How does security giant 'Dahua' use AI to secure airport security?

Aircraft travel is the safest of all modes of transportation. However, in recent years, due to the intrusion of drones and the frequent occurrence of terrorist acts, people are particularly cautious when choosing to travel by air. In addition to the public, as a channel to the outside world. And the city's transportation hub, which has always been the primary concern of the relevant government departments.

In this context, in order to effectively cope with various emergencies, the concept of 'smart airport' has quietly emerged, and relevant manufacturers have successively launched airport intelligent security systems. As a domestic security giant, Dahua shares have many plans to land in this area.

Airport security management challenge

It is reported that Dahua Co., Ltd. established the Great Transportation Division in 2015, and the airport is a sub-industry line of big traffic. With this, Lei Feng network in an interview to Dahua Technology Company Limited resolve Han Wei Zhong program manager, he chatted with the relevant focus of airport security.

Han Weizhong told Lei Feng. Dahua has a good say in this block. Since the establishment of the business unit, Dahua has participated in the construction of some airport security systems at home and abroad as a product supplier. He said that in fact, these programs are 'concerned'. The objectives generally cover three categories: aircraft (aircraft), people (passengers, staff) and vehicles (social vehicles, related work vehicles).

In the view of this solution manager, there are currently three problems that need to be solved in airport security.

1. First of all, airports of all sizes require complex collaboration of multiple (sub)systems such as monitoring systems, alarm systems, access control systems, network transmission systems, and management platforms. However, this is not the case. Han Weizhong said that the security systems of major airports are basically built in phases, and the various security subsystems are basically separate. The system including the video system, the access control system, the alarm system, the bayonet system, the parking system, and the like form an information island. The administrator needs to log in to multiple systems for operation, and the system cannot perform emergency linkage.

2. Last year, the State promulgated the Anti-Terrorism Law, which stipulated the requirements of the video surveillance system in the airport, that is, the video surveillance HD 1080P, no blind spots, and storage for 90 days. At present, the CCTV system, which was built in China 14 years ago, is basically based on analog systems. These videos cannot meet the requirements of the Anti-Terrorism Law. The video tracking and intelligent analysis after the incident have caused great results. difficult.

3. Various areas of the airport, including the dock area, parking lot, office area, cargo area, front area, flight area, hangar, surrounding areas, etc., each of which is a vertical management system. Due to the particularities of these areas, the complexity of the large number of people and goods entering and exiting, there are too many factors controlling risk.

How Dahua Smart Airport Solutions can solve such problems

For the above three issues, the Dahua Airport solution combines a multi-functional HD surveillance camera with deep learning-based AI to analyze big data to achieve its goals, whether it is an airplane, a car or a human face.

For aircraft such targets, mainly including the panoramic display of the flight area: due to the improved performance of the front-end camera CPU, the current panoramic display technology has evolved from expensive back-end server splicing to cost-effective front-end camera panorama splicing, simple installation and splicing effect Well, if installed at the airport tower you can take a bird's eye view of the airport.

Targets such as people mainly include individual intelligence applications (face recognition) and group intelligence applications (behavior analysis). Face recognition is mainly used in key areas such as security checkpoints, boarding gates, VIP lounges, etc. Dahua has successfully implemented the security inspection system docking public security key personnel pool in the docking with the airport public security. Any suspect dares to appear in the airport area. Alarms will be triggered immediately; behavior analysis is mainly used for terminal entrances and exits (number of people), check-in counters (queue length analysis), quarantine shops (passenger flow analysis), gates (personnel gathering), etc. Prevention of incidents and deployment of security personnel.

For such vehicles, UOB uses an integrated camera to capture and identify vehicles that are illegal, stolen, blacklisted, and crime-related to the airport. In addition, through the overall layout, the overall control of the vehicle can be realized: from the airport overhead vehicle capture, to the airport terminal arrival floor, the departure layer of the illegal parking capture, and then to the airport parking lot to guide the reverse search, Dahua can Completely restore the trajectory of social vehicles entering and leaving the airport and locate specific parking spaces. Combined with the airport flight system and the airport public security vehicle management system, the owner information and flight information can also be located.

