• Font Size:
  • A
  • A
  • A

Feature Articles

Machine Vision Puts Bold Face On Growing Biometrics Industry

by Winn Hardin, Contributing Editor - AIA

Terrorist attacks in the U.S. since 1999 have changed the world in many ways. Until then, biometrics was a relatively obscure technology used by high-security military systems. Literally overnight, biometric market forecasts tripled from $700 million by 2006 to $2.05 billion, and for once, the market analysts were right. Based on the latest figures from the International Biometrics Group (IBG), a business group that tracks the biometrics market, biometrics is expected to generate $9.3 billion in revenues by 2014. This marks a 115% increase from 2010 revenues, which reached $4.3 billion, according to IBG.

The ensuing years and brought about success in the biometric market as well as a changing set of opportunities for machine vision companies. In a previous article titled, “Biometrics Matures; Vision Gets Nod”, www.MachineVisionOnline.org spelled out the changing needs of biometric cameras – the largest point of entry for machine vision equipment suppliers targeting the biometric market. In this article, we’ll look at developments and demand for various biometric solutions and how they have changed in recent years as this biometrics matures into a cornerstone of the security market.

Mutual Needs, Common Solutions

Almost all biometric systems use machine vision technologies, from the camera, to the image processing hardware and software. In the early days, biometrics was viewed as a large potential market for the machine vision industry because of the synergies between machine vision solutions and biometric application needs. But as markets matured, and volumes increased, more companies realized the potential benefits of becoming a ‘pure play’ biometric company. Does this mean that biometrics doesn’t offer the machine vision industry growth opportunities? Not hardly. Some software companies, and many hardware companies – especially camera manufacturers – continue to sell into the biometric security market with great success. Let’s look at how the applications have changed over the past few years.


Fingerprints are the oldest and most widely used biometric around the world. The FBI hosts the largest fingerprint database through its Integrated Automated Fingerprint Identification System (IAFIS). The FBI settled early on a wavelet form of lossless image compression to avoid artifacts, improve reliability and speed storage and search routines, but the system has come under fire in recent years based on cases similar to the Brandon Mayfield case. Mayfield is a lawyer and Muslim living in Oregon. After the Madrid bombings in Spain, the FBI erroneously identified Mayfield as a suspect based on a partial fingerprint match. Mayfield was eventually cleared, but reports indicate that Mayfield’s fingerprint match, which the FBI indicated was a 100% match, was actually one of 20 fingerprints that were similar to a partial fingerprint lifted from a bag of detonators found at the Madrid bombing site. While capacitive sensors were once seen as the leading method for acquiring a image of a person’s fingerprint because of low cost, they can be fouled by dirty skin and/or worn or mutilated skin. Capacitive sensors measure capacitance values that vary based on whether the sensor is in contact with a ridge or valley in the fingerprint and are relatively impervious to contamination. Today, optical sensors operating at 30 fps are the most commonly used in fingerprint scanners so that the system can select from multiple images as the individual roles the finger “nail to nail” across the sensor, resulting in a complete fingerprint. According to IBG, fingerprint readers make up 66.7% of the biometric market, or about $2.27 billion in 2009.


Hand and vein geometry can be used together, separately, or in conjunction with fingerprint scanning. Hand and vein geometry systems look at the unique sizes and position of fingers, fingers to knuckles, vein patterns, etc., and create biometric values or keys that describe the unique relationship of the features both individually and in relation to one another. There has been some interest in hand recognition systems that also offer high enough resolution to create images that can produces 500 dpi images of each fingerprint. To do acquire a hand and fingerprint image with sufficient resolution means the camera must be in excess of 10 megapixels.

Iris Recognition
The pattern of blood vessels on the back of the human eye offers one of the best ways to identify a person outside of analyzing DNA, however, iris-based biometrics are also among the most invasive. A person must put their head in a specially designed reader, and the camera should use infrared lights and sensors to generate the best possible image because of the thin layer of water on the eye that can be problematic for visible light sensors. Because the person’s eye must be volunteered to the system, iris recognition is typically used only in applications that require the highest security, such as military installations.


Facial Recognition
While there have been few major developments in the above-mentioned biometric systems, facial recognition has been actively pursued by many security interests. Facial recognition uses the size and relationship between major features of the face, such as eyes, nose, and mouth. Facial recognition rose to the top of the biometric security systems list because it holds the potential to work in uncontrolled environments, such as airports. In theory, a facial recognition system should be able to pick a known terrorist out of a crowd, however, making that theory a reality has proven difficult – if not impossible – using today’s commercial off the shelf (COTS) technology. However, facial recognition has proven its worth – most recently in New York State where the addition of facial recognition enrollment and verification technology has helped the state 150 imposters and wanted terrorists, including a Mafia hit man and bank robber.

One of the most active areas of development for security imaging systems is video analytics, or the ability to screen images for security dangers by making automated measurements, often by combining imaging with non-imaging sensors. As the world adds more and more cameras, security personnel are pushed further beyond their limits to review all the data from the world’s growing network of cameras. The answer must be cameras with greater intelligence. As cameras continue to get ‘smarter’ and utilize different parts of the electromagnetic spectrum – such as infrared – that are useful for biometrics and other security applications, the segment of biometric revenues that go to machine vision suppliers will continue to grow.

*The editor would like to recognize special assistance from Ben Dawson, Director of Strategic Development at DALSA Corporation.



There are currently no comments for this article.

Leave a Comment:

All fields are required, but only your name and comment will be visible (email addresses are kept confidential). Comments are moderated and will not appear immediately. Please no link dropping, no keywords or domains as names; do not spam, and please do not advertise.

First Name: *
Last Name: *
Your Email: *
Your Comment:
Please check the box below and respond as instructed.

Search AIA:

Browse by Products:

Browse by Company Type: