With 98 million networked cameras and 29 million HD CCTV units sold in 2017 worldwide, video surveillance is a hot commodity, and for good reason. Airport security managers are increasingly sensitive to risk and continue to deploy an abundance of cameras across their facilities.

But the jumble of video technology hasn’t delivered the overall security and safety that airports need. To truly address airport security needs, video surveillance solutions require two powerful areas of technological development: enhanced imaging and artificial intelligence.

Innovative camera technology covers broader areas with less equipment. Infrared and hyperspectral imaging identifies objects and individuals that would normally go unseen. Remotely controlled pan and tilt capabilities focus cameras on areas of interest. Thermal imaging can collect information that ordinary cameras cannot, because sensors can record information beyond the light humans can see. Plus, the right choice and design of equipment can cover more area per camera.

AI at the edge and in the cloud perform basic recognition and observation tasks that once required personnel to perform. By turning video into structured data, AI enables more thorough integration with other data. Airport security involves collaboration of multiple systems: fire detection, access control, parking monitoring, intrusion detection, and more. To reach the level of security they need, airports require systems to bring together all the data for enhanced analysis and operational readiness. An advanced unified video system can turn video from a self-contained unit into a holistic solution.

Combining Enhanced Imaging and Artificial Intelligence

Artificial intelligence and machine vision help automate the analysis of visual content, speeding routine processing. Systems can alert appropriate staff when critical and final decisions are necessary and take additional actions like locking down areas or restricting access.

“AI technology plays a role in analyzing, deepening, and excavating data. It can transform unstructured video information into structured data that can streamline processes and free up personnel time,” stated Kong Lingxin, Airport Solution Manager, Dahua Technology Co., Ltd.

“At the same time, AI technology can get higher accuracy than traditional methods. False positives and false negatives can decrease by 85 percent or more, which makes automated video analysis viable,” he added.

AI and deep learning combined with advanced and enhanced video create a more secure facility-wide environment so that:

• Artificial intelligence and machine vision automate visual content analysis. Systems can recognize potential threats, such as individuals or vehicles on watch lists.
• Automated systems can identify passengers by combining HD video cameras with other intelligent analytics, like facial recognition and e-passport verification.
• Wider fields of view enable the use of fewer cameras in many settings while not losing any visual information.
• High-definition, fisheye, and panoramic surveillance cover critical areas of the airport with wider angles, eliminating blank areas with enhanced haze and night imaging.

But there are challenges. Advanced video systems require intensive data processing. They must be compatible and interoperate with legacy systems. The Dahua More Intelligent More Secure for Airport solution does this and much more.

Get Smart with Facility-Wide Security

The Dahua end-to-end system is an example of how AI combined with advanced video significantly improves the outcomes of airport-wide surveillance. The system’s structure works in parallel with the physical layout of an airport and its extended surroundings, as Figure 1 shows.

Video and other sensors cover all parts of the airport, whether perimeter, parking lots, luggage areas, aprons and runways, roads, or shops. The system can work with a broad range of advanced Dahua and legacy cameras.

Benefits begin with the ability to offload much of the bulk video processing, which frees up personnel to perform more useful tasks than watching a screen or monitor. Systems can scale with increased activity at facilities, preserving risk management and safety. The combination of technologies can also reduce the total cost of security because so many routine tasks are automated.

The system can trigger alarms and alert appropriate staff when critical and final decisions are necessary. It can also take additional actions like locking down areas or restricting access.

By providing real-time analysis and actions, the overall result is better, faster, and more complete responses at a time when airports struggle to keep ahead of potential threats. Perimeter security increases and the system can reduce both false positives and false negatives for stronger risk management.

Computing Power from Edge to Cloud

The system has a layered approach, with cameras, sensors, and devices for threat detection and interaction with people, then inner layers of processors and software for analysis and control.

At the outer layer, Dahua cameras can come equipped with 180-degree cameras for wider coverage. Panoramic cameras provide a 360-degree view. Infrared imaging allows video in total darkness with heat signatures identifying objects, or even fires.

Edge processing is enabled by Intel^® technology, which provides computing power in a limited space, and allows specially outfitted cameras with a visual processing unit, or VPU, to perform first-level AI image analysis. The cameras can then pass along alerts when, for example, an unauthorized person crosses a boundary.

“Our system includes Intel AI acceleration technology, built into the camera. This means that deep-learning algorithms can be run efficiently,” said Kong Lingxin.

“Through the subsequent algorithm upgrade, Software Defined Video can be achieved. With AI acceleration on the camera side, edge calculation of video data can be realized and then integrated with the central cloud platform. A complete cloud-edge federated system architecture is implemented,” he added.

Network Video Recorders, also using Intel processors, are closer to the center of the system. They collect video data for later use, although it can also be streamed when necessary.

In addition to the point of presence that the cameras and NVRs provide, parking space and traffic detection sensors, alarm buttons, and other intelligent devices can send necessary data.

Cloud processors run the Dahua General Surveillance Management platform and Intelligent Video Server. They provide full AI deep-learning analysis and control. The cloud portion integrates with other systems at the airport to bring together all relevant data for better threat assessment and response.

As the challenges for airport security increase, the need for enhanced video and AI processing will become more acute. Systems like the More Intelligent More Secure for Airport from Dahua Technology help organizations strengthen their security capabilities to handle challenges more quickly and efficiently and preserve the atmosphere travelers, patrons, staff, and outlying communities want and deserve.