Privacy protection biological information matching method and system based on terminal equipment
By using deep learning algorithms on terminal devices to extract and split biometric features, and only storing parts that cannot be represented separately in the database, it solves the privacy threat problem caused by database leakage in traditional bioinformatics identification solutions, and achieves high-security biometrics Information matching and verification.
Patent Information
- Application Number
- CN202411014485.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-08
AI Technical Summary
When traditional biometric information identification solutions are leaked, user privacy faces a serious threat, and there is a risk of illegal users impersonation, which cannot effectively protect the security of biometric information.
Using the privacy-protected biological information matching method based on terminal devices, the biological features are extracted through deep learning algorithms, split into two independent parts, discarding the key part A, and only part B is stored in the database. The feature data synthesis algorithm is used when matching Reconstruct the feature dataset for verification to ensure that the original biological features cannot be restored even if the database is compromised.
It significantly improves privacy security and data integrity, ensures the accuracy and security of bioinformatics matching, and is suitable for high-security biometric authentication scenarios, such as finance, security, medical and other fields.
Smart Images

Figure CN118916862A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of biometric information security technology, and in particular to a privacy-preserving biometric information matching method based on a terminal device and a system thereof. Background Art
[0002] With the development of computer vision and deep learning technology, biometric information extraction technology has been widely used in identity authentication scenarios. During the registration process, the system extracts the user's biometric features and stores them in the database in the form of vectors, etc., for 1:1 identity authentication scenarios and 1:n identity recognition scenarios. The former provides the user's identity information, compares the biometric information collected on site with the biometric information of the corresponding user stored in the database, and authenticates the user's identity; the latter does not provide the user's identity information. After collecting the user's biometric information on site, it searches the database for biometric information that is sufficiently similar to it, thereby identifying the user's identity.
[0003] The development of computer vision has greatly facilitated the biometric-based identity authentication technology, but it has also brought new challenges to the security protection of user personal information. In order to solve the problem of user information security protection and prevent the illegal theft of user biometric information, encryption-based solutions have been proposed, including the use of functional encryption, fuzzy extractors, homomorphic encryption and other technologies to encrypt biometric information. Through these technologies, the user's biometric information is stored in the database in an encrypted form. Even if the attacker invades the database, the original biometric information cannot be restored from the encrypted information. Only when the legitimate user provides biometric information that matches the data in the database can the information be decrypted and the identity authentication operation be performed. In particular, the identity authentication scheme based on functional encryption and homomorphic encryption, in which the system can directly calculate the similarity between biometrics on the encrypted data, protects the user's biometric information. In addition, in order to improve the computational efficiency of identity authentication, technologies such as cloud computing are applied to identity authentication scenarios.
[0004] In traditional biometric identification solutions, each user's facial information is directly stored in the cloud database in the form of a vector. This storage method has great security risks. Attackers may steal data in the database and manipulate the data transmission channel to obtain the user's facial information, which not only leaks the user's privacy, but also creates the risk of illegal users impersonating.
[0005] In some technologies, biometric data is directly stored in the database. Once the database is attacked, user privacy is seriously threatened. Therefore, it is particularly important to develop a technology that cannot restore the original biometric information even if the database is leaked. Summary of the invention
[0006] The purpose of the present invention is to provide a privacy-preserving biometric information matching method and system based on terminal devices to solve existing technical problems.
[0007] In order to achieve the above object, the technical solution adopted by the present invention is as follows:
[0008] A privacy-preserving biometric information matching method based on terminal devices. When biometric information is collected and stored, the biometric information is feature extracted, and then a feature data set is generated. The feature data set is split, and then part of the information is discarded, and the other part of the information is stored in a database. When biometric information matching is required, the collected biometric information is processed according to the storage process, and then the information in the database is compared with the matched data. If the comparison is successful, the matching is successful, otherwise access is denied.
[0009] Furthermore, when biometric information is collected and stored, a deep learning algorithm F0 is used to extract features from the biometric information obtained through the camera and fingerprint collector, and a predetermined algorithm rule F1 is applied to generate a feature data set S based on the extracted features. The feature data splitting and transformation algorithm F2 is used to split S into two parts A and B, where A and B are independent of each other, and neither part can represent the original biometric feature alone. Part A is discarded, and only part B is retained and stored in a secure database to ensure that even if the database is illegally accessed, the complete biometric information cannot be restored.
[0010] Furthermore, when biometric information matching is required, the terminal device again extracts and generates a feature data set S' to be matched through F0 and F1, and similarly applies the F2 algorithm to split S' into A' and B'. Using the feature data synthesis algorithm F3, the A' generated by the current terminal is paired with the B data in the database to generate a test data set S1. The synthesized S1 is compared with the originally extracted S'. If the two match, the verification is successful, otherwise access is denied.
