A mass fingerprint encoding and comparison method based on multi-feature fusion

Abstract

The invention relates to a multi-feature fusion mass fingerprint feature encoding and comparison method, comprising: inputting mass registered fingerprints and a fingerprint to be identified, and marking a unique ID identification code for all the registered fingerprints. Construct the feature matrix and minutiae matrix of the fingerprints to be recognized and all registered fingerprints respectively; construct a fingerprint rapid screening tree. Register fingerprints with similar feature evaluation indicators at the leaf nodes of the screening tree to form a list of registered fingerprints of the same type; use the feature matrix of the fingerprints to be compared to retrieve registered fingerprints with similar feature evaluation indicators with the fingerprints to be compared in the fast screening tree List, count the occurrence frequency of each fingerprint ID identification code in the list, take out several fingerprints with the highest frequency to form a registered fingerprint suspect queue; compare the fingerprint detail point matrix of fingerprints to be compared one by one with the registered fingerprint detail points in the suspect queue The matrix is ​​compared, and finally the registration fingerprint with the highest similarity is compared as the comparison result of the method.

Description

technical field [0001] The invention provides a multi-feature fusion mass fingerprint encoding and comparison method, which belongs to the field of biological feature identification, and in particular relates to a fingerprint identification method. Background technique [0002] Compared with other biometric identification technologies, fingerprint automatic identification technology has excellent accuracy and high economic practicability. Fingerprint identification technology is divided into two processes: (1) fingerprint registration process. By reading the fingerprint image, the fingerprint area in the fingerprint image is found, and the features in the fingerprint area that can distinguish the fingerprints from each other are extracted. Save these feature data in the database and use it as the representative of this fingerprint to complete the fingerprint registration process; (2) feature comparison process. After extracting the features of the newly collected fingerpri...

AI Technical Summary

Problems solved by technology

For massive fingerprints (such as hundreds of millions or even billions of fingerprints), on the one hand, when extracting fingerprint features, it is not possible to construct fingerprint features with the ability to accurately distinguish massive fingerprints only by using traditional minutiae features or line features; On the other hand, the idea of ​​comparing the fingerprints to be compared with all the registered fingerprints one by one is not advisable due to the huge workload of the comparison, and it will consume the comparison time in units of months

Images