Human centric sensing by Android phones

Abstract

There is some existing indoor localization system mostly require support from external infrastructures, especially Wi-Fi hotspot. This project proposes a crowdsourcing localization system exploring the indoor geo-magnetism variance.

The project application relies on crowd generated the route for virtual map and to fill the map with magnetic fingerprints. As this will save the trouble to obtaining floor maps of the building, because it can only manually build the map or requires the owner to upload the map manually is available and accessible. Without using extra sensor such as Wi-Fi radio and accelerometer, this project only use magnetic field and gyroscope data to produce the virtual map and perform location estimation. We test the application in various building and shopping mall.

At the end of this project, the application done extensive experiments demonstrate to test. The result of this project is believed to have delivers satisfactory of localization accuracy, generate a virtual map and it is easily accessible to the general public. These results can be further improved by extending to other OS of the smartphone then android which can help more people to find their way in the shopping mall or large indoor building.