Lei Feng Network has learned that Dahua's products have already penetrated into Beijing Capital International Airport, Shenzhen Baoan Airport, Nanjing Lukou Airport, Xi'an Xianyang Airport, Wuhan Tianhe Airport, Guangzhou Baiyun Airport, etc. In addition, at some regional airports, Dahua Airport The overall solution has also been able to land: such as Dunhuang Mogao Grottoes Airport, Shandong Linyi Airport, etc., including land-based road vehicle capture system, front-end camera, storage system, alarm access control, monitoring center decoding large screen, integrated security management platform. It is worth mentioning that they plan to use intelligent video applications such as floor-standing thermal imaging and panoramic stitching in the second phase.

Here, he focused on the application of thermal imaging cameras in airports. He said that the Dahua Airport solution also took care of the surrounding areas of the airport. Since there is substantially no light in this area, special equipment such as thermal imaging cameras must be used to ensure the safety of the area.

Thermal imaging technology has the following advantages: including PTZ function, remote monitoring (can detect vehicles within 8.8 km), long-distance zoom, binoculars (optical and thermal imaging lenses), and intelligent (can detect intrusion). Compared with the current peripheral protection devices on the market, such as vibrating fiber sensors and infrared beam sensors, they have higher accuracy and fewer false positives, while directly providing video for verification.

Party A's needs and attitudes towards AI related programs

It is not easy to successfully land the project. The solution manager sighed that in addition to the technology itself, there are usually many problems encountered during the project landing process. At present, in the bidding process of the airport-related programs, Party A mentioned that the more demand mainly includes three points:

1. The airport unified security management platform can integrate various security systems and unified scheduling management;

2. Application of video intelligence in the terminal building (number of people, face recognition, personnel gathering, queuing detection, etc.); application of video intelligence in the flight zone (flight zone panoramic display, aircraft take-off and landing tracking, etc.);

3, the airport simulation monitoring system transformation needs (in response to anti-terrorism law requirements, simulation HD transformation).

In his view, these needs have been difficult to achieve in the past, unless they are viewed by human eyes and manually operated; but with the advent of AI, they can be solved very well. For example, the comparison of people's cards requires the security personnel to confirm with the naked eye. The AI ​​technology can further increase the accuracy. Before the occurrence of AI, the AOC and the TOC control center will then issue instructions to carry out the emergency plan operation. After the AI ​​technology is mature, it can be used for early warning before the event occurs. For example, when the check-in counter and the security checkpoint are gathered, an alarm message will be generated. The staff will temporarily open the window and configure the security through the pre-judgment of the alarm information. Personnel go to the site to deal with other measures to prevent it from happening.

Speaking of the rise, Han Weizhong also introduced to the front of the company the attitude of Party A on AI products. He said that Party A is still relatively conservative at the beginning of the intelligentization, mainly because the relevant technologies are not mature enough to be put on the ground, including setting up face-to-face access in the airport staff channel, and entering and exiting through the face, but after the application, it is found that the employees need to face each other. Access control identification, the surrounding light has a great impact, often after the failure to identify the card to pass, the effect is greatly reduced.

“But the face recognition technology has reached a new level. The recognition does not require active coordination. It can be identified and compared by the face recognition surveillance camera installed on the channel, and the efficiency is obviously improved,” he added.

There are many solution providers, how can Dahua win?

Subsequently, the solution manager said that there are many providers of airport intelligent solutions, and how to stand out from them is very important.

Nowadays, with more than ten years of technology accumulation in the industry, Dahua has its own core technology in processing video encoding, decoding and algorithms. In addition, Dahua is also actively working with NVIDIA, Genetec, Milestone, Avigilon, AXXON, ISS and so on. The cooperation of the company provides more opportunities for the Dahua Airport solution.

In the future, Dahua will continue to invest in cutting-edge technology research and development in civil aviation video surveillance solutions, ensuring seamless integration of front-end IPCs and back-end NVRs to establish a robust security system for airports.