Video forensics in cloud computing: the challenges & recommendations

Forensic analysis of large video surveillance datasets requires computationally demanding processing and significant storage space. The current standalone and often dedicated computing infrastructure used for the purpose is rather limited due to practical limits of hardware scalability and the associated cost. Recently Cloud Computing has emerged as a viable solution to computing resource limitations, taking full advantage of virtualisation capabilities and distributed computing technologies. Consequently the opportunities provided by cloud computing service to support the requirements of forensic video surveillance systems have been recently studied in literature. However such studies have been limited to very simple video analytic tasks carried out within a cloud based architecture. The requirements of a larger scale video forensic system are significantly more and demand an in-depth study. Especially there is a need to balance the benefits of cloud computing with the potential risks of security and privacy breaches of the video data. Understanding different legal issues involved in deploying video surveillance in cloud computing will help making the proposed security architecture affective against potential threats and hence lawful. In this work we conduct a literature review to understand the current regulations and guidelines behind establishing a trustworthy, cloud based video surveillance system. In particular we discuss the requirements of a legally acceptable video forensic system, study the current security and privacy challenges of cloud based computing systems and make recommendations for the design of a cloud based video forensic system.