Ravi Das

The biometric template creation process of hand geometry recognition
To start the enrolment process with a hand geometry recognition system, a combination of both prisms and Light Emitting Diodes (also known as LEDs) are used from within the scanner to capture the raw images of the hand.

All sides of the hand are taken into consideration, including the front, the back, and the palm. After these raw images are taken, a composite 3D picture of the hand is then created.

To capture the best raw images possible, five guiding pegs are located just beneath the camera to help guide the individual into properly positioning his or her hand.

A serious flaw with this process is that the images of these pegs are also captured. Thus, extra processing time is required to crop out them out from the raw images.

Because of this, the extraction algorithm cannot take into account any variances which can be caused by the rotation or different placements of the hand. To successfully create the enrolment template, the average measurements taken from the seven categories (as described in the previous section) are calculated.

These are then converted over into a binary mathematical file in a manner very similar to that of fingerprint recognition. In fact, the file itself is very small, only 9 Bytes.

Another flaw in creating both the enrolment and the verification templates, is that the geometric features of the hand share quite a bit of physiological resemblance with another. As a result, this can greatly hinder the unique feature extraction process.

To help alleviate this problem, a method known as Principal Component Analysis (also known as PCA) is utilised. This allows for a set of uncorrelated features to be “cloned” from the raw images so that the unique features can be extracted.

Hand geometry recognition also becomes the modality of choice for creating, deploying, and implementing a multi-layered security solution (it is best used for a Physical Access Entry application). It actually works very well even in a standalone mode, for both Physical Access Entry and Time/Attendance applications.

The strengths and weaknesses of hand geometry recognition
Just like retinal recognition, the advantages and disadvantages of hand geometry recognition can be classified according to the same seven criteria:

This is one of the biggest strengths of this particular modality. Most individuals have at least one hand that can be scanned for unique features, and the technology even has advanced to the point that even any physical deformities (up to a certain degree) can be taken into consideration. There are even hand geometry scanners which are designed exclusively for left-handed individuals. The technology is very easy to use, and to train individuals on. All that is needed is the proper placement of the hand in the five pegs.

Although most individuals possess some unique features to their hands, it does not possess rich data, like the retina.

This is probably one of the biggest disadvantages of hand geometry recognition. Because the hand can be considered to be an external component of the human anatomy, it is prone to the harshness of the external environment. The shape of the hand can also be greatly influenced to a certain degree by weight loss and weight gain, injuries, and even other types of ailments (such as rheumatism).

The raw images which are captured of the hand are not affected by the surface elements of the skin. This includes grime, dirt, and scars.

Despite the lack of the richness of data, the accuracy rate of hand geometry recognition is ranked very high. It has an Equal Error Rate (also known as the ERR, this is where the False Acceptance Rate and the False Rejection Rate equal each other). But, the device itself is very bulky, and this can prove to be a serious disadvantage regarding end-user perception and acceptance.

For the most part, hand geometry recognition is very well accepted by the general population worldwide. This can be attributed to the fact that it is perceived to be non-invasive. The only negative aspect is the concern of the hygiene of the actual platen on which the hand is placed, as this requires direct contact by the end user.

Resistance to circumvention
Hand geometry recognition is very difficult to spoof because this requires creating an entire 3D physical mock-up of the hand.

A case study: Yeager Airport

The problem
Ever since the tragic events of 9/11, security at the major international airports and airlines has obviously become of great importance. For this very purpose, biometrics has become one of the tools of choice to enhance the levels of security.

Although it is very important to screen passengers traveling through the airports, it is also equally important to screen the airport and airline employees as well. After all, they probably pose a greater threat to security because they know their way around through all the access points. Because of this, even the smaller to medium-sized airports across the United States have also started to implement biometrics.

The setting
A good example of this is Yeager Airport, located in Charlestown, West Virginia. It has a land space of about 740 acres, and six commercial airlines serve the airport and the surrounding communities. A primary reason for the increased need of security is that the control tower is located in the heart of the main passenger terminal.

Other security concerns included:

  • The main entrance doors to the airport open an average of five times per hour, around the clock;
  • The cost of maintaining a physical security presence was staggering – as much as $1,200 per day;
  • The airport’s HVAC system, as well as other types of sensitive equipment, are also located in a basement near the stairwell which leads to the control tower.

The solution
The management team at Yeager Airport ultimately chose the “HandKey II” devices, manufactured by Schlage Recognition System. Five scanners were initially installed, with many more being planned for the future.

The benefit
Once an airport employee has been properly verified and authenticated by the hand geometry scanner, only then will the airport doors open, and entrance to the control tower be granted. Also, by implementing the use of biometric technology, the costs of having to hire a physical security presence is no longer needed, thus resulting in huge cost savings for the airport.

Apart from enhancing airport security, it is heavily expected that the use of hand geometry recognition will grow quite rapidly in the federal government and law enforcement sectors. For example, although it is not currently being used in the e-passport infrastructure at present, serious consideration is being given to it.

Also, like the Automated Fingerprint Identification System (AFIS) is being continually updated by the FBI, it is quite likely that hand geometry recognition will also be used as a complement to fingerprint recognition.

Finally, although hand geometry recognition is considered to be the “oldest biometric technology,” it is by no means going to be replaced in the long term. For example, predictions are that the growth rate of this modality will continue at almost 23% in the coming years, up to 2019.

Our next article will deal with a very controversial biometric – facial recognition.

Keesing Technologies

Keesing Platform forms part of Keesing Technologies
The global market leader in banknote and ID document verification

+ posts

Ravi Das is a Cybersecurity Consultant and Business Development Specialist. He also does Cybersecurity Consulting through his private practice, RaviDas Tech, Inc. He also possesses the Certified in Cybersecurity (CC) cert from the ISC2.

Previous articleIndia eases regulations for reissue of OCI cards
Next articleCurrent and valid passports on eBay