Privacy-Preserving Watch List Screening in Video Surveillance System

摘要

Increasing number of surveillance cameras in public places raises legal issues and demands for a solution to protect individual privacy in video surveillance systems. In this paper, we propose a privacy-preserving watch list screening system that guarantees both security of public places and privacy protection in a video surveillance scenario. The proposed watch list screening system is proposed to discriminate groups of identities of interest without confirming the precise identity of people in order to ensure individual anonymity. Further, to prevent the revelation of private information, such as facial images, the proposed system utilizes a homomorphic cryptosystem, enabling secure computation in the encryption domain. Several experiments were conducted to demonstrate the feasibility of our proposed system using face images, considering verification accuracy and complexity aspects.