Estimation of Human Movement State by Smartphone Sensor Information without GPS

We developed a smartphone application for estimating movement states with an acceleration sensor, a magnetic sensor, and an electric wave strength. Life-log management system is also implemented in a Web application as an example of utilizing the results obtained by our estimating system. The verification results confirm that the system can estimate the movement states with high accuracy regardless of the users and models. All the train stations that were part of the user's course in the experiment were recognized by the system and obtained in the Web application. These results suggest that a detailed life log can be generated by our proposed system.