Guide to Physical Biometrics
Biometrics refers to a measurable anatomical or physiological characteristic, such as the iris or fingerprints, used for automated recognition. The term “biometrics” is also used to refer to the automated process by which these characteristics are measured and used to establish or verify the identification of a person. Biometric systems contain five main components: a sensor to capture and digitize data, signal processing algorithms that form the biometric template, a data storage unit, a matching algorithm that compares new templates to those previously recorded and stored, and a decision process that uses the results of the matching algorithm to accept or reject a new individual. Biometric systems can identify or verify the identity of an individual. A system used for identification recognizes the individual’s biometric template from the templates of users stored in its database. A verification system matches an individual’s biometric template to a previously stored example of that individual’s template. Biometric systems are used in law enforcement, in forensic science, in government, and in civilian applications. Companies use biometric systems to restrict access to sensitive areas or information, governments use biometrics to verify the identity of people entering the country, and law enforcement uses biometrics to identify the suspect of a crime. Biometric characteristics are unchangeable or highly resistant to alteration and cannot be forgotten, misplaced, or transferred to another person, which gives biometric systems an advantage over other verification or identification systems, such as PIN numbers or photo identification cards.
A camera captures the image of an individual’s face and constructs a biometric template and uses statistical and mathematical pattern recognition technology to extract patterns from the data and match them to the patterns of pre-existing templates. A still or video camera captures an image, the “probe image”, and detects a face. The system standardizes the probe image’s format to match that of the pre-existing images in the gallery and processes the biometric template extracted from the probe image. The template is compared to pre-existing templates associated with the individual to verify the person’s identity or to all pre-existing templates to find a match and identify the subject.
Facial recognition technology does not require physical contact and images can be captured without the subject’s knowledge by using cameras used for traditional visual surveillance, which gives this technology an advantage over other biometric systems for surveillance use. Limitations of facial recognition systems currently include difficulty in processing and matching images captured low lighting and difficulty processing images captured from the side or from above the subject. Facial recognition systems rely on the quality and quantity of previously-captured images, which can be another limitation to the technology’s widespread application. Facial recognition technology is shifting from two-dimensional images to high-resolution two- and three-dimensional images, which will improve system accuracy.
Fingerprints have been used for identification purposes for over a century because they are unique to each individual and do not change over time. The ease of acquiring fingerprints, the ability to obtain data from multiple sources (each of ten fingers), and the large databases of previously collected fingerprints maintained by law enforcement agencies make fingerprint recognition technology a popular biometric identification and verification system. Fingerprint recognition is automated through the use of digital fingerprint scanners and digital galleries of fingerprint images.
Fingerprints are collected digitally with optical, ultrasound, thermal, or capacitive sensors. Optical sensors, which take a picture of the fingerprint, are the most popular. Sensors may be equipped to take fully-rolled images of all ten fingers, such as the type used for government identity checks. Sensors may also take the image of one fingerprint where it touches the sensor, which are more commonly used to control access to certain areas. Fingerprint recognition requires physical contact, which makes this technology unsuitable for use in the surveillance of subjects without their knowledge. The large databases of fingerprints, the ease of data capture, and the high speed of data processing make fingerprint recognition technology suitable for identifying individuals at checkpoints.
Fingerprints are matched to pre-existing images with minutiae-based matching techniques or pattern matching. The most widely-used technique, minutiae-based matching, compares the location and direction of minutiae points. Minutiae points refer to the ridge endings and bifurcations of ridge paths of the finger’s friction ridge skin, the part of the finger that creates the black lines associated with fingerprints. Each fingerprint contains 30 to 40 minutiae points and no two people have more than eight points in common. Pattern matching compares the new image to pre-existing images to determine their similarity and is most often used to detect duplicate images.
- Fingerprint Identification: An overview of fingerprint matching techniques, fingerprint classification, and fingerprint image enhancement.
- Fingerprints: An introduction to fingerprint recognition and its use in law enforcement.
- Fingerprinting and Databases: A brief overview of the history of fingerprint identification and law enforcement fingerprint recognition and identification databases.
The random pattern of an individual’s iris is set by age one and does not change. A high-quality digital camera that illuminates the iris using infrared light captures the image of an eye. The system uses landmark features to identify the iris and maps the segments of the iris into vectors, which include information on the spatial frequency and orientation of the areas. The system then converts this information into a special code and compares the code to codes contained in the gallery.
The distance between camera and subject can vary between a few inches to a yard. Iris recognition systems capture and process data capture quickly: the camera requires a few seconds to capture the image and the database or gallery can be searched quickly because of the template’s small size. The specialized cameras used in iris recognition technology could be installed and used in surveillance situations, but the application of this technology is limited by the small size of iris databases and the short distance required between the camera and the subject. Subjects can also easily obscure their irises.
Nuclear DNA, or deoxyribonucleic acid, is located in the nucleus of cells and mitochondrial DNA is located the mitochondria. The combination and sequence of four chemical bases, adenine, guanine, cytosine, and thymine, form the DNA code. Short sequences of these base chemicals called “short tandem repeats” or STRs, repeat over the length of the DNA molecule. The number of times these STRs repeat is unique to each individual and is how DNA is used to identify people. Half of an individual's DNA is inherited from each parent, making DNA unique to every person.
DNA may be used for identification because it is found in every cell in the human body and only a small sample is necessary for analysis. DNA samples used for identification are commonly obtained through blood samples or cheek swabs. A mixture of chloroform and phenol isolates a strand of DNA from other nuclear material. A polymerase enzyme is added to the extracted strand of DNA, which replicates the strand to produce enough DNA for analysis. This method is called polymerase chain reaction. A restriction enzyme separates the DNA into shorter pieces that are sorted by size using electrophoresis. The DNA fragments are placed into a tube filled with gel. A negative charge is applied to the top of the tube and a positive charge to the bottom. As larger and smaller DNA fragments move toward the positive charge at different speeds, “bands” form on the gel, which are then compared to other samples.
DNA matching or profiling technology is not automated, which excludes DNA from the category of biometrics. DNA matching requires physical samples to be taken from individuals, which limits the use of this technology in live surveillance. Samples inadvertently left by subjects may be used to identify them after they have left an area.
Disclaimer: The website
www.homesecurity.us provides information in
articles such as a
home security checklist,
cell phone security tips and general
home security information as well as this
current article. The information is provided
as a service to our readers. The articles are
presented for information purposes only with no