Real-time arm tracking for HMI applications

Matthew Masters, Luke Osborn, Nitish V Thakor, Alcimar Soares

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

Abstract

Limb tracking is an important aspect of human-machine interfaces (HMI). These systems, however, can often be limited by complex algorithms requiring significant processing power, obtrusive and immobile sensing techniques, and high costs. In this work, we utilize a sensor fusion algorithm implemented in commercial inertial measurement units (IMU) to combine accelerometer and gyroscope measurements in an effort to minimize computational requirements of the limb tracking system. In addition, previously developed methods were implemented to eliminate sensor drift by including information from a magnetometer. We tested the accuracy of our system by computing the root mean squared error (RMSE) of the true angle between the headings of two sensors and the estimate of that angle through quaternion-vector manipulations. An average RMSE of approximately 2.9° was achieved. Our limb tracking system is wearable, minimally complex, low-cost, and simple to use which has proven useful in multiple HMI applications discussed herein.

Original languageEnglish (US)
Title of host publication2015 IEEE 12th International Conference on Wearable and Implantable Body Sensor Networks, BSN 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Print)9781467372015
DOIs
StatePublished - Oct 15 2015
Event12th IEEE International Conference on Wearable and Implantable Body Sensor Networks, BSN 2015 - Cambridge, United States
Duration: Jun 9 2015Jun 12 2015

Other

Other12th IEEE International Conference on Wearable and Implantable Body Sensor Networks, BSN 2015
Country/TerritoryUnited States
CityCambridge
Period6/9/156/12/15

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Health Informatics
  • Instrumentation
  • Signal Processing

Fingerprint

Dive into the research topics of 'Real-time arm tracking for HMI applications'. Together they form a unique fingerprint.

Cite this