Verification method for providing information verification for enterprise information server

Abstract

The invention discloses a verification method for providing information verification for an enterprise information server. The verification method comprises the steps of S1, multi-mode biological feature collection and primary processing, S2, biological feature information encryption and segmentation marking, S3, distributed object storage and verification request generation, S4, storage node information retrieval and segmentation verification, and S5, multi-node verification result integration and final judgment. According to the method, the biological characteristic information is processed by adopting an AES high-strength encryption algorithm and a segmentation marking mechanism, and the high reliability and availability of the distributed object storage system are guaranteed by virtue of the multi-node distributed layout and redundancy strategy of the distributed object storage system; during verification, the widely distributed storage nodes quickly retrieve information according to an accurate request and carry out specialized segmented verification, so that data security protection is enhanced, local attacks and single-point faults are effectively resisted, and the verification efficiency and accuracy are improved at the same time.

Description

Technical Field

[0001] The present invention relates to the technical field of information verification, and in particular to a verification method for providing information verification for an enterprise information server. Background Art

[0002] With the popularization of the Internet and the development of related technologies, a large amount of information is transmitted and stored on the network, and the authenticity, integrity and security of information are facing severe challenges.

[0003] As the digital transformation of enterprises and organizations accelerates, more and more business processes and data assets have been digitized. Information verification technology has become a key support for ensuring the safe and reliable operation of digital businesses. For example, enterprise resource planning (ERP) systems and customer relationship management (CRM) systems all require strict verification of user identities and operating permissions to ensure the security of the company's core business data.

[0004] The traditional enterprise information server-based verification method has low security. In today's complex and changeable network security situation full of malicious attacks, whether the data is stored in a single server within the enterprise or a simpler distributed storage architecture is used, it is difficult to effectively resist complex attacks. Once it is attacked by complex attack methods such as advanced persistent threat attacks, distributed denial of service attacks, and malicious data tampering attacks, the integrity and security of the data will face extremely severe tests. Accordingly, this application proposes a verification method for providing information verification for enterprise information servers. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a verification method for providing information verification for an enterprise information server.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A verification method for providing information verification for an enterprise information server comprises the following steps:

[0008] S1. Multimodal biometric collection and preliminary processing. When a user needs to access the enterprise information server, the system immediately starts the multimodal biometric collection process. Through special biometric collection hardware or terminal devices with corresponding functions, the system simultaneously collects multiple biometric data of the user, including fingerprints, facial images, irises, and voiceprint information;

[0009] S2, biometric information encryption and segmentation marking, encrypt each pre-processed biometric information separately, use the AES encryption algorithm to generate an independent encryption key for each biometric, then segment the encrypted biometric information, divide the biometric into multiple data segments, and add unique marking information to each segment, which includes the segment number, biometric type, and association information with other segments, so that it can be accurately reassembled and identified later;

[0010] S3, distributed object storage and verification request generation, the segmented and marked biometric information is stored in the distributed object storage system, which is composed of multiple storage nodes, which are distributed in different network areas or geographical locations within the enterprise. Each node stores part of the biometric information segment, and according to the redundancy strategy of distributed storage, the reliability and availability of data are ensured. When the user completes the storage of the biometric information, a verification request is initiated to the enterprise information server. The request contains the user's identity, the biometric type used in this verification plan, and the corresponding segmentation tag information;

[0011] S4, storage node information retrieval and segment verification. After receiving the verification request, the enterprise information server searches for the corresponding storage node in the distributed object storage system according to the tag information in the request. Each selected storage node obtains the corresponding biometric information segment from the local storage according to the received tag information, decrypts it using the pre-stored decryption key, and verifies the decrypted segment using the local biometric verification algorithm.

[0012] S5. Integration of multi-node verification results and final judgment. Each storage node transmits the verification results back to the enterprise information server. The enterprise information server uses a specific fusion algorithm to comprehensively consider the verification results of multiple nodes and ultimately decide whether to allow them to access the enterprise information server.

[0013] Preferably, in step S1, the multimodal biometric feature uses a high-precision fingerprint recognition algorithm to extract detailed feature points such as ridge endpoints and bifurcation points from the collected fingerprint information, and converts them into digital feature vectors in a specific format;

[0014] Perform image preprocessing operations on facial images, including grayscale, noise reduction, face detection and positioning, and then use deep convolutional neural networks to extract key facial features, such as facial contours and facial texture;

[0015] For iris information, iris recognition technology is used to obtain and encode the iris texture features, pigment distribution features, etc.

