Machine learning enhances video surveillance at outdoor events

Video surveillance plays a critical security role at large outdoor events. Advances in AI and machine learning can offer enhanced tools for spotting potential threats.

Facial and behavior recognition technologies are improving video analytics, providing invaluable new tools to security professionals. Stephanie Weagle, Chief Marketing Officer at BriefCam, a company that specializes in using cameras and video analytics for security, said video-analysis technology could be trained to recognize “signs of problematic behavior,” as well as keep an eye on unauthorized entry, exceedingly high foot traffic or loitering.
 
Such video technology can be integrated with other services to trigger follow-on actions after potential threats are detected. Proactive alerting, triggering calls to action in response to suspicious or unusual behavior, can enhance the overall efficiency of a security team.
 
Advanced video surveillance tools increased “situational awareness,” Weagle said. By mounting cameras with behavioral-recognition technology in various locations, security personnel can have a better grasp of what’s going on in the area.
 

Weagle added that the continuous, ongoing surveillance provided by cameras also minimized the risk of “human error and distraction.”
 
Facial recognition plays a critical role in video analytics. Proliferation of better-quality cameras and advances in video analytics had helped to improve the accuracy of facial recognition, Weagle said.
 
One such advance is the ability to store photos of suspects or past offenders on a video-analytics engine and alert security personnel when such individuals are spotted by a camera. Security personnel can then be dispatched quickly to deal with the potential threat. 

Weagle noted facial recognition technology could not only be used to identify criminal suspects but also VIP guests, “to ensure important visitors are being properly engaged and ushered around as needed.”
 
As facial recognition is still relatively new and not yet widely adopted, countries have different policies and rules on the technology’s use.
 
Nonetheless, event organizers should consider “how the technology can be used in ways that ensure fairness, transparency, accountability and privacy compliance, while maximizing public safety,” Weagle said.
 
Data collected by video sensors could also help to “map out activity trends and discover insights” to improve security readiness and response, she added.
 
One challenge facing event organizers is vehicle and crowd congestion creating bottlenecks at entry and exit points.
 
Video analytics could address this problem by collecting traffic flow data, before coming up with optimized pedestrian or vehicle navigation routes to ease the traffic. This would improve the guest experience and enhance security, Weagle said. She added it would also enable “evacuation and security dispatch routes to accelerate emergency responses” to be planned in advance.
 
“This information can also be used to understand staffing requirements, ensuring that crowds are guided through the event, queues and bottlenecks are prevented and employees are upholding safety regulations,” Weagle said.