Map Assisted Pedestrian Orientation Detection by Smart Phone

Song Zheng -,Ma Jian -,Wang Wendong -,Xiong Yongping -

DOI: https://doi.org/10.4156/jcit.vol8.issue3.53

2013-01-01

Journal of Convergence Information Technology

Abstract:Accurate indoor navigation system based on smart phones has been extensively studied and considered to be of good prospect for the reducing cost and increasing popular rate of smart phones. Dead reckoning method which measures steps and direction of each step is one of the most popular pedestrian positioning methods on the smart-phone platform, and its accuracy is mainly limited by the error of orientation detection. This paper presents a lightweight algorithm(MAPOD) that can measure pedestrian’s orientation in an accurate and real-time way by the following procedures: indentify certain locations such as corridor-corner in an indoor environment by signatures on multiple sensors to establish pedestrian’s initial orientation or correct orientation error, establish pedestrian’s temporary orientation when a change in smart phone’s mounting method is detected, calculate orientation by initial orientation and orientation sensor’s reading when no mounting change is detected. Results from three different mounting scenarios demonstrate 90% direction error of 5 °. MAPOD gains more accuracy with no significant increase in system energy consumption, is effective to improve the indoor positioning accuracy and availability in smart phones.