Video Privacy in the Age of Automated Surveillance

Automated digital video surveillance has come of age and a new threat to personal privacy is posed. Traditional physical security coupled with information systems has slowly been eroding personal privacy. But the advent of automated video analytics and the latest integrated information systems will truly open Pandora’s Box. The possibility that your employer will know when you arrive at work, when you loiter in front of your workplace, and when you leave with a package is already here. The possibility that your local retail merchant will know when you shop, with whom you shop, and what you do when you shop is around the corner. The potential for the misuse of this information is of increasing concern. However, the same technology which can be used to compile personal information about you can also be used to protect your privacy. In this paper, we present several examples of automated surveillance techniques which can be used to limit the information garnered from digital video surveillance. Exploiting user-defined alarm conditions which can be arbitrarily complex, we show how privacy can be preserved using vision technology in situations where privacy concerns might otherwise be overlooked. Specific examples are given at locations such as industrial entry points, parks and schools, and a retail environment.