Abstract
With ever increasing volume of video data, ensuring effective access control to video content has become one of the most important and pressing problems in the domain of information security. The traditional access control mechanisms are inadequate due to the complex nature of video access control where the requirements are not only to restrict access to videos based on their IDs, but also ensuring security of content within the videos (e.g., people, objects and events). For example, even if two users have access to the same video, they might not have access to the same objects or sequence of frames.
While there is significant work in the area of video access control models, it is primarily limited to specification and does not satisfactorily address evaluation. In this paper, we present a novel way to overlay authorizations on the video context indexing structure to allow efficient evaluation of authorizations.