FaceVACS-SDK

Cognitec's Face Recognition SDK 5.0 delivers face recognition to a broad range of applications and allows to use face recognition as a biometric authentication model.

FaceVACS-SDK uses modern techniques of programming with consistent, logically clear programming interfaces. It covers the basic functions (use cases) of Enrollment, Verification, Identification and defines abstractions to support user defined applications. Additionally, FaceVACS-SDK 5.0 supports the finding of human faces and eyes in images.

FaceVACS-SDK 5.0 supports ISO Biometric Interchange Format 19794-5 and allows to determine certain characteristics of portrait images. The portrait image quality or suitability for photo-id card documents and for facial recognition can be checked.

FaceVACS-SDK 5.0 can not only work with popular two dimensional intensity data but also with corresponding range data (3D data).

FaceVACS-SDK is a product that Cognitec offers to Value-Added Resellers who intend to develop their own system, using FaceVACS technology. Please contact sales@cognitec.com for details.

Features

• A12T6, B2T6 algorithm for latest biometric performance improvements
• A11L1T6 for backward compatibility
• State of the art face recognition API (C++)
• Complete Microsoft .NET programming environment integration
• Support for enrollment, verification and identification operations
• Multiple face and eye finding
• Thread safety
• Support intensity image gathering from image streams (live video) and image collections (image databases)
• Intensity image interface with support for color images
• Support of Intel Performance Primitives (IPP) Libraries
• Abstract interfaces to support easy adaption to various video streaming hardware and databases
• BioAPI 1.1 Verification Engine BSP Compliant
• Suite of tools to perform biometric evaluations on data: e.g. generation of identification match lists and similarity matrix data
• Portrait characteristics determination (used for image quality assessment), i.e. pose deviation, exposure determination, glasses detection, eye closed detection, uniform lighting detection, image and face geometry determination
• ISO 19794-5 format, JPG 2000 support
• Image transformation, such that ISO Token Frontal, ISO Full Frontal images can be created
• Detailed documentation including API reference and user guide
• Fully documented examples illustrating the main use cases and providing hints how to create customized implementations

Major features that are new compared to the previous version 4.0 are:

• Improved algorithms A12T6 and B2T6
• Improved face and eye finder
• Improved sample quality measure