Secure iris privacy protection method based on K-nearest neighbor

Abstract

The present invention relates to a secure iris privacy protection method based on K-nearest neighbor, comprising the following steps: S1, generating a random binary classification data set; S2, generating an iris privacy protection. The secure iris recognition technology of the present invention has good performance and secure biometric recognition technology, with high-precision recognition effect and high security.

Description

Technical Field

[0001] The present invention relates to the field of biometric data privacy protection in biometric recognition technology, and more specifically, to a secure iris privacy protection method based on K-nearest neighbors. Background Art

[0002] In recent years, with the rapid development of biometric technology, fingerprint, facial, and iris recognition technologies have become increasingly commonplace, widely adopted in mobile phones, computers, and access control systems. While biometrics bring convenience to people's lives, they also pose serious security risks. Unlike traditional passwords, users can update their passwords if they are compromised. However, due to the uniqueness and high importance of personal biometric information, the damage caused by such a leak is irreversible. Traditional biometric recognition solutions rely directly on plaintext for identification, posing a serious risk of privacy leakage. If personal biometric data is stored in plaintext in a database, theft by an attacker could have irreversible consequences for individuals and society. Therefore, many advocate the need to protect biometric data to prevent the leakage of personal biometric information. According to ISO / IEC 24745, privacy protection for biometric templates must meet the following requirements: irreversibility, revocability, and non-linkability. Iris recognition, as one of the most important biometric recognition technologies, offers many advantages over other biometric technologies, such as stability, anti-counterfeiting, and anti-spoofing, making it suitable for identity verification. Traditional iris recognition typically involves image preprocessing, iris localization, iris normalization, image equalization, and iris feature extraction. Iris image data is converted into feature vectors or binary codes for recognition. Because the iris contains rich texture information, a well-developed recognition algorithm can achieve high accuracy and has broad application prospects.

[0003] Currently available biometric template protection methods can be divided into two categories: biometric encryption systems and revocable biometric technologies. Based on how the key is generated, biometric encryption systems can be further divided into key-bound encryption systems and key-generating encryption systems. In key-bound encryption systems, the key is generated independently of the data. In key-generating encryption systems, the key is generated from the data. Revocable biometric technology transforms data into another domain using a function, performs operations and matching in the transformed domain, and ensures that it is difficult to recover the original data from the transformed data. Revocable biometrics can also be divided into biometric information salting and irreversible transformation. Biometric information salting transforms the biometric using a function defined by a user-specific key. Irreversible transformation protects the biometric template by applying an irreversible transformation function to the feature template. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a secure iris privacy protection method based on K-nearest neighbors that can meet the privacy requirements of iris template protection and support effective iris recognition.

[0005] The technical solution adopted by the present invention to solve the technical problem is to construct a secure iris privacy protection method based on K nearest neighbors, including the following steps:

[0006] S1, generate a random binary classification dataset;

[0007] S2. Generate iris privacy protection.

[0008] According to the above scheme, step S1 specifically includes the following contents:

[0009] S101. Randomly generate a 32-bit binary string using the Random function;

[0010] S102, dividing the binary string into two blocks of binary strings, each block having 16 bits;

[0011] S103, randomly inverting the two binary strings by x bits;

[0012] S104, repeatedly generating five different 32-bit binary strings;

[0013] S105, dividing the newly generated string into two blocks again;

[0014] S106, randomly invert the first block by y bits, and keep the second block unchanged;

[0015] S107, repeatedly generating five different 32-bit binary strings;

[0016] S108, save the binary strings generated by S101, S104, and S107 into a txt file at once;

[0017] S109. Repeat the operation N times or more as the final random binary classification data set.

[0018] According to the above scheme, step S2 specifically includes the following contents:

[0019] S201, performing block processing on the binary iris data x;

[0020] S202, calculating the distance between the iris data block i and the first block of the txt text in the classification data set, using Hamming distance for distance calculation;

[0021] S203, calculate the distance between the i-th block and the classification data set, sort them, and select the k closest binary classification strings to save;

[0022] S204, the k most recent binary classification strings are replaced by the second block, voting is performed using the second block, and the last bit of each bit is saved as the final result;

[0023] S205, after repeated operations of steps S202-204, each block is operated in turn and then spliced;

[0024] S206, saving the final result obtained in step S205 as an iris template.

