The ICAO standard (ISO 19794-5) is the set of rules for the use of facial recognition at airports, so that crew and passengers can board in the aircraft without having to show physical documents.
In Brazil, in November 2021, Congonhas airport began testing this technology to authorize the boarding of pilots and flight attendants without the presentation of identification documents.
To do so, the International Civil Aviation Organization has initiated a project for automatic validation of personal identity through the electronic passport.
However, for this technology to be safe, the International Standardization Organization (ISO) has developed an international quality standard so that facial recognition can be used in electronic passports.
Thus, the ICAO standard (ISO 19794-5) was established from four requirements. Get to know about them below!
Find out what the ICAO (ISO 19794-5) requirements are
Blurring level
The first of the requirements of the ICAO standard (ISO 19794-5) is the verification of blurring level, performed from searching the information of the high frequencies, in addition to calculating the variance of pixels after Sobel’s high-pass filtering.
In this case, the developed algorithm reached an error rate of only 6.2%.
Presence of hair in front of the eyes
Then, the analysis of the presence of hair in front of the eyes is obtained by detecting vertical edges in the ocular region. From this, it is possible to identify strands of hair in the eyes region.
This algorithm recorded an error rate of 11.9%.
Presence of hat
Among the requirements of the ICAO standard, the analysis of the presence of a hat is performed from the segmentation of the skin in the forehead region.
The algorithm developed reached 7.2% error rate.
Eyes covered by glasses frame
Finally, the possibility of eyes covered by glasses frame is also evaluated with the detection of horizontal edges and lines in the eye’s region.
To accomplish this, the method developed identifies horizontal line segments caused by the frame in the eye region.
This algorithm recorded an error rate of 11.6%.
Understand how facial recognition works at airports for crew members
In practice, when passing through the access control to the Security Restricted Area, instead of presenting documents, the crew members undergo biometric reading equipment, which performs facial reading and validates the parameters obtained.
This validation is carried out with the Digital CHT database (crew identification document), to confirm that the individual is a civil aviation crew member and ensure the authenticity of the document.
If this identification is confirmed, the crew member will have their access to the AIRPORT restricted area without the need to present documents.
However, if the system does not authorize the entry, it will be necessary to present the CHT and the identification document of the crew member to the airport operator agent, responsible for access control to the Security Restricted Area.
It is worth mentioning that this access control procedure, through facial recognition, does not exclude the crew from undergoing airport security inspection.
The goal of implementing this technology, based on ICAO standard, is to make the airport boarding process more efficient, agile and secure.
Is it possible to implement this facial recognition technology in other projects?
Although it looks like a complex system, it is currently possible to implement this technology in different applications from software development kits (SDKs).
As such, you don't need to develop a system from scratch. Just implement the SDK to your project, with the possibility of including features according to your needs, since there are options with Low Code development.
Get to know BioPass ID and implement a facial recognition system in your app
After understanding how the biometric facial recognition system works, as well as the main requirements of the ICAO (ISO 19794-5) standard of the airport-tested system, the time has come to implement this technology in your project.
To do so, it is important to have a solid partner and a robust and easy to implement system.
In this case, BioPass ID, a modern cloud-based biometrics and artificial intelligence platform, offers the Quality API package, with functions for assessment of biometric data, with REST API, availability from any device with internet connection and easy to code.
Our feature can be applied to any project with simplicity, efficiency, and is compatible with a wide variety of applications. Get to know the main features:
- Identity Management.
- Biometric deduplication.
- Latent editor.
- Faces and fingerprints matching.
- Biometric adjudication.
- Multibiometric cloud database.
- Verification of facial compliance with ICAO standard (ISO 19794-5).
- Automatic face cropping.
- Fingerprint quality check.
It is worth noting that the BioPass ID technology is used by law enforcement authorities and is available for all types of applications. Get professional results without any expertise in biometrics.
See our differentials:
- Quick deployment from cloud-based APIs and ready-to-use model SDK.
- Everything you need to create complete applications with biometric support.
- Quickly integrate high-performance biometric and capture functions into your application.
- Constant updates to the API and Software Development Kit.
- Technical support via email, chat or phone.
- Documentation and source code of sample applications.
Try BioPass ID for free through our cross-platform demo applications, with facial recognition and ICAO-standard facial quality analysis.