We have been leading the development of biometric algorithms ever since we started. This means that we can now present the industry’s fastest response time from touch to identification and authentication (≤ 0.2 seconds on average). To us at Fingerprints, security is very important, and we are leading efforts for development and standardization in this area.
OUR PROPRIETARY ALGORITHM
Fingerprints’ patented Distinct Area Detection (DAD) algorithm is a feature-based algorithm that looks for things that are unique in its surroundings. In our application, the algorithm locates distinct areas in the three-dimensional fingerprint image derived from the sensor.
The design of the fingerprint-recognition algorithm is crucial to the performance of an embedded system. Not only is the performance with respect to FRR and FAR affected, but the requirements of processor power and memory are strongly dependent upon the type of algorithm. Getting this right can help deliver a low-cost system with low power consumption that is suited to an embedded environment such as a smart card. Fingerprints’ focus in algorithm development is on the design of verification and identification algorithms based on the DAD method.
The DAD algorithm works to locate distinct areas in the three-dimensional fingerprint image derived from the capacitive sensor. A ‘distinct area’ is a part of the image containing characteristic information and includes, but is not limited to, the fine details of the fingerprint.
The algorithm consists of two steps – enrolment and verification/identification. In the enrolment step, a number of distinct areas are extracted from the fingerprint image. The areas together with their geometric relationships form a template that is unique to each fingerprint. In the verification/identification step, the template is used as an operator acting on the fresh fingerprint image. If the match is approved, the authentication of the person is completed. One of the benefits of using this technology is that to get a match you will need both the fresh 3D fingerprint image as well as the stored template.