Key Tech Trends
How modern video analytics make an impact on the industry
- By Scott Dunn
- Dec 01, 2022
Video analytics is in the midst of a revolution. For decades, the video surveillance goal was primarily to record video for later review in the event of a security-related incident. True, there might be a security guard keeping an eye on a handful of wall monitors to look for signs of suspicious activity, but by-and-large the technology was reactive. Of course, that has all changed in recent years, and the advancement of video analytics has made security technology more proactive, alerting on incidents in real time and allowing security personnel to respond quickly and effectively.
That technology continues to evolve and improve. Those improvements have come in the form of better camera (image) quality, more powerful computing at the edge, greater cloud integration and expanded use cases that go beyond security—all of which has played an important role in changing the way today’s organizations use video analytics. As we move into 2023, those changes are worth keeping an eye on—both because of how they have already affected the industry, and how they promise to affect change it in the future. As the analytics revolution continues, a number of key analytics trends that will change the way businesses operate in ways that go far beyond security.
Trend 1. Hybrid Deployments Will Continue to Grow
The advent of cloud computing was a significant development for analytics, allowing organizations to take care of vast amounts of computing power beyond their on-site capabilities. For video analytics, cloud was somewhat limited in its application—after all, sending an entire video feed to the cloud for analysis would require vast amounts of not only bandwidth, but storage space as well. Fortunately, the recent development of Deep Learning Processing Units (DLPUs) has dramatically increased the processing power of modern IOT Devices such as cameras, allowing them to conduct onboard processing and run analytics natively at the network edge. As a result, only the relevant metadata (and maybe a few seconds of video) needs to be sent to the cloud.
What does this mean? In practical terms, it means that key decisions, like whether to issue an alert or trigger an automated recording, at the edge, while the cloud is often used to conduct further forensic analysis. This provides for faster situational awareness and better sharing of the computational workload. Major cloud providers realize this. They have shifted operations to process actionable insights at the edge, using cloud computing to engage in trend analysis and pattern recognition. The result is that hybrid deployments have become increasingly popular, with organizations relying on both edge and cloud processing to generate the real-time and long-term analysis they need. The benefits are clear and we will see greater adoption in the future, and evolution, of hybrid computing.
Trend 2. The Metadata Revolution Is Here to Stay
This plays into the hybrid deployments trend, but it is also important to recognize the specific role that metadata plays. Processed at the edge, video is tagged and categorized with key identifiers. This includes vehicle color, license plate number, direction of travel, time of day and more. All of the information about the scene is now in once place, and if something strikes the system as interesting or suspicious, a human operator can go directly to the edge device and pull the specific video clip they want to see (or even a live feed), rather than sifting through hours of video.
This has obvious security uses, but it also allows businesses to leverage this technology for operational efficiency and business intelligence purposes. By tracking the number of people in a store, how they travel throughout the space, what products they gravitate towards, and other insights, retailers can improve the customer experience in significant ways. That is just one use case—as this technology becomes more common; we can expect to see more industries leverage metadata to generate valuable insights to further both their business and security goals.
Trend 3. Computer Vision Will Become the Norm
The trend toward embracing computer vision is one that will only continue to grow. Computer vision, essentially, is a set of eyes that can analyze what is happening within its field of vision. Because today’s cameras have perception capabilities that far exceed the human eye, computer vision can be used to implement and monitor safety and quality assurance protocols more accurately and effectively.
A bottling company no longer needs to rely on a human to notice a half-filled bottle—a camera armed with computer vision will see it every time. Likewise, a camera equipped with thermal monitoring capabilities can track the temperature of machinery and alert if it becomes too hot or too cold.
Here, again, DLPUs are essential—andd ubiquity is making computer vision technology commonplace. In the past, an organization would need expensive and highly-specialized cameras in order to use computer vision.
Today, DLPUs are not just for high-end cameras—they are becoming the standard. That means just about anyone with a truly modern surveillance camera can leverage computer vision for a wide range of uses. Better still, it reduces the number of devices organizations need to deploy, as the same camera they already use for security monitoring is now used for quality assurance and other purposes. There are also devices today that embed the DLPU functionality into their own chip which can minimize device size and mitigate heat issues at the same time. Innovation continues at accelerated speed.
Trend 4. Interoperability Will Make Organizations More Efficient
As computer vision allows companies to use individual devices for multiple purposes, organizations will be able to allocate resources more efficiently. Security teams, facilities teams, marketing teams, and others can all use the same camera deployments for wide-ranging purposes. This means that an organization can reduce the number of devices in use, and it can draw funding for those devices from across multiple budgets. This might not seem like a big deal on its face, but in a world where competition for limited resources can be fierce, the ability to use a device already being used for security purposes to conduct marketing research or monitor for maintenance issues is a major development. It can help organizations ensure that all departments are pulling in the same direction—a trend that they will certainly want to continue.
Trend 5. Mobile and Temporary Video Surveillance Will Grow in Popularity
Pop-up events and other temporary attractions are a regular occurrence in cities all around the world, ranging from craft fairs and outdoor concerts to marathons and other major sporting events. In days gone by, it was difficult to secure those events, after all neither cities nor event organizers are likely to spend the money to deploy permanent security for a temporary event. This can be a problem, especially for a major event like a championship game, which might attract thousands of people outside the event—let alone inside of it. Special events, entertainment options, food vendors, and more are popping up all the time, and the ability to extend the reach of security to encompass them is critical.
Fortunately, improved wireless networks (and 5G technology) have made temporary security deployments much more feasible. Today, event planners can deploy thermal sensors, PTZ cameras and other security devices using temporary accommodations that can be erected, and dismantled in just a few days.
Because it is now easier to transmit information wirelessly to venue owners, event organizers, and law enforcement, it is more feasible to keep temporary events secure. Since those events are not going anywhere, look for temporary security measures to grow in popularity over the coming years.
Looking Ahead to the Future of Surveillance
Surveillance technology has evolved, and so has the way organizations put it to use. Modern analytics can do far more than simply keep people and property safe—they can provide critical business intelligence insights, giving businesses the information they need to better serve their customers and improve their bottom line.
Computer vision and metadata analysis are helping those same businesses generate and combine even more valuable data, with new analytics and use cases emerging on a seemingly daily basis. Finally, nontraditional security deployments are helping cities keep pop-up shows and other temporary events safe, thanks to improvements in wireless data transmission.
These developments are here to say, and they set the stage for the next several years. As developers continue to dig into the possibilities that modern surveillance technology has unlocked, these trends promise to only grow more interesting—and more exciting.
This article originally appeared in the November / December 2022 issue of Security Today.