[0011] A system for a privacy-preserving biometric information matching method based on terminal devices includes a system data center, networking equipment and a plurality of embedded biometric information input integrated terminal devices. The plurality of embedded biometric information input integrated terminal devices are connected to the system data center via the networking equipment. The plurality of embedded biometric information input integrated terminal devices perform calculation processing on the input biometric information before storage, and then store the calculated data in the system data center. The stored biometric information is not complete biometric information. When biometric information matching is required, the collected information is calculated and matched with the information in the system data center.
[0012] Furthermore, the embedded biometric information input integrated terminal device includes a collection module and an embedded terminal device. The collection module transmits the collected data to the embedded terminal device for processing, and the embedded terminal device implements storage and matching of incomplete biometric information data.
[0013] Furthermore, the embedded terminal device includes a processing module, a feature extraction algorithm F0 module, a feature set arrangement algorithm F1 module, a feature data transformation algorithm F2 module and a feature data synthesis algorithm F3 module. The feature extraction algorithm F0 module, the feature set arrangement algorithm F1 module, the feature data transformation algorithm F2 module and the feature data synthesis algorithm F3 module are all connected to the processing module, the processing module is connected to the system data center, and the processing module is connected to the acquisition module via the feature extraction algorithm F0 module.
[0014] The present invention has the following beneficial effects due to the adoption of the above technical solution:
[0015] The present invention enhances privacy protection: even if the database is hacked, the key part A is discarded and the leaked data cannot restore the original biometric features, which significantly improves privacy security and ensures data integrity: through the close integration of terminal devices and specific algorithms, the accuracy and security of biometric information matching are ensured, and it is suitable for various scenarios that require high-security level biometric authentication, such as finance, security, medical and other fields. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 is a biological information input flow chart of the present invention;
[0017] Figure 2 is a biological information matching flow chart of the present invention;
[0018] Figure 3 It is a block diagram of the system module of the present invention;
[0019] Figure 4 It is a schematic diagram of the principle structure of the system of the present invention. DETAILED DESCRIPTION
[0020] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and preferred embodiments. However, it should be noted that many details listed in the specification are only for the purpose of enabling the reader to have a thorough understanding of one or more aspects of the present invention, and these aspects of the present invention can be implemented even without these specific details.
[0021] like Figure 1-2As shown, a privacy-preserving biometric information matching method based on a terminal device is provided. When the biometric information is collected and stored, the biometric information is feature extracted, and then a feature data set is generated. The feature data set is split, and then part of the information is discarded, and the other part of the information is stored in a database. When biometric information matching is required, the collected biometric information is processed according to the storage process, and then the information in the database is compared with the matched data. If the comparison is successful, the matching is successful, otherwise access is denied.
[0022] In the embodiment of the present invention, Figure 1 As shown, when biometric information is collected and stored, a deep learning algorithm F0 is used to extract features from the biometric information obtained through the camera and fingerprint collector, and a predetermined algorithm rule F1 is applied to generate a feature data set S based on the extracted features. The feature data splitting transformation algorithm F2 is used to split S into two parts A and B, where A and B are independent of each other, and neither part can represent the original biometric feature alone. Part A is discarded, and only part B is retained and stored in a secure database to ensure that even if the database is illegally accessed, the complete biometric information cannot be restored.
[0023] In the embodiment of the present invention, Figure 2 As shown, when biometric information matching is required, the terminal device again extracts and generates a feature data set S' to be matched through F0 and F1, and also applies the F2 algorithm to split S' into A' and B'. The feature data synthesis algorithm F3 is used to pair the A' generated by the current terminal with the B data in the database to generate a test data set S1. The synthesized S1 is compared with the originally extracted S'. If the two match, the verification is passed, otherwise access is denied.
[0024] Feature extraction algorithm F0: It is mainly used to extract image features. The algorithm requires both efficiency and accuracy. Feature set arrangement algorithm F1: It sorts the features extracted by F0 according to certain rules to form a feature set S, which is convenient for subsequent effective splitting. Feature splitting transformation algorithm F2: It splits the feature set S sorted by F1 into two parts A and B, where A and B are independent of each other, and neither part can represent the original biological feature alone. Feature data synthesis algorithm F3: It is the reverse process of F2, which re-synthesizes A and B into S.
[0025] This solution is used in the security field, such as access control systems, safes, or any application that requires high-precision face recognition. Two factors need to be considered when selecting an algorithm: accuracy and reliability.
