Facial recognition

Enterprises using facial recognition include Amazon and Facebook. (Getty Images)

Innovation | Trends

Breaking down biometrics, from palm prints to facial recognition to vein scans

Identification falls into two categories—physical and behavioural. The physical includes measurable aspects of the human body such as the face, fingerprint, iris and veins. Behavioural includes characteristics such as gait, keystroke, and voice.

A Facebook IconFacebook A Twitter IconTwitter A Linkedin IconLinkedin An Email IconEmail

Breaking it down, biometric identification falls into two categories—physical and behavioural. The physical includes measurable aspects of the human body such as the face, fingerprint, iris and veins. Behavioural includes characteristics such as gait, keystroke, and voice. The elements are translated into code and stored to be accessed during the authentication process.

Here are some popular, and perhaps even familiar, solutions:

Person using fingerprint access door Companies currently using finger or palm prints include Apple Inc. and Samsung.

Finger/palm prints

Converting patterns including ridge lines and the valleys between the ridges (known as minutiae) into code, this data is stored and accessed for authentication when an employee next scans their finger or palm prints. The more minutiae patterns tracked, which vary across systems, the more accurate the authentication will be. Companies currently using or experimenting with this modality include: Apple, Inc. and Samsung, which allow users to unlock their phones with a fingerprint, PayPal Holdings, Inc., which partnered with Samsung so users to could pay for purchases with a fingerprint when using a Samsung phone, and Visa, which tested out fingerprint readers to replace passwords at ATMs owned by South Africa’s Absa bank. 

Facial recognition

Stored facial images are measured against a digital image, matching components such as nose width, distance between eyes, depth of eye socket, lip shape, cheek bone shape and jaw line to provider user access to computers, networks, or locations. Enterprises using facial recognition include: Amazon for users to make purchases, Facebook to identify those tagged in photos, and software company, ResolutionView, to track hours reported versus actual hours worked by employees.

A graphic of a head talking into a phone for voice recognitionCanadian financial institutions using voice authentication include RBC, Manulife Financial and TD Canada Trust.

Voice authentication

Analyzing physical traits such as the shape of the vocal tract and behavioural characteristics including pitch, cadence and tone, voice biometric solutions digitize words into segments, encoding the frequencies and formants, to create a unique voice print of the individual. This print is used to identify, verify and authenticate the speaker. Canadian financial institutions using voice authentication to verify customers on the phone include RBC, Manulife Financial, TD Canada Trust, amongst others.

Iris scans

Considered one of the most secure biometric measurements (10 times more reliable than fingerprint scans) due the unique characteristics of an iris that typically won’t change over time, these scans involve capturing a high-contrast image locating the centre of the pupil, the edge of the pupil, the edge of the iris, the eyelashes and eyelids. The patterns in the iris are translated into code and stored to be matched against individuals. Several banks including Wells Fargo and CitiBank, as well as airports such as Gatwick and Heathrow in London are using this technology.

Finger being scanned in a small device, with medicial read-out on monitorTop companies using vein pattern recognition include Fujitsu, Hitachi and NEC.

Vein scans

Vein pattern recognition, captured using the palm or finger, occurs when deoxidized haemoglobin absorbs infrared light projected by a scanner, illuminating a pattern. Reference points in the pattern are stored, typically as an image that may be encoded, and later used to identify the individual. Finger vein patterns are much smaller than palms. Retina vein pattern recognition is more complex and not as popular. A scanner projects infrared light through the eyeball. The blood vessels absorb the light and produce a pattern, which is stored as an image. This technology offers a very high level of security since vein patterns don’t change. Top companies using this biometric modality include: Fujitsu, Hitachi and NEC.

Person using computer using fingerprint accessCharacteristics including keystroke analysis are commonly used as an additional layer of security alongside fingerprint scans.

Behavioural biometrics

Characteristics including gait, keystroke and signature analysis, are commonly used as an additional layer of security alongside other biometric identifiers. For example, an employee may log into their computer using a fingerprint and then the system further authenticates their identity by analyzing their keystrokes on the keyboard.