A game theory-based block image compression method in encryption domain

With the development of digital imaging technology and the prevalence of mobile devices with camera, internet privacy has been a growing concern for public. Especially in some private chat groups in social network, for example, Facebook, WeChat, and MySpace, some people want to share their personal images by internet service provider without leaking this information while the internet service provider can also take some additional measures on these images, for example, reducing the network bandwidth. How to provide technologies or software with such functionalities, usually conflicting goals, becomes increasingly urgent in both academia and industry. Image encryption is a choice; however, it does not provide the additional function needed by internet service providers. Recently, game theory is widely used in network security to solve such problem with conflicting goals. In fact, there is a game theory between users and service providers. This paper proposes a block-based image compression method in encrypted domain which provides not only the privacy protection capability but also the additional operation capability needed by internet service providers. This block-based method can be formulated as a game theoretical problem and can be optimized by game theory. First, the image to be shared will be encrypted in block-by-block way by owner with simple encryption operation. Second, the service providers can send the part or full of the encrypted image stream according to the available bandwidth with an adaptive parameter sets. Finally, the intended receivers can decrypt the received encrypted stream to recover the image. Extensive experiments show that the proposed algorithm improves the compression performance compared with the-state-of algorithms.