Requirements for stereovision-based access control for physical spaces

Computer vision allows for computer systems to see, react and interact with the environment in which they are deployed. Current computer vision systems are based upon aging identification techniques such as blob and feature detection which can no longer be used in isolation to reliably identify actors within a given environment due to their two-dimensional limitations. Stereovision is becoming a viable computer vision technology for access control systems due to the growing processing power available for image processing that enables the use of multiple cameras to compute the third dimension of depth to aid in detection of objects within physical spaces. This paper evaluates the current state of computer vision with respect to access control for physical spaces and draws up a set of requirements for effective access control to be enforced using stereovision as a basis for the computer vision aspects of subject identification on the African continent. This paper then proposes an access control model for physical spaces based upon the three-dimensionality of stereovision to provide a more robust and reliable form of identification for actors within a system whereby access to physical locations should be limited.