Towards the Run and Walk Activity Classification through Step Detection - An Android Application

Melis Oner, Tolga Kaya, Patrick Seeling

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Falling is one of the most common accidents with potentially irreversible consequences, especially considering special groups, such as the elderly or disabled. One approach to solve this issue would be an early detection of the falling event. Towards reaching the goal of early fall detection, we have worked on distinguishing and monitoring some basic human activities such as walking and running. Since we plan to implement the system mostly for seniors and the disabled, simplicity of the usage becomes very important. We have successfully implemented an algorithm that would not require the acceleration sensor to be fixed in a specific position (the smart phone itself in our application), whereas most of the previous research dictates the sensor to be fixed in a certain direction. This algorithm reviews data from the accelerometer to determine if a user has taken a step or not and keeps track of the total amount of steps. After testing, the algorithm was more accurate than a commercial pedometer in terms of comparing outputs to the actual number of steps taken by the user.
Original languageEnglish
Title of host publicationProc. of the 34th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBS)
StatePublished - Aug 2012

Fingerprint

Dive into the research topics of 'Towards the Run and Walk Activity Classification through Step Detection - An Android Application'. Together they form a unique fingerprint.

Cite this