Graphical model-based privacy policy editing for smart video surveillance

Poor usability and, as a consequence, human errors can render powerful security and privacy mechanisms useless. Borrowing from the concept of visual programming we introduce a graphical editor for authoring privacy policies for smart video surveillance systems, i.e., systems involving computer vision technology to some extent. We built this editor upon a meta model, which we derived from a generic architecture for smart video surveillance. Employing our tool, privacy-related requirements can be assembled from readily understandable graphical blocks and exported into machine-readable usage control policies. We give a demonstration based on a smart video surveillance system employed for fall detection in medical facilities.