Source: Campus Safety/Fredrik Nilsson – Think of it this way: to centrally process analytics on a 20-camera server deployment, a campus would need a very high-end processor, like a quad core Xeon. By comparison, if they handle analytics at the edge, all they’d need is a Core 2 duo, which is really a desktop motherboard. The cost differential could be several thousand dollars. There are the less obvious costs as well: PC warranties are generally cheaper than server warranties and it costs more to service a server than it does a PC.
Finally, there’s a green element to all of this. If you put analytics on the server, you need more energy to power and cool the equipment. With in-camera analytics, you’re reducing your power consumption and saving BTUs on air conditioning.
Kattan: We have seen that the adoption of IP technologies in general has accelerated the adoption of video analytics. In most professional camera installations, there is a need to record the video and store it for certain periods of time. The ability to implement the analytics in the camera can actually smooth the process of pre-processing, or tagging, relevant video segments, and in some cases, reduce the flow of “uneventful” video over the network. The ability to store only relevant video segments typically saves tens of thousands of dollars in storage costs over the lifetime of a surveillance system.
Q: Nilsson: Being able to tag and quickly retrieve video footage is essential. In an emergency or criminal situation, it can help drastically reduce the time of suspect apprehension from weeks to hours or even minutes. Specific to the education and healthcare communities, what are some of the more commonplace uses for video analytics?
Kattan: At the most barebones level, many campuses use basic motion detection – monitoring building perimeters to protect against breaches to the premises and property. The analytics can help security personnel detect crowding, which may indicate a fight, an injury or safety hazard.
Thanks: Campus Safety/Fredrik Nilsson