Automatic gait recognition using dynamic variance features

Y. Chai, Jinchang Ren, R. Zhao, J. Jia

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

13 Citations (Scopus)


Human gait recognition is currently one of the most active research topics in computer vision. Existing recognition methods suffer, in our opinion, from two shortcomings: either much expensive computation or poor identification effect; thus a new method is proposed to overcome these shortcomings. Firstly, we detect the binary silhouette of a walking person in each of the monocular image sequences. Then, we extract the pixel values at the same pixel position over one gait cycle to form a dynamic variation signal (DVS). Next, the variance features of all the DVS are computed respectively and a matrix is constructed to describe the dynamic gait signature of individual. Finally, the correlation coefficient measure based on the gait cycles and two different classification methods (NN and KNN) are used to recognize different subjects. Experimental results show that our method is not only computing efficient, but also very effective of correct recognition rates over 90% on both UCSD and CMU databases

Index Terms

o Controlled Indexing

computer vision , feature extraction , gait analysis , image classification , image resolution , image sequences , video databases
o Non Controlled Indexing

automatic gait recognition , computer vision , dynamic gait signature , dynamic variation signal , feature extraction , human gait recognition , image classification methods , image resolution , monocular image sequences , walking person binary silhouette


* Abstract
* Index Terms

Brought to you by

* Your institute subscribes to:
* IEEE-Wiley eBooks Library , IEEE/IET Electronic Library (IEL)
* What can I access?

Terms of Use

Original languageEnglish
Title of host publicationProceedings of 7th International Conference of Automatic Face and Gesture Recognition (FG2006)
Place of PublicationSouthampton, UK
Publication statusPublished - 2006


  • face recognition
  • gait recognition


Dive into the research topics of 'Automatic gait recognition using dynamic variance features'. Together they form a unique fingerprint.

Cite this