Privacy Preserving: Hiding a Face in a Face

摘要

This paper proposes a detailed framework of privacy preserving techniques in real-time video surveillance systems. In the proposed system, the protected video data can be released in such a way that the identity of any individual contained in video cannot be recognized while the surveillance data remains practically useful, and if the original privacy information is demanded, it can be recoverable with a secrete key. The proposed system attempts to hide a face (real face, privacy information) in a face (new generated face for anonymity). To deal with the huge payload problem of privacy information hiding, an Active Appearance Model (AAM) based privacy information extraction and recovering is proposed in our system. A quantized index modulation based data hiding scheme is used to hide the privacy information. Experimental results have shown that the proposed system can embed the privacy information into video without affecting its visual quality and keep its practical usefulness, at the same time, allows the privacy information to be revealed in a secure and reliable way.