You are here

Iris Image Quality Tool Suite for Biometric Recognition

Description:

Biometric system performance depends on the quality of the acquired input samples. If sample quality can be improved, whether by sensor design, user interface design, or standards compliance, better performance can be realized. For those aspects of quality that cannot be designed-in, an ability to analyze the image and identify recognition-related defects and problems is needed. The ability to quickly acquire knowledge about sample quality can aid in deciding whether to initiate reacquisition from a subject, in real-time selection of the best sample from a set of samples, and in the selective invocation of different processing methods. Automated quality assessment is also useful for other purposes. In an enterprise context, image quality assessments can reveal trends, or time and location specific instances of poor image capture practices. For example, if operational collections consistently produce poor iris quality samples, it can indicate that additional training is needed or that screening may be better served by using another biometric modality.

Quality analysis is a technical challenge because it is most helpful when measures reflect the performance sensitivities of one or more target biometric matchers. To help facilitate universal interoperability of iris data across cameras and matchers, the National Institute of Standards and Technology (NIST), supported by the Department of Homeland Security, completed Iris Exchange (IREX) II Iris Quality Calibration and Evaluation (IQCE). This work identifies iris image properties that influence different vendor’s recognition accuracy and quantifies their effects. It is expected that this work will inform the development of the international standard ISO/IEC 29794 Biometric sample quality - Part 6: Iris image data.

This topic area seeks to enable innovative research that will contribute to efforts to develop, evaluate, and publish precise computation methods for quality metrics and increase the availability of commercial grade iris image quality software tools capable of operating on embedded, desktop, and server platforms.

PHASE I: Develop, evaluate, and document precise computation methods for quality metrics that are identified by IQCE to influence recognition performance.

PHASE II: Develop a suite of software tools designed to function on a variety of platforms to provide image quality assessment information implementing the quality metrics established in Phase I.

PHASE III: COMMERCIAL APPLICATIONS: Optimize and validate computation methods that yield quantitative scores that predict cross camera and matcher recognition outcomes on avariety of hardware platforms. Develop, market and license software packages tailored to operate on different platforms that will perform this function.

US Flag An Official Website of the United States Government