With the increase of mobile devices, and their continuous improvement we could track human behaviour with more and more complex methods. Furthermore, the number of systems and services based on this idea are ever increasing. The position and state of people could be vital in smart homes, or at workspaces, and can be utilized in computer games. Nowadays the various computer-vision based object tracking systems are world-wide, which we even can use to follow the daily pattern of humans, however the implementation of such systems demands more resource then many could afford.
In this document I study the methods of position detection based on mobile phones, putting an emphasis on the various sensor data and camera that the phone provides. Both data sources have their advantages and drawbacks, making them incapable to provide a full-fledged and efficient solution to the problem. While making this document, I am going to design and implement a system, that can determine the path of a phone with more reliability and precision by fusing the data of the two sources. At the end of this document I will check the correctness and constraints of the developed system and based on the results in term of speed and accuracy I will mention a few possible ways to improve the system.