Title :
Foot-mounted inertial measurement unit for activity classification
Author :
Ghobadi, Mostafa ; Esfahani, Ehsan T.
Author_Institution :
Mech. & Aerosp. Eng. Dept., Univ. at Buffalo SUNY, Buffalo, NY, USA
Abstract :
This paper proposes a classification technique for daily base activity recognition for human monitoring during physical therapy in home. The proposed method estimates the foot motion using single inertial measurement unit, then segments the motion into steps classify them by template-matching as walking, stairs up or stairs down steps. The results show a high accuracy of activity recognition. Unlike previous works which are limited to activity recognition, the proposed approach is more qualitative by providing similarity index of any activity to its desired template which can be used to assess subjects improvement.
Keywords :
biomechanics; patient monitoring; patient rehabilitation; activity classification; daily base activity recognition; foot motion; foot mounted inertial measurement unit; human monitoring; physical therapy; stairs down steps; stairs up steps; template matching; walking; Acceleration; Accuracy; Foot; Legged locomotion; Motion segmentation; Sensors; Vectors;
Conference_Titel :
Engineering in Medicine and Biology Society (EMBC), 2014 36th Annual International Conference of the IEEE
Conference_Location :
Chicago, IL
DOI :
10.1109/EMBC.2014.6945068