Alignment-free fingerprint matching method and calculation circuit based on encryption circuit

Abstract

The invention discloses an alignment-free fingerprint matching method based on an encrypted circuit. The alignment-free fingerprint matching method based on the encrypted circuit comprises the steps that a detail point template set of fingerprints and an onsite sample detail point set are established, a line characteristic set OT and a local structure characteristic set DT are established based on the detail point template set, and a line characteristic set OQ and a local structure characteristic set DQ are established based on the onsite sample detail point set; point-by-point comparison is carried out on elements in the line characteristic set OT and elements in the line characteristic set OQ and on elements in the local structure characteristic set DT and elements in the local structure characteristic set DQ, and thus an NT*NQ matrix UO and an NT*NQ matrix UD are formed; similarity computing is carried out on the matrix UO and the line characteristic set OQ, the similarity computing is carried out on the matrix UD and the local structure characteristic set DQ, and the matching of the fingerprints is completed according to similarity. The invention discloses a computing circuit at the same time. The alignment-free fingerprint matching method based on the encrypted circuit and the computing circuit achieve the purpose that the fingerprint matching can be achieved without fingerprint alignment on the basis that private data of users and private data of authentication servers are protected.

Description

technical field [0001] The invention relates to the field of information security, in particular to an encryption circuit-based alignment-free fingerprint matching method and a calculation circuit. Background technique [0002] When conducting identity authentication based on biometric identification in an open network, an important challenge is that the two sides of the authentication - the server and the user often do not trust each other. Both parties only want the authentication result, but do not want the other party to know their private biometric data. . In the existing method, the recognition program running on the server or client needs to match the pre-stored template with the user's on-site biometrics to obtain the authentication result, which requires knowing the data of both parties. In order to solve the contradiction between data privacy and authentication in an open network environment, in recent years, researchers have designed authentication protocols with...

AI Technical Summary

Problems solved by technology

But alignment accuracy cannot be guaranteed

All fingerprint matching methods based on global features need to align two fingerprints, which requires complex algorithms and is difficult to implement with SFE technology. At present, some schemes use auxiliary data, but often require the fingerprint area in the image to be complete and clear. The fingerprint area is unavailable

Images