MBARK - Multimodal Biometric Application Resource Kit
Once a tool primarily used by law enforcement to help identify criminals, biometric technologies increasingly are being used by government and the private sector to authenticate a person's identity, provide security at the nation's borders and restrict access to secure sites—both buildings and computer networks. New software and other tools that can be used to help build improved biometric applications are now available from the National Institute of Standards and Technology (NIST).
Most biometric systems are "unimodal," meaning they rely on a single distinguishing physical characteristic—such as a fingerprint—for authenticating identity. But using a single feature can present problems. Poor illumination could make a face image unrecognizable; dirty or damaged sensor plates can affect fingerprint equipment.
A multimodal system that has several sources of information, including fingerprint, face, and iris data, can be more flexible and reliable. But most biometric equipment, including the sensors that capture data and the database that stores the information, are not interoperable. Organizations must either purchase a complete system or develop "middleware"—custom integration software—to link together applications.
Despite existing efforts, building modern biometric applications (or clients) that are flexible with respect to these problems—changes in sensors, workflow, configuration, and responsiveness—remains both difficult and costly.
The Multimodal Biometric Application Resource Kit, or MBARK, reduces the complexity and costs of implementing such an application. MBARK is public domain source code that may be leveraged to develop the next generation of biometric and personal identity verification applications.
Incorporating the MBARK libraries can yield a variety of enhancements critical for the success of any real-world system. For example, MBARK provides a usability tested and consistent user interface. MBARK provides operators the means to quickly recover from both minor mistakes and major hardware failures. In addition, the use of Extensible Markup Language (XML) facilitates true sensor interoperability via plug-ins and allows for changes in workflow on-the-fly.