Biometric Surveillance Solutions based on Gait Recognition

Project description

Normally surveillance involves the monitoring of behavior and activities of humans.  Biometric systems are technologies which measure and analyze human physical and/or behavioral characteristics for authentication, identification or screening purposes.  Some physical characteristics which can be measured are fingerprints, iris, DNA and facial patterns. The two most common behavioral characteristics are voice and human gait (which refers to the manner of walking).  

Recent biometric technologies have moved towards real-time non-cooperative surveillance from a distance. For example both facial and iris recognition systems operate within two-meters of the human and may soon be able to operate within ten-metres.   However gait identification may prove to be one of the most promising emerging biometrics for surveillance such as screening of individuals in high-security civilian or military facilities and monitoring potential targets of theft or terrorism. The gait of individuals checking in at an airport could be compared to an existing database, perhaps even before they enter the airport concourse or during initial security checks.  Such data compared with existing CCTV footage could be used to track suspect terrorists or criminals who may be disguising their features or carrying forged documents. A key advantage of gait identification is that identification at a distance is possible where the face and/or iris are not visible [1].  The ability to identify a possible threat from a distance provides a longer time frame in which to react before a possible suspect becomes a real threat without the knowledge of the suspect.  An ideal gait identification biometric example is that of CCTV surveillance technology where there is a need to capture biometric data without human contact or cooperation at a distance.  A recent case described where “one man was convicted of a burglary after podiatrists compared CCTV images of him on his way to commit a crime with images of him in custody” [2].

Most of the literature shows that existing appearance based gait feature representation methods suffer from clothing and carrying object covariate factors. Recent research at Ulster has proposed new gait features with promising potential to address the issue of clothing and carrying objects (covariate factors) [3, 4].  While these methods provide a good recognition rate there is still a possibility of obtaining a better recognition rate by using better appearance based gait feature representations.  This project aims to extend Ulster’s work to further create biometric surveillance solutions based on gait recognition using well established modeling techniques.  The idea behind this project is to identify key components in a person’s gait or motion and  develop new model(s) for characteristic, descriptive properties resulting into enhanced recognition rate.

Personnel Involved

First Supervisor: Condell, J Dr
Second Supervisor: Prasad, G Dr

Collaboration: This project does not involve collaboration with another establishment

Synopsis:

Most of the literature shows that existing appearance based gait feature representation methods suffer from clothing and carrying object covariate factors. Recent research at Ulster has proposed new gait features with promising potential to address the issue of clothing and carrying objects (covariate factors). While these methods provide a good recognition rate there is still a possibility of obtaining a better recognition rate by using better appearance based gait feature representations. This project aims to extend Ulster's work to further create biometric surveillance solutions based on gait recognition using well established modeling techniques. The idea behind this project is to identify key components in a person's gait or motion and develop new model(s) for characteristic, descriptive properties resulting into enhanced recognition rate.

Return to list of PhD Opportunities