FACE RECOGNITION
SYSTEMS AT AIRPORTS
Kanan Habibli Adigozal,
Azerbaijan State Oil and Industry
University, Baku, Azerbaijan
E-mail: habiblikanan@gmail.com
Abstract. The identification of a person in public transportation, workplaces, and international travel industries is always one of the challenges that require accurate, secure, and fast methods. In today’s world, especially in Europe, traveling to foreign countries is easier because of the Schengen agreement with European countries. Also, some countries grant a no-visa policy for their friendly neighbors which represents new challenges for authorities and law enforcement. Considering that humans live in a surveillance state in the modern world, biometric solutions such as facial recognition are seen as one the most effective ways to solve the problem. Face recognition systems offer fast and secure identification services across multiple transportation sectors. Even though the advantages of the software, up until the COVID-19 epidemic, investments in this technology sector were not enough. The success of the software during the epidemic has proved its usefulness and vision for the future and its market size has gone over $4 billion as of 2020. This paper will talk about the usage and challenges of face recognition in general and how it can be used at international airports in the future.
Key words: face recognition, face analysis, biometric security, security, airports, crime prevention, transportation
Introduction
Ever since the mid-50s, the speed of innovations in technology keeps increasing. These innovations have been focused to speed up daily events and processes. Since its coining in 1959, machine learning has been one of the most useful and fast-growing sectors of technology. The term itself consists of multiple parts that can be used in various fields in human society. It has the potential to be the future of humanity. The technology is also known as self-teaching software. It means that the software learns from its mistakes and previous experiences and gets better at each turn, so it becomes faster and more accurate. This is why it is considered to be used in transportation sectors of the state that need multiple steps to fully identify and verify a person.
Getting ready for a trip - doesn’t matter if it is for tourism purposes or not - requires a lot of effort from a documentary point of view. International airports must check and verify each person carefully and thoroughly so it takes a lot of time and effort. So, as the some parts of the world such as Schengen zone present a free movement zone for the travelers that have type C visa or the countries give a no-visa permission to their friend countries, the authorities confront a challenge on how to create a faster, more accurate and more secure identification system to prevent unwanted events such as trespassing, smuggling or the ones like Brady Cooper case.
As the technology is relatively new, there are a couple technical and scientific challenges ahead to perfect the software. The practice makes it perfect and some of the airports in the world such as Helsinki-Vantaa airport in Finland have started to use face recognition systems.
What is Face Recognition?
Face recognition is a software to identify and confirm an individual identity using a camera and their face. The technology relies on detecting and measuring human faces’ features in a given image or a video. Today’s facial recognition technology is able to detect human faces and based on the results, whether the faces in the image or the video belong to the same person or not. It can also look for a certain person’s face in a given array of multiple images collection.
Facial recognition is considered as a part of biometric security. Technologies such as recognition of fingerprint, voice and iris are other categories of biometric security. Facial recognition as its other peers is often used to increase security at public transportation and to fasten the process of searching for law enforcement.
How does face recognition work?
The process of usage of face recognition can be divided into multiple parts. The first step is detecting a face. The camera detects the face of the person and its position who may stand alone or in a crowd. The challenge is the person’s facing direction may change in each case. There are 2 main challenges that must be taken into consideration in these kind of systems:
The first one is facial expressions such as the gender, the effect of the age and facial expression (usually the person’s resting face). The posing of the person matters the most so at the airports officers ask the passengers to look directly at the camera. This realignment helps to achieve a better image to identify a face. This is also called the intrinsic factor of facial recognition.
The second is closely related to the environment in which the image or the video was taken. The viewing angle and the geometry of the face, the illumination and the shadowing of the place and the objects in it, and the image processing of the photo. These are the extrinsic factors of facial recognition
The second is analyzing the face. As the tool captures the face in the image, it starts to analyze the face. The software looks at the face from a 2D perspective because it makes it easier to read and compare the geometry of the captured face in public images and databases. The human face geometry has a couple of key factors: the distance between the eyes and the depth of their sockets, the shape of the cheekbone of the face, the distance between the chin and the forehead, and the distance from the nose to the mouth.
As the face is read by the software, the next step is to convert the image to the data. The process converts analog information to digital information. The concluded information is called a faceprint which is a string of characters consisting of numbers and points. This numerical code is unique for each person and can be used the other way around to get the analog version back.
The final step is to recognize the identity of the owner of the face. Using the faceprint, the system looks for a match among the other already recognized faces. Today’s software companies keep the facial information of their users in their databases for later use.
Facial verification algorithm promises accurate identification and verification of a person when the requirements such as good illumination, status positioning, well-controlled imaging environment, good camera quality and image resolution, are met.
The Categories of Facial Recognition Systems (FRS)
All FRS uses the steps above during daily tasks, but there are two categories of them in which environment and for what reasons they are used; verification (authentication ) method and identification (recognition) methods.
The verification method presents more control to the airport so it is considered more reliable. The process is also much simpler since it basically takes and stores the person’s picture and when it is needed it just fetches the picture from the database and compares the latest taken image with it. Because the image is taken in a better controlled environment (the airport itself), the illumination and the quality of the picture is better and verification can be done much easier. This method can be used as an authentication method for the internal airport systems so workers and officers of the airport can log in the computers.
The identification method is considered more complicated because instead of taking a single person’s photo from close range, this method is used to identify a person in a crowded environment in a lower quality from any distance. For this method to work, the wanted person’s picture must be in a database so it can be compared from the image in a crowd.
Challenges of FRS Usage at Airports
All the international U.S. airports, some of the Indian airports such as Pune and Kolkata can be introduced as the example airports that use facial recognition systems from around 2020. These experiences are the opportunities to review this technology and find its technical and social flaws and cope with them.
Most of the FRS requires a pre-registration procedure which requires the users to pay a certain fee. The users get a smart pass card which allows them to use the system. The card provides going through border control and plane boarding procedures without waiting in a queue. So basically, you have to pay for the usage of the software.
On the other hand, there are privacy concerns of the travelers. Like other biometric security systems, FRS scans human features to work faster and it sometimes causes inaccurate results which may result in misleading actions by the authorities. Statistics show that 1 out of 25 travelers will get the wrong identification with the current accuracy of 96% of the FRS. The current state of the system may also work differently according to the face or the gender of the traveler and by some considered not to be ready for the use of the public.
Data vulnerabilities also can be a cause of concern. Even though authorities declare that after a given amount of time, the images taken will be deleted from the databases to prevent the system from violating human rights, from a technical point of view, if one can access even the hash of the image, it can be restored and used for other reasons.
Conclusion
In conclusion, we can say that there are a lot of flaws in the FRSs. The increase of the usage of the technology will present us with more feedback all around the world in different situations. There are multiple cases that can’t be foreseen today. Such as gender, races, facial features and wear like beard, makeup, glasses etc. can create confusion at the decision making of the system. People can change drastically as the cause of gaining/losing weight naturally or because of a disease. System is also very dependent on environmental causes like illumination, dryness, resolution and quality of the camera. However, the problems of the current state of the system can be accepted as normal, considering the fact that the technology is still new and in development.
References:
1. URL: https://aws.amazon.com/what-is/facial-recognition/
2. URL: https://www.kaspersky.com/resource-center/definitions/what-is-facial-recognition
4. URL: https://www.analyticsinsight.net/what-is-the-importance-of-facial-recognition-in-todays-world/
Телефон: +79503125696 E-mail: ontvor@yandex.ru
Сайт: on-tvor.ru; vk.com/ontvor
