A new technique for foot-off and foot contact detection in a gait cycle based on the knee joint angle using microsoft kinect v2

A. Amini, K. Banitsas, S. Hosseinzadeh

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

20 Citations (Scopus)

Abstract

The Microsoft Kinect RGB-D sensor has been proven to be a reliable tool for gait analysis and rehabilitation purposes. Although it is accurate for detecting upper body part movements, even the second iteration of the Kinect sensor lacks the accuracy when it comes to lower extremities. while detecting foot-off and foot contact phases of a gait cycle is an important part of a gait performance analysis, The Kinect's intrinsic inaccuracies make it an unreliable tool to detect them accurately. We propose a new Kinect based technique for detecting foot-off and foot contact phases in a gait cycle that solely relies on a subject's knee joint relative angle. The system was tested on 11 healthy subjects walking in pre-defined pathways in 12 walking sessions while the Kinect v2 camera was placed at different heights ranging from 0.65 to 1.57 and angles ranging from 0 to 45 degrees to the ground. The algorithm's accuracy was also compared to another footstep detection method based on the subject's ankle joints height to the ground. The results showed 86.52% accuracy in detecting foot-off and foot contact events on average for both feet.
Original languageEnglish
Title of host publicationProceedings of 2017 IEEE EMBS International Conference on Biomedical and Health Informatics
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages153-156
Number of pages4
ISBN (Electronic)9781509041794
DOIs
Publication statusPublished - 13 Apr 2017

ASJC Scopus subject areas

  • Health Informatics
  • Computer Science Applications
  • Biomedical Engineering

Fingerprint

Dive into the research topics of 'A new technique for foot-off and foot contact detection in a gait cycle based on the knee joint angle using microsoft kinect v2'. Together they form a unique fingerprint.

Cite this