Video Surveillance Systems
- 19.1 Occupancy, People, Counting and Energy
- 19.2 Video Smoke Detectors
Video surveillance systems are used to deter crime and identify criminals when a crime has been committed. A video surveillance system can monitor stores and stock, provide a visible presence that video cameras are used in a building, allow building management to see what is happening at any time of the day, identify exact times when crimes have been committed, provide an identification method by which people can be screened before entering a building, and allow security personnel to check who is in a building at any given time. A complete video surveillance system consists of cameras, a control station, servers, hardware, operator work stations, software, cable, infrastructure, and junction boxes.
Video surveillance systems should be integrated with intrusion detection, access control, and electronic personal protection systems. The surveillance system may be comprised of both fixed and pan, tilt and zoom (PTZ) high resolution color cameras for monitoring the perimeter and interior of the building.
When we think of analytics related to building systems we generally think of predictive analysis or fault detection and diagnostic software tools related to HVAC systems. Video
Figure 19.1 analytics, that is software that can analyze and identify people, objects, and events in many ways can be just as important in providing information on building use and performance. Video cameras can be multifunctional. The analysis of digital images addresses physical security, but, goes way beyond that to provide data and information for building life safety, energy management, and overall building performance. This one device, the video camera, has a variety of uses for sensing and gathering data about the building condition and performance. This is a good thing, as more high quality and relevant building data is critical in generating actionable information and is a key to better building management and performance.
If you assume that the video camera is an extension of the human eye, the analytical software is the extension of the human brain. Cameras currently available can detect smoke or fire, identify specific people, detect motion, determine if objects have been moved, and provide occupancy data including the actual number of people in a space. Generally, if you can develop a pixel template of the event or condition you are trying to track, the video analytic software can detect the event or condition.
The video analytic process starts with cameras capturing successive digital video images of a coverage area. The digital image consists of pixels, a contraction of the words picture element and the smallest element of the
digital image. The analytic software first analyzes pixels, their patterns, the adjacency of pixels, the changing of pixels over time, and then compares the pixels to a database of templates of objects, conditions, and events. When the software gets a reliable match between the digital image of the coverage area and its database of templates or conditions, the video system identifies an event, state, or situation.
Video cameras are a staple of physical security systems. In the past, you typically had a security operations center where personnel viewed the feeds from the cameras and subjectively determined whether an event or action had taken place that warranted action. One of the largest benefits of using analytics in a typical video surveillance security system is improved detection and identification of threats, conditions, and events (machines outperforming humans). The software is working 24/7 with a constant level of accuracy. Also many video surveillance operations are not real time, with video simply being archived, available for search and review after an incident. Even if the system is manned, the attention span of personnel in a security operations center is oftentimes very short and inconsistent.
The analytic tools related to video cameras are extensive. As one would expect, most are geared towards some aspect of security and they include:
b Facial recognition: Video cameras can be used for recognizing people who are then given access to a building or space.The camera can recognize people that may be threats.
b Motion detection: The analytics can detect motion within a camera's coverage area, triggering an alert.
b Missing objects: By comparing digital images the analytics can detect if objects are missing or if a new object has been placed in the coverage area.
b Reading license plates: The cameras can read license plates to determine if particular vehicles have access to parking garage or building.
Some of the more innovative and interesting aspects of video analytics are people counting for occupancy and using video as a detector of fire and smoke. The following section contains an overview of these two applications.