Palm vein biometric identification is something someone has probably heard about but not often experienced. It is a technology developed in last ten years, which is widely used there. It provides high security (up to one million reference points) together with the highest accuracy that a biometric technology can offer. The sensor uses a 850 nm and 940 nm long near-infrared waves, detecting low levels of oxygen vein blood and distinguishing it from living tissue.
It is then transformed into a pattern which describes veins inside a person’s palm. The pattern is converted into a template — this is a mathematical model. The template gets encrypted and stored into a database and is matched with what is already stored for proper identification.
Palm vein patterns are unique to each person and research has shown that these patterns remain unchanged as an individual ages. Palm vein patterns are more complex when compared to fingerprints or the finger vein patterns and have more unique features on which to base matches. This allows for the construction of extremely precise digital models, which can be used for the identification of users whose patterns are collected in large databases. The inner side of the hand is used for identification since it is less prone to changes in skin color than the outside, since it receives less exposure to the sun.
The main advantage of this technology is that it provides a non-intrusive mode of identification. High technology reliability has been achieved with low FAR levels (~0.00008 %) and research has confirmed results using databases containing up to 150 thousand palms. Other identification technologies, for example fingerprints, encounter issues with reading dirty or wet fingers, due to seasonal changes in the papillary diagram, cuts, etc. On average, 5 8 % of the people have fingers that are difficult to read using fingerprint identification technologies. The success of palm vein recognition, however, is not dependent on whether the hands being scanned are dirty, wet, or have surface variations. In addition, this technique is both non-invasive and hygienic.
Principle of the Method
The method is based on the principle of reading emissions of the reflection of person’s hand in the multispectral near-infrared lights. Since recovered hemoglobin in the blood absorbs infrared light, the venous vessels have emissions of lower intensity. As a result, a unique diagram is formed and can be elaborated with a mathematical method with the output written into a template. Next, the template is compared to a sample in the template database (in the case of 1: N identification) or is compared in pairs (in the 1:1 verification mode).
Steps of Creating a Biometric Template
- Filtration of the raw graphic image — this process selects important areas of the palm and reduces distracting noise and highlights
- Binarization — brings all of the images to a common view and reduces the impact of different focus lengths and image contrast
- Selection of the scan area — finds «membranes» between fingers
- Splitting of the obtained image into discrete areas with the coordinates of the control points and curve angles and further saving these points into the file, which represents a mathematical
model of the points recorded. This also makes it impossible to reverse generate the original palm vein image from the template.
Identification
- The process is based on comparing a scanned template with a template previously saved to the database
- This comparison relies on a correlation mechanism that carries the major processor load for the calculation system
- Identification speed can be increased by employing a pre-selection mechanism that uses general details of the manner in which the venous image is built (hash code)
Advantages of Palm Vein Identification Technology
- The technology is extremely accurate and secure
- Forging is next to impossible
- It is a non-intrusive method
- Identity can be verified whether palms are wet or dry
- Well-suited for dry climates and people of any age