Large events and festivals draw crowds that can cause challenges for public transport systems and traffic around a venue. Crowd flow analytics and location data can be used to coordinate these events efficiently and safely ensuring that safe crowd density levels are maintained. During half-time at sporting events or intermissions at concerts, thousands of people leave their seats to buy refreshments and use restrooms. Overcrowding is almost guaranteed, so operations staff must determine the approximate crowd density and raise awareness of locations where crushing may occur.

Ipsotek VISuite uses AI to accurately count people within the camera field-of-vision (FOV), whether walking, using stairs, or using escalators. This capability is ideal for venues and stadiums, which typically have multiple entrances and exits, as the results can be displayed in real-time in the analyzed camera video feed. The aggregate values and other operational statistics can be displayed in a dynamic dashboard in the control room. This method of counting people can also be used with cameras that are next to venues on bridges, around pop-up retail areas, and food and beverage locations to assist authorities with managing crowds pre and post-events. If issues occur, Ipsotek VISuite provides real-time crowd management alerts such as crowd density estimation, people counting, GIS coordinates, and live heatmaps and reports. In addition to providing real-time alerts, Ipsotek VISuite can create heatmaps that track where people go and how they use spaces around them. This information can be incorporated into health safety plans, determine F&B placement, and identify changes to venue layouts which can drive enhanced revenue streams.

Queue monitoring and capacity estimation can be applied in many operational areas, including building entrances and exits, pinch points, staff areas, turnstiles, and VIP areas.

Ipsotek VISuite video analytics can provide metrics including:

• Queue Capacity Estimation
• Egress Rate
• Average Queuing Period
• Average Time Spent at Desk or Counter
• Unoccupied Desk Recognition

The Forensics upgrade to the Eviden Computer Vision Platform bundle adds the following additional capabilities:

Also, the Investigation upgrade adds additional support for the following:

Applications can include:

Queue capacity solutions can use several types of cameras and sensors for different applications. For example, Perimeter Protection is often implemented with Thermal/IR cameras where restricted area Intrusion detection and virtual perimeter management require GIS PTZ cameras for effective tracking.

Ipsotek VISuite uses Elastic Search as its database to capture incident-related metadata for display in multiple formats as part of a dynamic dashboard using Kibana. This gives operators a real-time, holistic view of operations across the venue or stadium. Examples include showing queue times at turn styles, crowding in food and beverage areas (F&B) and merchandising zones, and more to gain operational efficiencies, new business revenues, or be sent to other big data platforms for analysis.

In large venues, Ipsotek VISuite can detect vehicle intrusion in pedestrian areas, route passengers to open parking locations, if supported by the VMS or Command and Control Center (C3), or be used at taxi queues to monitor the queue length and wait times.

Computer vision can help with crowd management by analyzing video footage and providing real-time insights into crowd behavior, density, and movement patterns. Use cases for overcrowding and crowd management include monitoring walkways, venue approaches, stairways, bridges, lobbies, hallways, ticketing areas, and many more. This information can be used by event organizers, security personnel, and law enforcement officials to make informed decisions about how to manage the crowd and ensure safety.

The Ipsotek VISuite solution can provide information about the following types of real-time insights:

• People Counting Reports
• Person or Group Loitering Out of Hours
• Social Distancing Monitoring
• Slip and Fall Detection
• Abandoned Object
• Vehicle in Restricted Area
• Overcrowding detection
• Loitering near a critical area
• Panic Detection

The Forensics upgrade adds the following additional capabilities:

• Incident Response

• Forensic Search

• Lost Child Search

Also, the Investigation upgrade adds additional support for the following:

• Multi-Camera Tag and Track

Perimeter protection with computer vision typically involves using cameras and machine learning algorithms to monitor the boundaries of a defined area. The computer vision algorithms analyze the video feed from the cameras in real time, looking for specific types of movement or behavior that could indicate a security threat. Perimeter intrusion detection can be applied to many venue areas, including the outside perimeter, storage areas, roof areas, and access points, and creating virtual perimeters. Perimeter protection solutions are most commonly implemented with Thermal/IR equipment, while restricted area intrusion detection and virtual perimeters typically require GIS-capable PTZ tracking cameras.

Perimeter protection with computer vision can provide a highly effective way to monitor the boundaries of a defined area and detect potential security threats in real time. By automating the detection process with advanced algorithms, these systems can provide a much higher level of situational awareness than traditional security measures.