For electrical contractors installing video surveillance and networked Internet protocol (IP) cameras, now is a great time to think outside of the proverbial box and look beyond the typical security applications of viewing and recording images of people, places and things. A host of opportunities is emerging as video analytics (also referred to as video content analysis) offers more power, greater reliability and easier deployment. All of this equates to being able to offer customers additional functionality, which enables them to use their video cameras for more than security—such as gathering actionable data and business intelligence on the company and its operations.


Essentially, video analytics is computer software that scans the pixels collected from camera images to provide metadata. The software behind video analytics can be configured to recognize certain features and characters and respond based on those visual cues. They are referred to as “rules,” and the exceptions to those rules trigger analytics functionality.


Analytics’ primary purpose is to hone in on alerts and exceptions without requiring staff to watch video 24/7, a task for which humans are ill-suited, being prone to error and fatigue. Without requiring additional manpower, analytics simultaneously examines numerous video feeds or locations. Basically, analytics offers the same observations as humans, only better, faster and more accurate.


Reliability guides mainstream applications


Video analytics is far from new, but, when it first came on the security scene, it got a black eye because of unreasonable expectations fostered by overexuberant development companies. It was prone to false alarms (false positives) and was difficult to deploy, especially outdoors. Recently, that has changed. Video analytics and its corresponding software have matured, mining and culling data more efficiently. Increased computing power is responsible for many of these improvements. Software algorithms have also become more sophisticated and more customizable.


Central processing units are more powerful and less expensive, and video analytics have become precise in determining what’s happening, what a person is doing and what the specific target is, according to Eric Olson, vice president of marketing, PureTech Systems Inc., Phoenix. 


“Greater intelligence equates to the same amount of coverage or better with fewer cameras and smarter information for less cost,” he said.


PureTech Systems is a software company that develops video analytics algorithms, and Olson said the future is putting metadata on every image frame—determining what’s on the frame, what direction it is moving and other information, such as whether the object is male or female. 


“We’ve also added another dimension, which is geospatial reference, especially useful in security applications,” he said. “We know the exact longitude and latitude of the object or person for forensic searches.”


The market for video analytics continues to hit its stride. According to Homeland Security Research (HSR), Washington, D.C., the video analytics industry is forecast to experience decades of rapid growth.


HSR’s report “Global Video Analytics, ISR, Intelligent Video Surveillance and Object Recognition Market 2015–2020,” indicates that the global intelligent video surveillance, intelligence, surveillance and reconnaissance and video analytics industry revenues will grow at 14 percent compound annual growth rate from 2014–2020. The report attributes the increases to the following:


• Security-related video surveillance boom


• The rapidly growing Internet of Things applications is driving the “Video Imaging of Everything.”


• CCTV cameras cost-performance is following Moore’s Law—
high-resolution camera prices are dropping significantly.


• Technology maturity: video analytics algorithms, processors, applications and products underwent a decade of technological evolution to intelligent video processing, based on advancements in image processing, enabling automatic and semiautomatic detection and identification of signatures.


• Cost reduction of video analytic systems is driven by the falling prices of image processing/digital signal processing (DSP) and communication systems.


• Improved cost performance of new edge-based video analytics DSP technologies


How analytics work


There are two primary methods of collecting data using video analytics—at the camera level, also referred to as “at the edge,” and server-based applications. In addition, some technology partners are moving to cloud-based services, with analytics hardware and software rented out of a data center.


Cameras and their increasingly prolific SD storage cards have become small, powerful computers and are often used as motion sensors to trigger analytics. The most popular types of analytics include object detection/classification, license plate recognition, loitering/crowd detection, and wrong direction/speed threshold.


This greater processing power and enhanced stability is opening new areas of specification for installing contractors. Now, small and medium businesses (SMBs), as well as enterprise customers, can deploy analytics and harvest data for security and other uses. The retail market is one of the emerging, primary users of video analytics technologies, driving more widespread use. Retail and other SMBs are now deploying analytics for customer behavior insights, employee training, identifying key performance metrics, leveraging real-time business intelligence, and assessing store compliance and safety.


One area in particular taking off in the retail market is activity heat mapping and dwell times. Video analytics can create visual reports and dashboards that provide a snapshot of patron density. They can provide information on where people in a store or location are going on an accumulated basis as well as their behavior in those areas. Heat-mapping shows aggregated periods of time and activity around certain end-caps or different areas where people congregate, along with a path analysis of where they traveled.


Another trending area in video analytics applications is license plate recognition (LPR) technology. Retailers with drive-up windows are using LPR to get orders ready. Typically, the retailer has a mobile application to place orders and cameras capture the license plate on the first entry so that, by the time the customer enters the drive-through facility, their order is ready to go.


Facial recognition used indoors and with a fixed population may also represent an untapped opportunity—for example, finding children lost in a mall. Facial images are captured with a high-resolution camera at the entrance or exit in the event children wander off.


Moving to business intelligence


Rick Spillane, senior vice president of global sales for 3VR, San Francisco, said both the retail and banking industries are hot markets for video analytics. While loss prevention and marketing already have a traditional vested interested in the data from cameras, he said that operations departments are also looking at analytics to promote greater in-store efficiencies.


“Using heat maps, operations can tell what times of day a store or location may be most active and in which areas,” he said. “They can use this to determine when they may need additional cashiers, for example. With that information on traffic, they can study other anomalies on what’s actually happening in the store environment. So in addition to installing product to protect against loss or theft, they can use the same money spent on cameras and analytics to extract even more data and intelligence—­sharing these tools and the budget for deployment across the organization.”


Spillane said analytics have become much easier to use from the installer’s perspective, with the ability for the user to determine, through the software, what’s important to them or the specific data that needs to be extracted from the environment.


“Programming used to be much more labor-intensive,” he said. “We’ve provided sophisticated algorithms that are extremely easy to use. 3VR satisfies the end-user with the robust way analytics perform and also the integrator by making the product less labor intensive.” 


He added that video analytics is a great opportunity for electrical contractors.


“Today’s software is no more complicated than controlling lights; it’s very accessible and extremely easy,” Olson said.


He added that being able to share the data gleaned from analytics by different parts of an organization for various purposes provides lower total cost of ownership and a sales advantage to the installing company.


Electrical contractors installing networked video cameras can generate additional profitability by deploying video analytics beyond security and reaching into areas like business intelligence and operational data. Being able to offer more services beyond security only strengthens the contractor’s value proposition with the customer.