[0016] The voiceprint information is extracted through speech signal processing technology to extract characteristic parameters such as frequency, amplitude, and harmonic structure.

[0017] Preferably, in step S2, a unified segmentation marking standard is formulated to clarify the coding rules of the segment number, the identification method of the biometric type to which it belongs, and the format and content of the associated information. The segment number is encoded in hexadecimal and the biometric type is represented by a specific two-letter code;

[0018] At the same time, the scalability of segmented tags is taken into consideration to adapt to possible future updates of biometric information or the addition of new biometric features. Specific tag bits or fields are reserved to indicate the version number of biometric information or the identification of new feature types, ensuring that the existing storage and verification mechanisms can be smoothly transitioned when the system is upgraded or expanded.

[0019] Preferably, in step S3, the distributed object storage is mainly composed of storage nodes, metadata servers, access nodes and other parts. The storage nodes are used for actual data storage, the metadata servers manage the metadata of the objects, such as object size, creation time and other information, and the access nodes are responsible for receiving client requests and coordinating the storage and reading of data between storage nodes.

[0020] Preferably, in step S4, a search algorithm based on a distributed hash table DHT is adopted, and the segmented tag information is used as a hash key to quickly locate the node storing the corresponding biometric information segment through hash calculation.

[0021] Preferably, in step S5, the specific fusion algorithm used is a fusion algorithm based on credibility weighting, which comprehensively considers the verification results of multiple nodes and assigns different weights according to the accuracy and stability of different biometric features in the overall verification;

[0022] If the final verification score is higher than the set threshold, the user is deemed to be a legitimate user and is allowed to access the enterprise information server;

[0023] On the contrary, if the verification score does not reach the threshold, the user is judged as an illegal user, access is denied, and the corresponding security alarm mechanism is activated to record relevant verification failure information, such as verification time, type of biometrics used, verification results of each node, etc., for subsequent security analysis and tracking.

[0024] The present invention has the following beneficial effects:

[0025] 1. The AES high-intensity encryption algorithm and segmentation marking mechanism are used to process biometric information. The distributed layout and redundancy strategy of the distributed object storage system's multiple nodes ensure its high reliability and availability. During verification, the widely distributed storage nodes quickly retrieve information based on precise requests and conduct professional segmentation verification, which not only strengthens data security protection, but also effectively resists local attacks and single point failures, while improving the efficiency and accuracy of verification.

[0026] 2. The storage nodes are widely distributed and have the advantage of distributed defense. They can ensure the stable and continuous operation of the verification system when facing local attacks or single point failures. Each node can carry out professional verification based on the stored information segment, giving full play to the professionalism and efficiency of specific biometric verification, and effectively improving the accuracy and speed of verification of a single biometric segment, thus laying a solid foundation for the reliability of the overall verification results.

[0027] 3. Verification and integration of results by multiple storage nodes can greatly improve verification accuracy. Different nodes conduct verification based on different hardware, network and data characteristics. Diversity can avoid misjudgment caused by algorithm deviation, data damage or environmental factors in a single node, effectively reduce false rejection rate and false acceptance rate, and effectively protect enterprise information security. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 The present invention is a flowchart of a verification method for providing information verification for an enterprise information server. DETAILED DESCRIPTION

[0029] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0030] A verification method for providing information verification for an enterprise information server comprises the following steps:

[0031] S1. Multimodal biometric collection and preliminary processing. When a user needs to access the enterprise information server, the system immediately starts the multimodal biometric collection process. Through special biometric collection hardware or terminal devices with corresponding functions, it simultaneously collects multiple biometric data of the user, including fingerprints, facial images, irises and voiceprint information. Multimodal biometrics uses high-precision fingerprint recognition algorithms to extract detailed feature points such as ridge endpoints and bifurcation points for the collected fingerprint information, and converts them into digital feature vectors in a specific format. It performs image preprocessing operations on facial images, including grayscale, noise reduction, face detection and positioning, etc., and then uses deep convolutional neural networks to extract key facial features, such as facial contours and facial texture. For iris information, iris recognition technology is used to obtain and encode iris texture features, pigment distribution features, etc., and voiceprint information uses speech signal processing technology to extract feature parameters such as frequency, amplitude, and harmonic structure.

