DigitalPersona announced at the CARTES & IDentification 2011 conference in Paris, France, that the company has open sourced its new MINEX-certified FingerJetFX™ fingerprint feature extraction technology. FingerJetFX, Open Source Edition (OSE), is free, portable software that device manufacturers and application developers can use to convert bulky fingerprint images into small, mathematical representations called fingerprint “templates” for efficient storage or comparison. Full source code is available from digitalpersona.com/fingerjetfx and is designed for use across a wide range of platforms and operating systems. The software has been released under the Lesser GNU Public License (LGPL), giving developers the flexibility to adapt the extractor to their particular platforms and applications. DigitalPersona® is also offering a separate commercially-licensed version that adds fast fingerprint matching for both identification and verification while still running in a small amount of memory.
International standards for fingerprint data formats enable applications to exchange biometric data across a wide range of devices, databases and applications – but only if the underlying fingerprint data is consistently and accurately obtained. FingerJetFX fills this gap. It can be embedded into chips, fingerprint sensors and mobile devices, or run on PCs and large database servers, providing uniform biometric data that can be used with any standards-based fingerprint matching engine. With FingerJetFX, solution providers, from mobile device manufacturers to national ID program integrators, can more reliably mix and match fingerprint-enabled hardware and backend biometric software from multiple vendors.
FingerJetFX is both fast and compact, requiring approximately 128KB of code space and 128KB of data space. Based on DigitalPersona technology that has been used on millions of devices, FingerJetFX can run on 32-bit and 64-bit CPUs as well as 32-bit microcontrollers, and does not require any special hardware such as DSPs or floating point. FingerJetFX is written in self-contained C++ and is designed to run on Linux, Android, Windows, Windows CE, real-time operating systems and even embedded systems without an OS. FingerJetFX is device-independent and can be used with simple 8 bits-per-pixel grayscale image data from any brand of fingerprint sensor or fingerprint image database. It meets and exceeds the PIV requirements for fingerprint template interoperability and is compliant with NIST’s MINEX Ongoing Test (see www.nist.gov/itl/iad/ig/ominex_test-results.cfm, SDK 3F).
“Quality and consistency of biometric data are becoming critical for new generations of fingerprint-enabled systems, from the smallest embedded devices to the largest database servers” said Jim Fulton, vice president of DigitalPersona. “With FingerJetFX, DigitalPersona is giving integrators and solution providers around the world a way to close a potential gap in their biometric systems before it creates problems that could undermine their success. By offering both open source and commercially-licensed options, we’re giving businesses, governments and other organizations the flexibility they need to address this growing issue.”