Online motions recognition method using mobile phone accelerometer

Abstract

Accelerometers can obtain data about the move­ments of sorne person. Ihe continued study of this information can help to a doctor to establish a correct diagnosis or a rehabilitation plan for a person with mobility problems. This paper focuses on a new method to implement a motion recognition process with accelerometer sensor data contained in a mobile device. Ali the steps are described, from the data collection to motion recognition through statistical study data and machine leaming algorithms. Nowadays, mobile phones have big processing ca­pacity to execute complex programs. In the last step of this process, a classification function is im­plemented in a mobile phone for online motion recognition. The practical experience showed an overall accuracy of 91 % when recognizing four activities.