[0025] The implementation of the K-nearest neighbor-based secure iris privacy protection method of the present invention has the following beneficial effects:

[0026] 1. The secure iris privacy protection method of the present invention is a technology that supports effective identity recognition functions while ensuring the security of iris data. It converts and encrypts the extracted iris feature vector or iris binary code. When converting and encrypting iris data, the iris biometric information will change to a certain extent, reducing the impact of the change on recognition accuracy. The performance recognition technology is good and the accuracy is high.

[0027] 2. The secure iris privacy protection method based on K nearest neighbors of the present invention can simultaneously meet the irreversibility, revocability and irrelevance proposed by the international standard ISO / IEC24745 for iris template protection, while having high-precision recognition effect and high security, and the method can obtain good recognition performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0029] Figure 1 Schematic diagram of a secure iris privacy protection method based on K nearest neighbors of the present invention;

[0030] Figure 2 It is a principle diagram of the secure iris privacy protection method based on K nearest neighbors of the present invention. DETAILED DESCRIPTION

[0031] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described in detail with reference to the accompanying drawings.

[0032] like Figure 1-2 As shown, the K-nearest neighbor-based secure iris privacy protection method of the present invention includes the following steps:

[0033] S1. Generate a random binary classification dataset, the specific contents are as follows:

[0034] S101. Randomly generate a 32-bit binary string A using the Random function;

[0035] S102: Divide the binary string generated in step S101 into two blocks, each with 16 bits;

[0036] S103, for the two blocks generated in step S102, randomly invert x bits;

[0037] S104, repeat step S103 to generate five different 32-bit binary strings B;

[0038] S105. Based on step S104, the newly generated string is divided into two blocks again;

[0039] S106: After the block division in step S105, randomly invert the first block by y bits, and keep the second block unchanged;

[0040] S107, repeat step S106 to generate five different 32-bit binary strings C;

[0041] S108, storing the binary strings generated in steps S101, S104, and S107 into a txt file at once;

[0042] S109. Repeat the operation N times or more as the final random binary classification data set.

[0043] S2. Get iris privacy protection, the specific contents are as follows:

[0044] S201, performing block processing on the binary iris data x, dividing x into n blocks of binary strings of length m;

[0045] S202, calculate the distance between the iris data processed in step S201 and the first block of the txt text in the classification data set in sequence, using Hamming distance for distance calculation.

[0046] S203: For step S202, calculate the distance between the i-th block and the classification data set and sort them, select the k nearest binary classification strings and save them, neighbors = {neighbor1, neighbor2, ..., neighbor k};

[0047] S204: For the most recent k binary classification strings obtained in step S203, replace the first block with the second block, perform voting based on the second block, and save the last bit of each bit as the final result;

[0048] S205, after repeated operations of steps S202-S204, each block is operated in turn and then spliced;

[0049] S206: Save the final result obtained in step S205 as an iris template.

[0050] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are protected by the present invention.

Claims

1. A secure iris privacy protection method based on K nearest neighbor, characterized in that: The following steps are involved: S1, generate a random binary classification dataset; S2, generate iris privacy protection; The step S1 specifically includes the following contents: S101. Randomly generate a 32-bit binary string using the Random function; S102, dividing the binary string into two blocks of binary strings, each block having 16 bits; S103, randomly inverting the two binary strings by x bits; S104, repeatedly generating five different 32-bit binary strings; S105, dividing the newly generated string into two blocks again; S106, randomly invert the first block by y bits, and keep the second block unchanged; S107, repeatedly generating five different 32-bit binary strings; S108, save the binary strings generated by S101, S104, and S107 into a txt file at once; S109, repeat the operation N times or more as the final random binary classification data set; The step S2 specifically includes the following contents: S201, performing block processing on the binary iris data x; S202, calculating the distance between the iris data block i and the first block of the txt text in the classification data set, using Hamming distance for distance calculation; S203, calculate the distance between the i-th block and the classification data set, sort them, and select the k closest binary classification strings to save; S204, the k most recent binary classification strings are replaced by the second block, voting is performed using the second block, and the last bit of each bit is saved as the final result; S205, after repeated operations of steps S202-204, each block is operated in turn and then spliced; S206, saving the final result obtained in step S205 as an iris template.

Images