With the rapid development of biometric recognition technology, biometrics is highly adopted in our daily life. We unlock our mobile phone using fingerprint, and verify our bank transaction with face recognition. Biometrics has been applied for physical access control more than ever. Besides access cards, fingerprints, vein recognition, and facial recognition, are also often used to open the doors right now.
Biometrics to improve convenience
Biometrics can give an improved experience compared to an access card to identify a person. One can forget a card at home or lost it, but will never forget your finger or face at home, nor lose them. Currently, fingerprint recognition is most widely used technology to identify a person, due to its low cost, but still reasonable high accuracy. Moreover, using facial recognition or long-distance iris recognition, can provide users the ultimate “hands-free” access. The identification of a person happens when the face or iris is within the sensor range. Doors are open for you on the right moment when you walk to it, with no extra effort of presenting cards or fingers.
Biometrics to strengthen security
Biometrics is considered to be a better way to verify “you are the person who you claim you are”, compared with PIN code verification. As a knowledge, PIN can be forgotten or shared with multiple person, while biometrics can be neither forgotten nor shared. Therefore, in a high security environment, biometrics is used as a secure way for verification, in additional to access cards.
Areas of concern
Although biometrics have huge advantages to use, there are still areas of concern, and even potential risks, when using it.
- Accuracy. Biometric recognition technology provides a probability for identifying a person. It cannot provide 100% certainty of recognition results using fingerprint, vein, iris, or face. Each biometric recognition technology has its own false acceptance rate and false rejection rate. In general, from high accuracy to low accuracy, we have iris, vein, fingerprint, and face. There are also other external factors which influence the accuracy, such as the sensor quality, the senor hygiene, sunlight, dirt, etc.
- Privacy. Biometrics is classified as sensitive personal data in European General Data Protection Regulation (GDPR). It requires special attention while using and storing the biometric information of a person. In some countries, it is not allowed to store any biometric data in a central database.
- Fail to enroll. There will be small group of person cannot be enrolled using certain biometric recognition technology, due to various factors, i.e., dry finger, or finger with vague minutiae, or eye disease, or too dark iris. Therefore, for large scale population, often multiple biometric technology are applied.
- Liveness detection. A system attacker can fool the biometric device with a photo print of a fingerprint, or a picture a person, if the biometric device does not equipped with liveness detection.
How to choose
The following table lists the most common seen biometric recognition technologies used in physical access control, and the comparison of their characteristics.
|Fingerprint 2D Optical sensor||**||*||***||*||***|
|Fingerprint 2D Multispectral||***||**||***||*||***|
|Iris long distance||****||*****||****||***||**|
Number of * indicates from the scale low to high. * is lowest, and ***** is highest.
Liveness detection and fraud detection often are implemented as add-on functionalities along the basic biometric recognition technologies. The common technology used are to detect the warmth, textual, active blood flow of the investigated body parts, or even the gait of a person. Each biometric recognition system might integrate with one or more detection technologies; or sometimes none, to pursue the lowest cost. It is advised to choose your system carefully based on the different scenarios.
To complied with local privacy regulations, sometimes biometric templates are requested to be stored in access cards, instead of in a database somewhere. Each person owns and carries their own biometric templates with them. If one person lost his or her card, or the card is stolen and compromised, only this person is affected. This minimize the security risk of all. However, this also determines that biometrics can only be used as verification method, while identification is still done by access card. Moreover, due to the limited storage in an access card, normally only one or two biometric templates are stored in a card. Compared with centrally stored templates, those templates in cards often have lower resolution, which provides less characteristics to be compared with. Although it might be sufficient for 1:1 matching (verification), it definitely lowers the accuracy of the matching.
In general, 95% of the PACS customer chooses 2D fingerprint. The main reason of choice is cheap price. Among 2D fingerprint sensors, multispectral fingerprint sensor is often a better choice than the optical sensor, because of higher accuracy and more reliable performance. It is also slightly more expensive than optical sensor. In case of the high security environment, iris would provide the best accuracy, followed by palm vein recognition and 3D fingerprint recognition. Facial recognition and multi-finger solutions are still often used for hospitality and eye-catching purposes, due to their high price.
What AEOS Offer:
As an open platform, AEOS has published its Biometric Restful API and Device Integration Protocol so that various biometric systems can easily integration with AEOS.
- AEOS biometric interface is privacy-by-design. By default, no personal information is synchronized between AEOS and biometric systems. More information can be synchronized only when user activates it.
- AEOS biometric interface is security-by-design. All the communication via IP between AEOS and biometric systems are secured.
- AEOS integrate with various biometric systems, offering our customer wide range of choices from different types of biometric technologies and various price range. We are working closely with TBS, Idemia, Suprema, Recogtech, Icognize, and delivering fingerprint 2D, 3D, iris, palm vein solutions.
Want to know more?
Download our whitepaper.