[0026] A system for privacy-preserving biometric information matching method based on terminal devices, such as Figure 4As shown, it includes a system data center, networking equipment and several embedded biometric information input integrated terminal devices. Several embedded biometric information input integrated terminal devices are connected to the system data center via the networking equipment. The embedded biometric information input integrated terminal devices perform calculations on the input biometric information before storage, and then store the calculated data in the system data center. The stored biometric information is not complete biometric information. When biometric information matching is required, the collected information will be calculated and matched with the information in the system data center.
[0027] In the embodiment of the present invention, the embedded biometric information input integrated terminal device includes a collection module and an embedded terminal device. The collection module transmits the collected data to the embedded terminal device for processing, and the embedded terminal device implements storage and matching of incomplete biometric information data.
[0028] In the embodiment of the present invention, Figure 3 As shown, the embedded terminal device includes a processing module, a feature extraction algorithm F0 module, a feature set arrangement algorithm F1 module, a feature data transformation algorithm F2 module and a feature data synthesis algorithm F3 module, the feature extraction algorithm F0 module, the feature set arrangement algorithm F1 module, the feature data transformation algorithm F2 module and the feature data synthesis algorithm F3 module are all connected to the processing module, the processing module is connected to the system data center, and the processing module is connected to the acquisition module via the feature extraction algorithm F0 module. The system can effectively protect the user's biometric information privacy while ensuring the accuracy of biometric recognition.
[0029] The system enhances privacy protection: even if the database is breached, the key part A is discarded and the leaked data cannot restore the original biometrics, significantly improving privacy security. Ensure data integrity: through the close integration of terminal devices and specific algorithms, the accuracy and security of biometric information matching are ensured. Wide application: suitable for various scenarios that require high-security biometric authentication, such as finance, security, medical and other fields.
[0030] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.
Claims
1. A privacy-preserving biometric information matching method based on terminal devices, Features: When biometric information is collected and stored, the features of the biometric information are extracted, and then a feature data set is generated. The feature data set is split, and then part of the information is discarded, and the other part of the information is stored in the database. When biometric information matching is required, the collected biometric information is processed according to the storage process, and then the information in the database is compared with the matched data. If the comparison is successful, the match is successful, otherwise access is denied.
2. According to claim 1, a privacy protection biometric information matching method and system based on terminal equipment, Features: When collecting and storing biometric information, a deep learning algorithm F0 is used to extract features from the biometric information obtained through cameras and fingerprint collectors. A predetermined algorithm rule F1 is applied to generate a feature data set S based on the extracted features. The feature data splitting and transformation algorithm F2 is used to split S into two parts A and B, where A and B are independent of each other and neither part can represent the original biometric feature alone. Part A is discarded and only part B is retained and stored in a secure database to ensure that even if the database is illegally accessed, the complete biometric information cannot be restored.
3. According to claim 1, a privacy protection biometric information matching method and system based on terminal equipment, Features: When biometric information matching is required, the terminal device again extracts and generates the feature data set S' to be matched through F0 and F1, and also applies the F2 algorithm to split S' into A' and B'. The feature data synthesis algorithm F3 is used to pair the A' generated by the current terminal with the B data in the database to generate a test data set S1. The synthesized S1 is compared with the originally extracted S'. If the two match, the verification is passed, otherwise access is denied.
4. A system for a privacy-preserving biometric information matching method based on a terminal device according to any one of claims 1 to 3, Features: It includes a system data center, networking equipment and several embedded biometric information input terminal devices. Several embedded biometric information input terminal devices are connected to the system data center via networking equipment. The embedded biometric information input terminal devices process the input biometric information before storage, and then store the calculated data in the system data center. The stored biometric information is not complete biometric information. When biometric information matching is required, the collected information will be calculated and matched with the information in the system data center.
5. A system for a privacy protection biometric information matching method based on a terminal device according to claim 4, Features: The embedded biometric information input integrated terminal device includes a collection module and an embedded terminal device. The collection module transmits the collected data to the embedded terminal device for processing, and the embedded terminal device realizes the storage and matching of incomplete biometric information data.
6. A system for a privacy-preserving biometric information matching method based on a terminal device according to claim 5, Features: The embedded terminal device includes a processing module, a feature extraction algorithm F0 module, a feature set arrangement algorithm F1 module, a feature data transformation algorithm F2 module and a feature data synthesis algorithm F3 module. The feature extraction algorithm F0 module, the feature set arrangement algorithm F1 module, the feature data transformation algorithm F2 module and the feature data synthesis algorithm F3 module are all connected to the processing module, the processing module is connected to the system data center, and the processing module is connected to the acquisition module via the feature extraction algorithm F0 module.
