Individual classification through autoregressive modelling of micro-doppler signatures

Guillaume Garreau, Nicoletta Nicolaou, Julius Georgiou

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper introduces the use of autoregressive modelling (AR) to characterize individual human gait signatures from micro-Doppler data. AR models are fitted to micro-Doppler data obtained while 6 subjects walk towards a custom-made ultrasonic transceiver module. The estimated AR coefficients capture individual movement characteristics. Such features can be used to identify different subjects quickly and with low computational cost. In the best configuration, average performance higher than 98% was obtained.

Original languageEnglish
Title of host publication2012 IEEE Biomedical Circuits and Systems Conference
Subtitle of host publicationIntelligent Biomedical Electronics and Systems for Better Life and Better Environment, BioCAS 2012 - Conference Publications
Pages312-315
Number of pages4
DOIs
Publication statusPublished - 1 Dec 2012
Externally publishedYes
Event2012 IEEE Biomedical Circuits and Systems Conference: Intelligent Biomedical Electronics and Systems for Better Life and Better Environment, BioCAS 2012 - Hsinchu, Taiwan, Province of China
Duration: 28 Nov 201230 Nov 2012

Conference

Conference2012 IEEE Biomedical Circuits and Systems Conference: Intelligent Biomedical Electronics and Systems for Better Life and Better Environment, BioCAS 2012
Country/TerritoryTaiwan, Province of China
CityHsinchu
Period28/11/1230/11/12

Keywords

  • autoregressive models
  • individual recognition
  • Micro-Doppler
  • ultrasonic device

Fingerprint

Dive into the research topics of 'Individual classification through autoregressive modelling of micro-doppler signatures'. Together they form a unique fingerprint.

Cite this