Encrypted Biometric Matching Using Euclidean Distance Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The leakage of biometric data during the biometric matching process poses a significant risk to user privacy, as the data is stored and transmitted in plaintext form, exposing it to potential security breaches.
Innovation Solution
A biometric matching method that encrypts biometric vectors using private keys at both the terminal device and server, forming encrypted data with a computational operator for Euclidean distance, allowing matching without exposing plaintext data, thus ensuring security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If biometric data is stored and transmitted in plaintext form for matching, then matching efficiency is improved, but data security deteriorates
Solution Approach 1:
The patent introduces encrypted biometric data as an intermediary form that enables matching operations without exposing plaintext data. The encryption scheme allows the server to perform matching on ciphertext, and the terminal device to verify results, thus mediating between security requirements and matching functionality.
Solution Approach 2:
The patent transforms biometric data from plaintext to encrypted form, changing its parameter state. This parameter change (from readable to encrypted) prevents direct access to original biometric information while maintaining the mathematical properties needed for Euclidean distance calculation and matching operations.
2Object-affected harmful factors
If biometric data is encrypted to protect privacy, then data security is improved, but matching complexity increases
Solution Approach 1:
The patent performs preliminary encryption of biometric data before the matching process begins. The terminal device encrypts the query biometric data and sends the ciphertext to the server, which has already encrypted the reference biometric data stored in the database. This preliminary action eliminates the need for complex decryption operations during matching.
Solution Approach 2:
The patent replaces the traditional mechanical approach of decrypting data for matching with a cryptographic substitution approach. Instead of converting encrypted data back to plaintext for comparison, the system performs matching operations directly on encrypted data using cryptographic primitives that preserve Euclidean distance relationships.
3Measurement precision
If plaintext biometric data is transmitted between terminal and server, then matching accuracy is maintained, but privacy leakage risk increases
Solution Approach 1:
The patent uses encrypted biometric data as an intermediary that preserves matching accuracy while preventing privacy leakage. The encryption scheme maintains the mathematical structure needed for accurate Euclidean distance calculation, allowing the server to compute precise matches without ever accessing the plaintext biometric information.
Solution Approach 2:
The patent applies different quality requirements to different parts of the system. The server stores and processes only encrypted data with cryptographic properties, while the terminal device handles the encryption/decryption operations. This local quality differentiation allows matching accuracy to be maintained at the terminal while the server operates with enhanced security.
Data Source
AI summary
The present application discloses a biometric matching method, terminal device, server, system, and medium. The method includes: performing, based on a first private key, an acquired biometric vector to be matched, a preset generating element, a second private key, and first encrypted data, a number of interactions and processes with a server to obtain second encrypted data (S201), the first encrypted data being pre-obtained by the terminal device through encrypting a sample biometric vector using the generating element and the first private key and sent to the server; and sending the second encrypted data to the server to cause the server to obtain, using the second encrypted data, the generating element, the second private key, and a preset Euclidean distance matching threshold, a matching result for the biometric vector to be matched and the sample biometric vector (S202).