Ocular biometrics databases: Standards and data interchange formats

The current challenges to developing commercial Ocular Biometric systems are to create applications that tolerate unconstrained environments. Designs for sensors, user interfaces, and algorithms must perform well under variable illumination, poses, and camera distances, with low-quality images, and within other real-world conditions. Practical deployment requires interoperability of data and processing algorithms from multiple systems. Facilitating meaningful progress ultimately demands Biometric Databases that contain real-world data. Accelerated standardization offers practical means to organize the tremendous volume of emerging biometric information. Standards enable data exchanges that are grounded on comparable and compatible formats. This paper describes the development of a research-oriented Ocular Biometric Database that contains images acquired in conditions relevant to unconstrained environments. The data was collected with different sensors and imaging systems under diverse conditions representing real-world applications. Images and related data were stored in a database adhering to extended biometric standards requirements and biometric data-interchange formats. Assessment and applicability reviews were performed on existing national and international standards pertaining to ocular biometrics and were used during the development. The paper describes the design approach with a goal to achieve interoperability and standardization of data and results. It also addresses recent developments in biometric data acquisition, pre- and post-processing, information storage, exchange, and access control.