AI AND VIDEO ANALYTICS BLOG
Video Surveillance & Physical Security Industry Viewpoints
July 15th, 2020
Author: Lizzi Goldmeier

Cities Have New, Smart Ways to Protect Health and Safety in Public Spaces

Video content analysis for driving compliance with health and safety mandates

As the COVID-19 pandemic evolves, many local and state governments have lifted stay-at-home orders and are conducting phased re-openings, to allow some business and social activities to resume. As the public returns to work and resumes socializing, people will come into closer contact in spaces such as parks, buses, trains, and sidewalks. Public health mandates, including physical distancing and face mask wearing, have been recommended to help prevent the spread of the disease. Monitoring and enforcing compliance has become yet another unprecedented challenge for public safety officers: As public and private facilities reopen, cities must be aware at all times of how many people are in a particular space – as well as the benchmarks for safe occupancy for that area – to comply with the mandates and, over time, understand where violations and problems tend to arise.

Smart Cities that rely on data-driven approaches to increasing efficiencies have somewhat of an advantage in this pandemic, because they already are leveraging technology to improve residential quality of life, optimize municipal operations, and secure the community. For instance, cities that already employ video content analysis to maximize their investments in surveillance networks, optimize traffic flows, increase public safety and drive communal wellbeing, can extend their usage for public health and safety applications: Whereas manually monitoring for compliance violations or contact tracing to curb virus spread would put undue pressure on law enforcement and city operators, intelligent video surveillance makes video searchable, actionable and quantifiable for accelerating investigations and video search; increasing situational awareness with real-time alerts; and deriving intelligence for operational decision making.

Powered by deep learning and artificial intelligence, the technology extracts, identifies, classifies and indexes valuable video metadata, so operators can proactively respond to developing situations, productively review footage, and uncover trends based on aggregated video data both for the current public health and safety considerations and for everyday city safety and wellbeing.

Increasing situational awareness with alerting

Once normal patterns are discerned or defined, video content analysis systems can be configured to detect anomalies. Real-time alerts empower law enforcement agencies to improve public safety through increased situational awareness, quick assessment of developing incidents and swift response coordination. This is helpful for detecting and responding to traffic violations, such as illegal U-turns, or to mounting traffic: For example, as vehicular congestion forms on city streets, operators can quickly investigate and understand the cause of the traffic build up and devise how to alleviate the congestion, whether through digital signage or deploying officers to redirect traffic, closing access points, or sending first responders in case of an accident – all before a full-blown traffic jam is formed.

For indoor and outdoor spaces, cities can also use count-based alerts to maintain safe human occupancy, as well, by setting up alerts to trigger calls to action when thresholds are exceeded. People-counting alerts can help operators proactively monitor and protect parks or public lobbies, for example, from becoming too crowded to the point that safe distances from others cannot be maintained. By alerting on people counts paired with identifying the proximity between people and the duration they stand closely together, cities can help prevent crowding and encourage compliance with social distancing, during the COVID 19 crisis. A video content analytics system can also be leveraged to detect and alert to when people are not wearing face masks – especially in crowded spaces or areas where crowds are forming.

Investigating incidents with speed and precision

Not all incidents can be proactively prevented; however, when crimes, emergencies or everyday events of interest transpire, municipal police need to be able to effectively investigate what occurred, quickly and thoroughly. With video content analytics, cities can accelerate law enforcement investigations and understand occurrences with the ability to rapidly search and filter video to focus on relevant information. Using object classification or face and license recognition capabilities, operators can limit video investigation based on known information and suspect leads and draw connections between people, places and events. During the COVID pandemic, video review has becomes especially important for contact tracing and helping notify anyone who interacted with someone confirmed to have Coronavirus. In this same way, this capability is leveraged every day to productively solve cases and support policing.

Collecting trend data and making intelligent decisions

Beyond understanding activity in the immediate aftermath of an incident – or while it is unfolding – video data can also be aggregated over time to generate long term statistical information and drive intelligent decision making. Through graphs, heatmaps, and histograms presented in dashboards, users can demonstrate and uncover critical patterns and trends. City officials can make better, more informed decisions, with actionable and quantifiable video intelligence.

One area where video data has a significant impact is urban and transportation planning:  To determine where critical infrastructure must be redesigned, replaced or improved, city planners can analyze traffic data captured by intelligent video surveillance throughout the city. In the same way, public transit managers can leverage this information, draw connections and make changes based on vehicular traffic patterns, pedestrians crowds and dwell durations at transit terminals and bus stops. From developing roadways and pedestrian pathways to relieve congestion to preventing bottlenecks and optimizing traffic flows, video content analytics help cities prepare and develop contingency plans for expected and unexpected activity.

Dashboard reports can similarly help cities understand to what extent the community is complying with public health and safety mandates, by providing accurate anonymous statistics, such as the average distance between pedestrians, face mask wearing compliance trends, and where and when violations tend to take place based on occupancy and face mask detection. By understanding whether visitors and residents are complying with recommended behaviors, cities can proactively prepare and – to the best of their ability – prevent outbreaks and help high-risk demographics stay informed about safe.

Hopefully, the COVID-19 pandemic will be relatively short-lived and, whenever normal operations resume, municipal agencies can continue leveraging video content analytics systems for traffic optimization, urban planning and crime prevention more than supporting public health. By analyzing the video data that is already accumulated with their extensive video surveillance networks, cities can derive much more value from their investment in those video networks.