Port authorities manage the movement of people and cargo, whether it be at a shipping port, an airport, or train station. Physical security is a major element of port operations, because they are large facilities spread out over extensive complex estates, with many elements that need to be secured: There are public-facing parts to every port, such as areas for ticketing and baggage collection, restrooms, restaurants, retail stores and security clearance. But there are also offices, terminals, loading docks, customs facilities, and maintenance, storage and parking areas. These all add up to vast expanses that require vigilant patrol and monitoring, to protect not only workers and visitors but also port assets and the goods that move through those ports.
Many port authorities already rely on video surveillance systems for increasing overall physical security, deploying staff for real-time remote monitoring from a control room or archived footage review for post-incident investigation. However, the camera installations of port authorities are extensive; to truly actively monitor all video would require significant investment of personnel and time; and – even with enough resources to watch or review the overwhelming volumes of video – the outcomes would still be prone to human error and distraction. For these reasons and more, hours and hours of video surveillance footage are saved, but the footage is almost never reviewed or put to practical use.
For this reason, port authorities are starting to invest in complementary AI-driven video content analytics software, which identifies, classifies and indexes objects in surveillance video footage to make it searchable, actionable and quantifiable. Based on metadata catalogued using Deep Learning techniques, leading video intelligence software solutions can search by filtering objects based on their classes and attributes, including face recognition, appearance similarity, vehicle or person clothing color, and size. It can also detect objects’ speed, paths and their direction, and dwell durations.
The ability to detect, identify, extract and catalog objects in video drives various analytic activities, including, but not limited to: alert triggering based on object classification and face matching; video search based on object filters (person (man, woman, child), vehicle, animal, and their attributes (clothing / vehicle color, etc.); and long-term data aggregation to derive operational intelligence.
Video analytics software can be configured to send alerts to security or operations staff when certain conditions are met. For example, if the number of people in a security clearance area or maintenance building exceeds the normal limit, as set by the operator, security staff can be alerted and dispatched to assess that area. Similarly, it can be used to detect anomalous behavior, such as when a vehicle dwells for a longer duration than a pre-determined threshold, or if there is movement in a selected area where there is supposed to be limited activity.
Here are more examples of how port authorities can leverage video content analytic alerting functionalities:
In jurisdictions where facial recognition is allowed, video analytics can pinpoint people of interest in real-time using digital images extracted from video, external image sources and pre-defined watchlists. Real-time face matching enables human operators to receive timely alerts when an image on a watchlist appears; operators can then proactively assess the situation to determine whether a person of interest has been located and whether continued tracking or intervention is necessary. For example, port operators might compile a “watchlist” of digital images of suspicious individuals that would need to be proactively apprehended were they to appear at an airport, rail station or shipping yard. Whenever face matches are detected on site, the system can send an alert to the operator, who can determine how best to respond to the potential threat as the situation develops.
It’s important to track human activity in loading areas, to prevent theft and make sure no employees or unauthorized persons are in danger of injury. In a cargo container shipping yard or an airport maintenance facility, port operators need to be aware and intervene when unauthorized persons are detected, whether the trespassing is intentional or accidental. When the analytics technology detects a person excluded from the watchlist of authorized persons, a real-time alert can automatically be sent to security staff to notify of his or her presence in a secure area under video surveillance. Even if there are no real-time alerts, operators can at least extract the image of the unauthorized person from the footage, to use face recognition to match the image against other recorded footage and live camera feeds moving forward.
Face recognition is a helpful capability for quickly identifying persons of interest in video; however, it is not always possible to utilize this technology – whether because of regulatory limitations on its use or if the camera installed isn’t optimally placed or doesn’t record high enough resolution video to enable face recognition. In these cases, relying on a comprehensive video content analysis solution, as opposed to a singular solution for face recognition, can be an advantage. Operators then have the ability to filter video based on known object descriptions, such as, a child wearing a “pink jacket,” or an “adult male wearing a red coat.” The video intelligence system isolates the objects in video that match the classifications described in the search query or triggers a real-time alert notifying security staff of a match. This technology can help security or law enforcement staff in a transit hub to locate children who are simply lost, or it can help detect appearances of known kidnapped children, if they are brought through a train station or airport on a given day.
When reviewing archived footage, security and law enforcement investigators can easily find persons of interest in video and draw pivotal investigative evidence. Airports, cruise ship terminals, train stations, and shipping ports can benefit from this technology, because it can help security or law enforcement agents detect known suspicious persons, respond preventatively to criminal or terrorist threats and drive post-incident investigation when a security breach cannot be proactively prevented.
Shipping areas, whether for trains, planes or ships, as well as their cargo, must be monitored to prevent theft or tampering. Cargo container entry and unloading points are under surveillance and video analysis so that, if a theft occurs, the video analysis software can be used to search and filter footage by criteria and to isolate areas of and objects of interest and focus the investigation to understand what occurred and track vehicles, containers and individuals.
Video content analytics also provides quantitative data reports for improving port logistics. For example, airport operators often need to understand statistics about pedestrian traffic flows, including how long groups of people dwell in a TSA checkpoint lines, at customs and immigration, or even taxi queues. Similarly, shipping port operators need to know much time it takes to offload or load a ship, move cargo into a warehouse and, generally, how long a ship is docked. With video content analytics technology, operators can visualize aggregated data based on captured video surveillance, and make intelligent decisions based on traffic pattern heat maps and reports. This comprehensive business intelligence is highly relevant and makes video surveillance useful to not only security managers, but also for operation and business managers, empowering them to make intelligent decisions and drive efficient port operations.
By investing in video content analytics, port authorities enable security and operations staff to maximize video surveillance resources, proactively prevent incidents (criminal, medical, or other), respond more quickly to incidents, accelerate post-event investigations, as well as gain quantitative insights about port pedestrian and vehicle traffic efficiency. These critical abilities and the resulting business intelligence are driving optimized safety and efficiency for leading ports worldwide.