Synchronization of oscillations for machine perception of gaits
Computer Vision and Image Understanding
Synchronization of oscillations for machine perception of gaits
Computer Vision and Image Understanding
ASB'03 Proceedings of the 1st international conference on Advanced Studies in Biometrics
Hi-index | 0.00 |
Existing gait recognition approaches do not givetheir theoretical or experiential performance predictions Therefore the discriminating power of gait as afeatureforhuman recognition cannot be evaluated. Inthis paper we first propose a kinematic based approachto recognize human by gait The proposed approach estimates 3D human walking parameters by performing aleast squares fit of the 3D kinematic model to the 2Dsilhouette extractedfrom a monocular image sequenceNext a Bayesian based statistical analysis is performedto evaluate the discriminating power of extracted features Through probabilistic simulation we not onlypredict the probability of correct recognition PCR withregard to different within class feature variance butalso obtain the upp er bound on PCR with regard todifferent human silhouette resolution In addition themaximum number of people in a database is obtainedgiven the allowable error rate This is extremely important for gait recognition in large databases.