A new generation of video analytics that transcends security and encompasses safety, business intelligence and logistics is arriving. Software and analytics more readily integrate with popular video management platforms and are flexible in application without having to write specific IT programming code. For systems integrators, analytics bring the possibilities of exciting new revenue streams and future-ready technology to surveillance projects.
Intelligent video surveillance and edge cameras with onboard processing, as well as refinements in artificial intelligence (A.I.), machine learning and computer vision have propelled analytics into new use cases—with rapid-fire deployments occurring across the globe.
Cameras have evolved into detection devices to provide immediate, proactive insights tailored to security, safety or business challenges. These devices are mini-computers in their own rights and are able to process scenes more effectively and detect and alert on specific rules and parameters. Event-driven video analytics from A.I.-based software have matured to actively assist monitoring operations in identifying dangerous behaviors, assessing situational awareness in real time and reducing false alarms.
Mainstreaming analytics
Armed with analytics, surveillance systems provide a variety of insights beyond security, including occupancy and usage, employee safety, people counting, hotspot detection and traffic flow. A.I. in connected cars can ease traffic jams, as witnessed in a live study during an experiment of the U.C. Berkeley-led CIRCLES (Congestion Impacts Reduction via CAV-in-the-loop Lagrangian Energy Smoothing) consortium. Designed to reduce instabilities in traffic flow, the project leverages the cloud for traffic conditions to create an overall speed plan. That plan was to broadcast to cars, which used A.I. algorithms to determine the best action to take.
A.I. is at the core of analytics, according to research firm Statista, which projects the market to reach over $100 billion by 2028. According to Statista, A.I. refers to a computer or machine’s ability to mimic the competencies of the human mind, with the current ecosystem consisting of machine learning, robotics, artificial neural networks and natural language processing. These features and algorithms are highly versatile and adaptable to users’ specific requirements, state Statista researchers, suitable for many different industries and applications.
While previous iterations of analytics in the physical security industry were prone to false alarms during detection and were generally unstable, those inefficiencies are a thing of the past. Rapid A.I. technological development coupled with the digital transformation has resulted in a reliable product category.
Analyzing video footage derives meaningful and actionable insights from images. This data includes overt intelligence, a crime captured in real time, but also may contain intelligence that is not readily visible, according to “Future Thinking Analytics at Your Fingertips,” published by Lumeo, designer of computer-vision solutions.
Lumeo won a 2021 New Products and Solutions Award from the Security Industry Association for its “no-code” open, flexible video analytics platform. The platform enables integrators to create and deliver custom solutions in minutes—rather than having to select from predetermined analytics only, as was common in the past.
Smarter works harder
“Advances in A.I. and machine vision have enabled entirely new solutions; like vision-based parking sensors, underwater fish disease detection, media curation, construction safety intelligence and threat detection, to name a few. Advances in A.I. let you identify and solve new problems while increasing accuracy by learning from previous experiences. The ability to learn and improve over time is a key difference between A.I.-based analytics and the classic heuristic-based computer vision, which makes the former more robust and future-proof,” according to Lumeo’s white paper.
Outdoors, geo-enabled analytics improve detection in environments subject to movement by trees, debris, small animals and other anomalies, using A.I. to ignore these alerts while detecting actual events. This is extremely useful for surveillance applications in critical infrastructure and thermal imaging.
Geo-analytics use location-based information to analyze contextual awareness and gain different perspectives on the data. By incorporating geographic location and other spatial details, users get a broader understanding of data and uncover new insights. For example, geo-analytics can procure data to help monitor equipment for proactive maintenance or create visual maps for retailers to gather regional sales performance information or track customer behaviors.
Analytics have evolved with A.I. and are more stable and flexible in application—a definite benefit for systems integrators. With analytics, cameras become intelligent detection sensors that can alert facilities managers to unsafe or insecure situations. A.I.-powered data gives your customers insights beyond security to make better business and operational decisions—and that’s a value proposition that’s hard to ignore.
SHUTTERSTOCK / NET VECTOR
