ParaScale Video Surveillance Systems

Many businesses, and even entire cities such as London, have installed video surveillance systems to record and subsequently mine video. A large-scale video surveillance environment includes thousands of video cameras writing streaming file content to a back-end storage system.

Three challenges dominate video surveillance.

First, the challenge of managing the huge volume of video files generated by video cameras, especially today's increasingly affordable HD cameras.

Second, due to the enormous volume of video, software mechanisms for automated processing (e.g., video analytics like license plate recognition and facial recognition) are essential to enable human attendants to comb through extraordinary amounts of data to identify problems and exceptions.

Third, certain videos are relevant long after they have been captured. This requires sufficient video retention capacity to avoid premature destruction of valuable imagery. If an organization undertakes the expense of capturing a video, it makes sense to retain it as long as practical.

ParaScale is ideal for meeting each of these requirements.

First, our global namespace helps managing hundreds of disks assembled for video repositories. A global namespace allows centralized management for rapid discovery and sharing of important video.

Second, ParaScale's open architecture permits video processing applications to co-exist with ParaScale software on Storage Nodes(SNs) eliminating the requirement for a separate video processing infrastructure. Moreover, processing local video resident on the SNs locally, reduces network and hardware requirements.

Third, ParaScale leverages the industry's most cost-effective hardware, therefore, storage is cheaper and the same budget can retain more historical video. Moreover, ParaScale allows centralized control, where administrators can set different video retention policies for different cameras, even varying retention by time of week. Alternatively, video analytics on the SNs can sample video, reducing either the resolution or frame rate as videos age. This approach, made possible as the SNs are open computing platforms (standard Linux x86 platforms), dramatically improved historical access to video with minimal impact on overall system performance and cost