Enrollment. As shown in Figure 1, the biometric sample of the individual is captured during the enrollment process (e.g., using a sensor for fingerprint, microphone for speaker recognition, camera for face recognition, camera for iris recognition). The unique features are then extracted from the biometric sample (e.g., image) to create the user’s biometric template. This biometric template is stored in a database or on a machine-readable ID card for later use during a matching process.

Matching. Figure 2 illustrates the biometric matching process. The biometric sample is again captured. The unique features are extracted from the biometric sample to create the user’s “live” biometric template. This new template is then compared with the template(s) previously stored, and a numeric matching (similarity) score(s) is generated based on a determination of the common elements between the two templates. System designers determine the threshold value for this verification score based upon the security and convenience requirements of the system.

Identification (one-to-many or 1:N comparison) determines if the individual exists within an enrolled population by comparing the live sample template to all stored templates in the system. Identification can confirm that the individual is not enrolled with another identity or is not on a predetermined list of prohibited persons. The biometric for the individual being considered for enrollment should be compared against all stored biometrics. For some credentialing applications, a biometric identification process is used at the time of enrollment to confirm that the individual is not already enrolled.

Verification (one-to-one or 1:1 comparison) determines whether the live biometric template matches with a specific enrolled template record. This requires that there be a “claim” of identity by the person seeking verification so that the specific enrolled template record can be accessed. An example would be presentation of a smart card credential and matching the live sample biometric template with the enrolled template stored in the smart card memory. Another example would be entry of a user name or ID number which would point to an enrolled template record in a database.

The selection of the appropriate biometric technology will depend on a number of application-specific factors, including the environment in which the identification or verification process is carried out, the user profile, requirements for matching accuracy and throughput, the overall system cost and capabilities, and cultural issues that could affect user acceptance. The table shows a comparison of different biometric technologies, with their performance rated against several metrics.

A key factor in the selection of the appropriate biometric technology is its accuracy. When the live biometric template is compared to the stored biometric template (in a verification application), a similarity score is used to confirm or deny the identity of the user. System designers set the threshold (match or no-match decision point) for this numeric score to accommodate the desired level of matching performance for the system, as measured by the False Acceptance Rate (FAR) and False Rejection Rate (FRR). The False Acceptance Rate indicates the likelihood that a biometric system will incorrectly verify an individual or accept an impostor. The False Rejection Rate indicates the likelihood that a biometric system will reject the correct person. Biometric system administrators will tune system sensitivity to FAR and FRR to get to the desired level of matching performance supporting the system security requirements (e.g., for a high security environment, tuning to achieve a low FAR and tolerating a higher FRR; for a high convenience environment, tuning to achieve a higher FAR and a lower FRR).

Some of the accuracy and usability limitations imposed by the use of a single biometric modality can be overcome by using multiple biometric modalities. Multi-modal biometrics enhance the overall matching accuracy through the use of multiple and independent biometric measurements. For example, the similarity score from a fingerprint measurement can be mathematically “fused” with an independent measurement of the vein pattern in the finger to yield a higher level of confidence in the identity of a person. In addition, multi-modal biometrics can provide a solution for those individuals who are unable to present a suitable biometric sample in one modality. An example would be offering the option to present either a fingerprint or iris for authentication. A person who has poorly defined fingerprint patterns due to age, occupation, or medical condition would be given the choice to enroll and use iris as their biometric modality of choice. If both sensors are present, the user can use whatever modality that they are best suited for. In this situation, there is no fusion of independent biometric measurements. As can be seen in Figure 3, multi-biometric systems can incorporate information from multiple modalities, instances, algorithms, sensors, samples, or any combination of the five. Arguably, such systems may also include other sources of information, including biographic or travel document-based information.

Biometrics and security have become an indispensable part of self-service, and Huabiao Technology has always adhered to security as its bottom line, providing users with safe and efficient services. For more information about the company and its offerings, visit the About Us page or reach out through the Support page.