EAI Endorsed Transactions on Mobile Communications and Applications 16(9): e5

Research Article

Real time event-based segmentation to classify locomotion activities through a single inertial sensor

Download56 downloads
  • @ARTICLE{10.4108/eai.14-10-2015.2261695,
        author={Benish Fida and Daniele Bibbo and Ivan Bernabucci and Antonino Proto and Silvia Conforto and Maurizio Schmid},
        title={Real time event-based segmentation to classify locomotion activities through a single inertial sensor},
        journal={EAI Endorsed Transactions on Mobile Communications and Applications},
        volume={16},
        number={9},
        publisher={ACM},
        journal_a={MCA},
        year={2015},
        month={12},
        keywords={gait events, dynamic segmentation, locomotion activities, inertial sensors, classification},
        doi={10.4108/eai.14-10-2015.2261695}
    }
    
  • Benish Fida
    Daniele Bibbo
    Ivan Bernabucci
    Antonino Proto
    Silvia Conforto
    Maurizio Schmid
    Year: 2015
    Real time event-based segmentation to classify locomotion activities through a single inertial sensor
    MCA
    EAI
    DOI: 10.4108/eai.14-10-2015.2261695
Benish Fida1,*, Daniele Bibbo1, Ivan Bernabucci1, Antonino Proto1, Silvia Conforto1, Maurizio Schmid1
  • 1: Roma Tre University, Rome
*Contact email: fida.benish@uniroma3.it

Abstract

We propose an event-based dynamic segmentation technique for the classification of locomotion activities, able to detect the mid-swing, initial contact and end contact events. This technique is based on the use of a shank-mounted inertial sensor incorporating a tri-axial accelerometer and a tri-axial gyroscope, and it is tested on four different locomotion activities: walking, stair ascent, stair descent and running. Gyroscope data along one component are used to dynamically determine the window size for segmentation, and a number of features are then extracted from these segments. The event-based segmentation technique has been compared against three different fixed window size segmentations, in terms of classification accuracy on two different datasets, and with two different feature sets. The dynamic event-based segmentation showed an improvement in terms of accuracy of around 5% (97% vs. 92% and 92% vs. 87%) and 1-2% (89% vs. 87% and 97% vs. 96%) for the two dataset, respectively, thus confirming the need to incorporate an event-based criterion to increase performance in the classification of motion activities.