Fingerprint SDK-SDK
Overview
STARTEK proprietary software is available in modules that perform image capture, gain control, minutiae extraction, fingerprint enrollment and one-to-one or one-to-many verification matching. The software modules are based on open architecture standards. The Imaging Suite modules and Verification Suite modules can be used for fingerprint reader from different technology such as optics or chip sensor.
The Software Development Kit (SDK) integrates biometric technology for a variety of applications. When the software adopts fingerprint technology, it fulfills the desired demand for biometric technology and saves time for usage. Fingerprint SDK supports the users' current system structure to adapt to a fingerprint verification login system. Fingerprint SDK is suited to various fingerprint readers and helps the user to choose the suitable one without modifying code. STARTEK Fingerprint SDK supports a high performance cross-device matching algorithm.
Image Capture Module
Support the function that captures the image from different input devices such as optical reader FC200I, FM100 and chip sensor reader from ST, Infenieon, Authentec, Veridicom etc. The captured image can either be stored in the memory buffer or saved to a BMP/RAW file in the GRAYSCALE or BINARY mode. Provide the auto-snap function, i.e. the fingerprint is captured automatically if the quality is judged to be good enough, otherwise the function reports the status of the fingerprint (too dry, too wet or not well-positioned). Perform a diagnostics function to diagnose the device. Gain Control Module Support the function that tunes the image parameters (the parameters are different depending on the device) so as to get the best image quality for different input devices. For example, one can tune the brightness and contrast for FC200I or the GAIN, OFFSET and GAMMA for FM100 to get the desired image quality.
Verification Suit
Minutiae Extraction Module Support the function that extracts the minutiae from the image captured from different input devices. Extract only the significant minutiae and eliminates the noisy minutiae and borders. Create a template, which is called pcode, with the size between 100 ~ 256 bytes. Fingerprint Enrollment Module Support the function that performs learning from the minutiae of three fingerprints extracted from Minutiae Extraction Module. This learning algorithm will merge these minutiae of three fingerprints and creates a final enrolled template, which is called fpcode and will contain the actual information of the fingerprint, with the size between 100 ~ 256 bytes.
Fingerprint Verification Module Support the function that can do one-to-one verification of the captured fingerprint minutiae pcode from the Minutiae Extraction Module and the stored template fpcode from the Fingerprint Enrollment Module. Support the function that can do one-to-many Identification if the captured fingerprint minutiae matches one of the stored template database at the score higher than the set threshold degree. The module provides a cross-device verification with high performance and high speed as long as the minutiae is extracted from any fingerprint devices that meets 500DPI. This gives the best solution for e-commerce. Provide different security levels for usage that needs different FAR/FRR.