Fingerprint identification is the most common and cost-effective biometric alternative to RFID cards and the good old pen & paper method. The identification process uses recognition of the ridges and spaces on the fingertips, that form structure unique to each person. Physiologically, a fingerprint is a structure that appears as pattern of papillary ridges formed by the dermal papillae that lie beneath them. A papillary image is a pattern unique to each individual, developed before birth, which does not change with age and rebuilds itself after damage.
Principle of the Method
At the heart of the biometric method of fingerprint identification is the technology for obtaining a graphic image of the fingerprint, the creation of its vector model and the subsequent search through the system’s template database for a matching template. The graphic image of the fingerprint is transformed into a special mathematical template for comparison. The original image of the papillary structure is discarded immediately after the initial transformation. This makes reverse generation of the biometric images used in this process impossible.
Steps of Creating a Biometric Template
- Improvement of the original fingerprint image by increasing the sharpness of the papillary line borders.
- Identification of the field of the fingerprint papillary line orientation. This is achieved by splitting the image into square blocks with sides of more than 4 pixels and calculation of the t-orientation angle of the fingerprint lines based on the brightness gradients.
- Fingerprint image binarization – conversion of the image to black and white by thresholding.
- Thinning image lines of the fingerprint – performed until the lines no longer reach 1 pixel width.
- Identification of minutia – unique characteristics of each fingerprint (end, split, break, etc.). The recorded image is split into the 3×3 pixel blocks and then the number of black (non-zero) pixels is calculated around the center.
- Saving the coordinates and angles of the minutia identified into vector format.
Fingerprint Matching Algorithm
- Obtaining an image from the scanner.
- Finding characteristic fingerprint points.
- Saving the identified points to the template.
- Comparing the scanned template to the previously saved images in the fingerprint database.
Advantages of the Fingerprint Identification System
- Easy and convenient to use.
- Most affordable biometric solution available.
- Identification is fast and reliable.
- Well-developed, reliable technology.