This report is an extension of a previous project called: ”Anvendelse af accelerometeret i en smartphone til estimering af distance og fart ved løb og gang”. In the previous project, an algorithm was developed that is able to estimate the step frequency, stride length, distance and speed of a walking or running user. To use this algorithm a manual calibration of the stride length is needed which decreases the usability.
The goal of this master thesis is to eliminate the need of a manual calibration by developing a method which integrates an automatic calibration of the stride length into the previously developed algorithm. The algorithm developed in this project is to be considered as a future supplement to GPS that can improve the estimation of distance and speed.
The master thesis consists of two phases.
The first phase of the master thesis examines the prerequisites in order to further develop the algorithm by using the sensors in a smartphone. A linear relationship between the step frequency and the stride length is established in this phase.
The second phase of the master thesis further develops the algorithm originated from the previous project. The development of the algorithm relates to identifying segments of a run that can be used for an automatic calibration that determine the linear relationship between a user’s step frequency and stride length. In the last part of the second phase it is examined how accurately the algorithm can estimate distances.