4.
Ključne riječi
smartphone; inertial sensor; systematic error; step detection; adaptive step length estimation; map matching
Hrčak ID:
156880
URI
Datum izdavanja:
25.3.2016.
Posjeta: 3.440 *