Persistent Authentication in Smart Environments using COTS hardware | Jorge Martinez
| | Abstract | This thesis studies a system that tracks principals from an authenticating point in the entrance trough his path as he moves in the building, this in order to provide authorization of services in the areas in which this user has the right privileges. Such authentication scheme allows the authorization of location-based services (for example automatic doors for restricted areas), in environments where low cost cameras are in use. The proposed system implements Persistent Authentication combining Blob detection techniques and different remote biometric factors using multiple Commercially Off-The-Shelf (COTS) cameras.
Background subtraction is performed on the video provided by the cameras, obtaining a shape around moving subjects in the environment, this shape is enhanced by different image processing techniques including Morphological transformations, providing a more complete model of the principals in the image. A series of biometrics is performed on the found model of principals, retrieving the identity made available during the principal authentication, finally the information resulting from multiple analyzed cameras is combined into one coordinate system giving as a result the position of each identity.
The implementation is used to study additional biometric factors not addressed by previous authors of Persistent Authentication and also to shine some light in the challenges presented by the usage of multiple cameras. | | Type | Master's thesis [Academic thesis] | | Year | 2017 | | Publisher | Technical University of Denmark, Department of Applied Mathematics and Computer Science | | Address | Richard Petersens Plads, Building 324, DK-2800 Kgs. Lyngby, Denmark, compute@compute.dtu.dk | | Series | DTU Compute M.Sc.-2017 | | Note | DTU supervisor: Christian D. Jensen, cdje@dtu.dk, DTU Compute | | Electronic version(s) | [pdf] | | Publication link | http://www.compute.dtu.dk/english | | BibTeX data | [bibtex] | | IMM Group(s) | Computer Science & Engineering |
|