Aesthetics is another benefit of megapixel products. Some campus officials prefer limiting the number of cameras deployed so they won’t encroach on a building’s decor.
Of course, storage is an issue with megapixel cameras. Fortunately, most campus applications don’t require a 30 images per second (IPS) recording rate. Frederik Nilsson, general manager for Axis Communications North America, has seen most healthcare and education surveillance systems using four IPS to 10 IPS rates. “There is also built-in intelligence that has the ability to scale down the IPS based on motion or activity in the scene,” he says.
This point leads to the next development…
Analytics and Other Intelligence are Less Expensive
The vast majority of time a camera is operating, it is recording non-event footage. Fortunately, many systems have analytics or motion detection that increase or decrease the IPS rate when appropriate, reducing the bandwidth being used when nothing is occurring. Additionally, analytics and other types of intelligence can flag events that deserve further scrutiny by campus police or security personnel.
“What we now see is that enough of that intelligence has shifted away from the PC and is being embedded in IP cameras and encoders,” says Dr. Bob Banerjee, IP video product marketing manager for Bosch Security Systems. “It will eventually be embedded inside people’s DVRs. We’re one step away from that.”
These advances have led to a decrease in the price of analytics. According to Banerjee, making a camera “smart” now costs $200-$400, as opposed to $4,000.
Lindsay Ryerson, product management leader for Honeywell Video Systems, says one effective use of analytics is graffiti detection. “That’s where we’re seeing analytics deployed because you’re looking for certain types of behaviors that are fairly well characterized,” he says. “You can increase the frame rate when you get someone who is loitering along a wall, for example, and get better quality images and more images of that person.”
Other applications include people counting and car counting at large scale events or in parking lots. According to Nilsson, analytics can provide 90-95 percent accuracy with these tasks. This solution can also help detect camera tampering, recognize license plates and track illegal parking.
