provided by: 
Originally published at Internet.comThe Neurotechnology (formerly Neurotechnologija Biometrics) product line is centered on the implementation of biometric recognition systems in applications and devices; with entries available facilitating both fingerprint and facial recognition. Available products in the line include FingerCell, FaceCell, VeriLook, VeriFinger, and MegaMatcher.
VerFinger represents the vendor's core fingerprint matching technology, with support for 1:1 and 1:Many matching at comparison speeds of up to 40,000 fingerprints per second. SDKs are offered in both Standard (for stand-alone applications) and Extended (for client/server apps) enabling the creation of applications that run on Windows, Linux, or Mac OS X platforms (some features of the technology will be unavailable on Mac OS X or Linux; see the vendor's site for a breakdown). Similarly, VeriLook represents the vendor's core facial recognition technology, boasting an enrollment time of less than .3 seconds, comparison speeds of up to 100,000 faces per second, and support for both 1:1 and 1:Many matching. The VeriLook SDK is also offered in both Standard and Extended (for Web-based or network application development; includes all the Standard features as well as a ready-to-use matching server) flavors.
The FingerCell and FaceCell offerings are EDKs (Embedded Development Kits) that enable device manufacturers and developers to implement biometric recognition specifically in low power, compact, and mobile devices such as doors, entry gates, handheld PCs and cell phones. Windows CE and Linux (ARM) are both listed as potential target platforms for the EDKs; and source-code versions of the products are available for potential porting to other systems.
The FingerCell EDK features the ability to recognize even rotated or partial fingerprints in 1-to-1 or 1-to-many scenarios; and can, according to the vendor, match up to 700 fingerprints per second. FaceCell, on the other hand, can match up to 3000 faces per second, with face detection and template extraction taking 1-2 seconds, depending on frame size and processor speed (the vendor recommends 400MHz or faster processors).
Finally, MegaMatcher provides combined fingerprint and/or facial matching engine technology targeted to large scale applications or AFIS (Automated Fingerprint Identification Systems). MegaMatcher is capable of matching up to 400,000 templates per second using the "fused" (both fingerprint and facial) algorithm on a stand-alone PC with 3GHz CPU, with the facial recognition engine itself able to match up to 500,000 faces per second and the fingerprint engine matching up to 60,000 fingerprints per second. Clustered deployments are also possible for increased performance. The MegaMatcher SDK is offered in Light (client/server applications) or Standard (large scale clustered deployments) flavors.
Also available from the vendor is the Smartcard Finger-Match and Smartcard Face Match Add-ons, which allow both the storage of the fingerprint/facial template as well as the matching process to be performed directly on a JavaCard (the matching technology occurs via a microprocessor embedded on the card)-the user's tempalte data therefore doesn't have to leave the card for verification.
New features of the latest VeriFinger and VeriLook product releases include:
- Support for the development of Web-based apps, via provided samples, tutorials, and concurrent licensing options
- The aforementioned Extended version of the VeriLook SDK
- Live face detection, multiple template recognition, and multiple-face recognition in VeriLook; the product can detect multiple faces in a single frame and process all of them simultaneously, and can enable a face template to be formed from multiple images (front, side, etc.)
- New licensing options, including separate matcher/extractor licenses (for example, one matcher license for a Web server but multiple extractor licenses) and concurrent network licensing (unlimited installs with limited concurrent execution)
The Neurotechnology product line is available now; visit the vendor's Web site for further information.Author: EITPlanet Staff
Read article at Internet.com site