Biometric Technology	Back to Homepage
Collectively, biometric technologies are defined as, "automated methods of verifying or recognizing the identity of a living person based on a physiological or behavioral characteristic" (Industry Information: Biometrics, 1996). In analyzing the definition of biometrics, several distinct terms must be elaborated upon to completely understand the framework of biometric technology. The phrase "automated methods" refers to three basic methods connected with biometric devices: (1) a mechanism to scan and capture a digital or analog image of a living personal characteristic; (2) compression, processing and comparison of the image to a database of stored images; and (3) interface with applications systems (Industry Information: Biometrics, 1996). These methods can be configured in a number of different topographies depending upon the biometric device and application. For example, a common issue is whether the stored images (reference templates) reside on a card, in the device or at a host or database. (Industry Information: Biometrics, 1996). The term "living person" may seem obvious, but it is an important component in defining biometrics. One of the first questions newcomers to the field ask is, "What about a latex finger, digital audio tape, plaster hand, or prosthetic eye." The answer is that biometric devices can incorporate specific algorithms that determine whether there is a live characteristic being presented. The term "living" also separates the biometric industry from the forensic identification field, although basic principles transcend both areas (Industry Information: Biometrics, 1997). Physiological vs. Behavioral When referring to a biometric technology, it is important to distinguish between physiological and behavioral human characteristics. A physiological characteristic is a relatively stable human physical characteristic, such as a fingerprint, hand silhouette, iris pattern, or blood vessel pattern on the back of the eye. This type of measurement is unchanging and unalterable without significant duress (Industry Information: Biometrics, 1997). Alternately, a behavioral characteristic is a reflection of an individual's psychological makeup, although physical traits, such as size and gender, have a major influence (Industry Information: Biometrics, 1997). Examples of behavioral traits used to identify individuals include a person's typing patterns at a keyboard, commonly referred to as keystroke dynamics, and the unique characteristics of how one speaks or speech identification and/or verification. Identification vs. Verification Sometimes verification and identification are interpreted as similar terms but they have two distinct meanings. Identification occurs when an individual's characteristic is being selected from a group of stored images. Identification is the way the human brain performs most day-to-day identifications (Industry Information: Biometrics, 1997). For example, if a person encounters a familiar individual, the brain processes the information by comparing what the person is seeing to what is stored in memory. Biometric devices that implement identification techniques can be quite time consuming. Often anywhere from five to 15 seconds or more are required in identifying the appropriate individual. In many cases, verification is used to authenticate a user's identity. A biometric device that uses verification requires that the individual make a claim of identity by presenting a code or a card. The matching formula or algorithm then needs only to compare the live and enrolled images of the user's characteristic. The question put to the machine is, "Are you who you say you are?" instead of, "Do I know who you are?" (Industry Information: Biometrics, 1997). Verification can be viewed as adding another level of security. A good analogy is when a person goes to add a dead-bolt to a door. In this case, a dead-bolt is usually added to increase the security of the door or entrance because generally a lock of some sort that was on the door beforehand. There are other notable details to consider in addition to the terms used to define biometrics. Other biometric performance factors that need to be thoroughly investigated include accuracy, speed, reliability, acceptability, resistance to counterfeiting, enrollment time, database storage requirements, intrusiveness, and cost. Biometric Testing The identifying power of a particular biometric encompasses two terms: False Rejection Rate (FRR), or a Type I Error, and False Acceptance Rate (FAR), or a Type II Error. False Rejection Rate and False Acceptance Rate are complementary in determining how stringent a biometric device is in allowing access to individuals. As a result, biometric devices commonly include features to allow for variable threshold or sensitivity settings. For example, if the false acceptance rate threshold is increased to make it more difficult for impostors to gain access, it also will become harder for authorized people to gain access. As FAR goes down, FRR rises. On the other hand, if the false acceptance threshold is lowered as to make it very easy for authorized users to gain access, then it will be more likely that an impostor will slip through (Industry Information: Biometrics, 1997). Hence, as FRR goes down, FAR rises. In understanding the impact of FRR and FAR rates consider an automated teller machine (ATM) access system: a "False Acceptance" means you may lose a few dollars, whereas a "False Rejection" means you may lose a valuable customer (FAR POINT Consulting Inc., 1997). Another good example in understanding the inverse relationship of FRR and FAR rates, involves a car alarm. When your car alarm is very sensitive, the probability of the bad guys stealing it is low. Yet the chance of your accidentally setting off the alarm is high. Reduce the sensitivity, and the number of false alarms goes down, but the chance of someone stealing your car increases (Recognition Systems, Inc. 1999). While the terms "false reject" and "false accept" are still commonly used in quantifying a biometrics' ability to rightfully identify an individual, the federal government has recently adopted a new standard of error rate measurement. Dr. Jim Wayman, Director of the United States National Biometric Test Center, has promoted the terms "false match" and "false non-match" as the new de facto terminology in determining a biometrics identifying power. According to Dr. Wayman, the problem with the terms "false accept" and "false reject" and even more so with "Type I" and "Type II" errors is that their meaning depends upon the claim of the user. For example, depending upon the biometric application, users make either a positive or a negative claim to identity. In a positive identification system, a rejection occurs if a person in not matched to a claimed record. In a negative identification system, a rejection occurs if a person is matched to a non-claimed record. Consequently, the words "false reject/accept" have opposite meanings, depending upon whom you are speaking to (Wayman, J. 1999).