Fingerprint synthesis method and system based on orientation field model and Gabor filter

Abstract

This invention discloses a fingerprint synthesis method and system based on new direction module and improved Gabor filter, including steps: generate fingerprint template image, create fingerprint profile image for the first step, and on the second step generate fingerprint orientation field through new joint direction field module, the third step is to generate density map with a new presentation method of fingerprint density map, on the fourth step create fingerprint ridge texture through improved Gabor filter; fingerprint generated according to pressing image, a series of transformation has been generated on the fingerprint template images, including adding scratches, texture translation, the expansion or corrosion of the ridge, the elastic deformation of ridge line, noise and smooth of ridge line, translation and rotation of image, change contrast, add background noise. The system includes: fingerprint template image module and fingerprint image compression unit. The present invention can generate realistic image of the fingerprint. It plays an important value in biological identification technology.

Description

technical field The invention relates to pattern recognition, image processing and information processing technology, in particular to image filtering and method using direction field to realize fingerprint synthesis processing. Background technique Fingerprint identification is currently the most widely used biometric identification method. In the research of automatic fingerprint identification algorithm, a large-scale fingerprint database is needed to test and optimize the algorithm. However, there are many shortcomings in the traditional method of collecting fingerprints to build a database. First of all, it takes a lot of cost and a lot of time to collect a large number of fingerprints. Secondly, no matter for the person who collects fingerprints or the person whose fingerprints are collected, the collection process is very cumbersome and error-prone. Moreover, collecting fingerprints also involves personal privacy. If the method of synthetic fingerprints can be use...

AI Technical Summary

Problems solved by technology

But it still cannot accurately represent the direction field, and in many cases can only achieve approximate results

Images