Security

Vision products have been used in the physical security market for a number of years. The technology is commonly referred to as video content analysis (VCA) or intelligent video in the physical security industry. It is used in virtual tripwire applications, providing perimeter intrusion detection for high-risk facilities such as airports and critical infrastructure.

Typical projects use video analytics to generate real-time alerts when an object is identified and tracked through a pre-defined area. Operators, based in the central control room, can manage more video streams as they are not required to identify intruders purely from watching the monitors. Instead, the vision product will identify and alert the operator when there is a potential security issue. It is then up to the human operator to decide how best to deal with the threat.

Other applications of vision products in the security space include wrong-way detection and trajectory tracking. These types of algorithm can be used in airport exit lanes, where security managers want to identify people entering the restricted area from the main terminal. This responsibility has historically been given to security guards; however, vision technology can reduce security guard costs for airport operators and increase the accuracy of the security system.

Other applications in the security market include behavior recognition algorithms such as “slip and fall” and abandoned/left object detection. These applications are less developed than the perimeter detection systems, but will provide powerful tools for security managers over the coming years.

What are the primary vision products used in security?

The primary vision products in the physical security market are network cameras, encoders, NVRs (network video recorders), DVRs (digital video recorders) and intelligent appliances. Established video surveillance vendors supply the majority of these products. These vendors either develop analytics in-house or embed third party algorithms from video content analysis software developers. Consequently, growth in the market is heavily driven by what these vendors decide to do and the extent to which they commit to intelligent video.

Over the last couple of years, the video surveillance market has also seen “low-end” analytics, such as camera tamper detection and video motion detection, become a standard feature on many vendors’ product ranges. These applications have developed from a distinct video analytics market to a product feature used as a differentiator. “Low-end” analytics are embedded on cameras, encoders and storage devices.