Artificial Intelligence Enhances Campus Video Deployments—and is Becoming More Affordable – Campus Security & Life Safety Magazine
Artificial Intelligence Enhances Campus Video Deployments—and is Becoming More Affordable
Analytics that use AI can help identify people (instead of a curious animal) loitering around a campus at 3 a.m. It can assist in investigations: quickly finding, for example, a burglary suspect…….
Artificial Intelligence Enhances Campus Video Deployments—and is Becoming More Affordable
Analytics that use AI can help identify people (instead of a curious animal) loitering around a campus at 3 a.m. It can assist in investigations: quickly finding, for example, a burglary suspect reported as wearing a yellow shirt. It delivers invaluable situational awareness since campus police cannot be everywhere at once to monitor for incidents.
While many in education have the desire to be early adopters of new technologies, budget constraints and path dependency (sticking with the same old technology, or none at all, because it’s too difficult to change) are often obstacles that interfere with deploying the latest advancements.
Campuses should, however, consider incorporating artificial intelligence (AI) and analytics into their video surveillance solution. AI has a variety of uses that can benefit all departments, not just security—making it justifiably affordable.
Technically Speaking: How is AI Used in Video Surveillance?
Video analytics uses AI and deep learning to create searchable, actionable and quantifiable intelligence from live or recorded video and develop rules on how to respond. This means that AI can use video footage to recognize and extract objects (as well as information) about the type and attributes of these objects. These can include people, loiterers, numbers of people and people counting, cars, animals, travel direction, and the like. Alerts are sent to a preconfigured recipient, helping quickly notify a responsible party who can then act on the information.
Video analytics can be run on the edge at the device (for example, in a deep learning video surveillance camera) or can be added to a video management solution via an integration. An advantage of using edge computing in the camera is that the information processing power sits as close to the source as possible. Edge analytics tends to have both greater accuracy and the ability to distinguish between multiple classes of objects, which immediately reduces the rate of false positives and saves unnecessary investigation time. As a result, edge analytics can deliver a more appropriate and timely response.
In a traditional model—when analytics takes place on a server—video is often compressed before being transferred. However, this often results in the analysis being undertaken on video of degraded quality. Additionally, when analytics is centralized on a server and when more cameras are added to the solution, more data is transferred, and more servers are needed to handle the analytics. Deploying powerful analytics at the edge means that only the most relevant information is sent across the network, reducing the burden on bandwidth and storage.
This article originally appeared in the March / April 2022 issue of Campus Security & Life Safety.
