The importance of 'scalability' in video surveillance architectures

The choice of the video coding technology strongly influences the design (and costs) of digital video surveillance architectures. Today, these limitations are no longer there and digital video coding is required to provide additional functionalities in addition to pure compression performances: e.g. low delay encoding-decoding, fast random access and "scalability". Scalable video coding schemes are designed to ease the simultaneous access to a large number of video sources from multiple clients (users), each with different constraints and requirements, through heterogeneous digital networks, each with different quality of service (QoS). Instead of multiple encoding each video source so as to provide the optimized bitstream to each client, scalable coding provides a unique bitstream whose syntax enable a flexible and low complexity extraction of the information so as to match the requirements of different devices and networks. Recently, scalable video coding schemes have been successfully installed on large video surveillance infrastructures such as metropolitan public transportation networks (e.g. Paris and New York Subways), airports (e.g. Toronto, Zurich) and highways (e.g. Portugal, Bern). Recently, to answer a real industrial need, the JVT a joint MPEG (ISO) and VCEG (ITU) video team is finalizing the standardization of the MPEG-4 SVC: the scalable video coding extension of MPEG-4 AVC. In this paper we introduce the principles of scalable video coding and discuss its importance to answer the requirements of several typical applications in video surveillance architectures. A brief overview of the current status of the JVT standardization process is provided.