Fingerprint Biometrics

Our rigorous scientific development process has allowed us to achieve unprecedented levels of accuracy, speed, and efficiency. We’re able to correctly classify and enroll a greater number of ‘legacy’ fingerprints (such as rolled fingerprints captured through inking techniques). These unique, home-grown biometrics solutions are designed to support a wide range of use-cases: from time & attendance, to access control, border security, digital payments, smart cards, smart citizen programmes, and more. Through our market-leading innovations, we’re creating the most secure, most convenient biometric solutions. Our system includes a number of ‘world-first’ modules

Foreground Centroid for Rotation Immunity Module

Patent Application 2017/05007

Our geometric method detects the fingerprint foreground centroid as a reference point – negating the need for us to use the core as a reference point (which is the traditional approach). In this way, we’re able to classify even Plain Arch fingerprints (that normally require a topological method to be dealt with), alongside Left Loops, Right Loops, Whorl, and Tented Archs. Prints are automatically and instantly rotated from any angle, and translated from any side, fast-tracking the matching process and ensuring that valid prints are always accepted.

Quasi-singularity Detection Module

Patent Application 2017/05008

Our unique system is capable of accurately determining missing singular points, and of classifying fingerprints even when the delta is missing. By short-circuiting the scanning process and inferring where the transition line is drawn, we can rapidly increase the speed of fingerprint classification – ultimately giving users the best experience. For ink to digital classification, quasi-singularity detection enables a far greater percentage of prints to be enrolled.

Template Matching

Patent Application 2017/05009

Our custom-developed algorithm enables ‘everytime’ accuracy when comparing the query template with the reference template. We enhance the ridges and cater for variations caused by the skin’s condition: like wetness, markings, scars, dirt, rotation, translation issues and more. Our approach completely removes the problems associated with intraclass variations, and neutralises elastic deformations caused by mapping a three-dimensional finger, to the two-dimensional fingerprint scanner surface. The sophisticated fingerprint matching problem is, hence, trivialized to just establishing minutiae correspondences.

Download Brochure

Download Technical Architecture

To find out more, get in touch with us at biometrics@mmapro.co.za


ANPR & VMMR

Automatic Number Plate Recognition and Vehicle Make & Model Recognition are a special form of optical character recognition (OCR), software that enables computer systems to read automatically the registration number (license number) of vehicles from digital pictures.

Reading automatically the registration number means transforming the pixels of the digital image into the ASCII text of the number plate. While license plate recognition has special type of OCR technology, today optical character recognition (OCR) technology is considered strictly a type of technology - mainly software - that lets you scan paper documents and turn them into electronic, editable files. From the LPR/ANPR point of view the image quality is always key factor. Capturing of fast moving vehicles needs special technique to avoid motion blur which can decrease the recognition accuracy dramatically. To ensure the right image quality short shutter time need to be used with the combination of high- power illumination.