[0032] S2. Biometric information encryption and segmentation marking. The pre-processed biometric information is encrypted separately, and the AES encryption algorithm is used to generate an independent encryption key for each biometric. Even if the encrypted information of a biometric is leaked, other biometric information can still be properly protected due to its independent key system.

[0033] The encrypted biometric information is then segmented, the biometrics are divided into multiple data segments, and unique marking information is attached to each segment. The marking information includes the segment number, the biometric type, and the association information with other segments, so that it can be accurately reassembled and identified later.

[0034] Establish a unified segmentation marking standard, clarify the encoding rules of segment numbers, the identification method of the biometric type to which they belong, and the format and content of the associated information. The segment numbers are encoded in hexadecimal. The use of hexadecimal can easily represent a larger range of segment number values, and can represent information more compactly during data storage and transmission.

[0035] The type of biometric feature is represented by a specific two-letter code, such as FP for fingerprint, FC for face, IR for iris, VP for voiceprint, etc. Through this standardized short code, the type of biometric feature can be identified quickly and accurately, which facilitates information classification and processing in complex verification systems.

[0036] At the same time, the scalability of segmented tags is taken into consideration to adapt to possible future updates of biometric information or the addition of new biometrics. Specific tag bits or fields are reserved to record the processed version of the current biometric information, which is used to indicate the version number of the biometric information or the identification of the new feature type. This ensures that when the system is upgraded or expanded, the system can quickly identify and assign the corresponding processing logic and storage structure to it, thereby achieving a smooth transition with the existing storage and verification mechanisms.

[0037] S3, distributed object storage and verification request generation, the segmented and marked biometric information is stored in the distributed object storage system. This system is composed of multiple storage nodes, each node stores part of the biometric information segment, and according to the redundancy strategy of distributed storage, even if an individual node fails, the complete biometric information can still be restored through the data of other nodes, thereby ensuring the reliability and availability of the data. When the user completes the storage of the biometric information, a verification request is initiated to the enterprise information server. The request contains the user's identity, the biometric type used in this verification plan, and the corresponding segmentation marking information.

[0038] Distributed object storage mainly consists of storage nodes, metadata servers, access nodes, etc. Storage nodes are used for actual data storage. These storage nodes are distributed in different network areas or geographical locations within the enterprise. This decentralized layout effectively reduces the risk of data loss due to local disasters or network failures.

[0039] The metadata server manages the metadata of the object, such as the object size, creation time, etc. These metadata are essential for quickly locating, retrieving, and managing biometric information segments;

[0040] As the interface between the system and the outside world, the access node is responsible for receiving client requests and coordinating the storage and reading of data between storage nodes.

[0041] S4. Storage node information retrieval and segment verification. After receiving the verification request, the enterprise information server searches for the corresponding storage node in the distributed object storage system based on the tag information in the request. Each selected storage node obtains the corresponding biometric information segment from the local storage based on the received tag information, decrypts it using the pre-stored decryption key, and verifies the decrypted segment using the local biometric verification algorithm.

[0042] A search algorithm based on the distributed hash table DHT is adopted, and the segmented tag information is used as the hash key. The node storing the corresponding biometric information segment is quickly located through hash calculation. There is no need to rely on the central index server. Each node can participate in the data positioning process according to the rules of the hash table, which not only improves the search efficiency, but also enhances the scalability and reliability of the system.

[0043] S5. Integration of multi-node verification results and final judgment. Each storage node transmits the verification results back to the enterprise information server. The enterprise information server uses a specific fusion algorithm to comprehensively consider the verification results of multiple nodes and ultimately decide whether to allow them to access the enterprise information server.

[0044] A specific fusion algorithm is used, which is a credibility-weighted fusion algorithm. The verification results of multiple nodes are comprehensively considered, and different weights are assigned according to the accuracy and stability of different biometric features in the overall verification. If the final verification score is higher than the set threshold, the user is judged to be a legitimate user and is allowed to access the enterprise information server. Otherwise, if the verification score does not reach the threshold, the user is judged to be an illegal user, access is denied, and the corresponding security alarm mechanism is activated, and relevant verification failure information is recorded.

[0045] Such as verification time accurate to seconds, type of biometric features used (such as fingerprint, face, iris, voiceprint), verification results of each node (pass or fail), etc., are used for subsequent security analysis and tracking. These data are used to conduct in-depth research on the patterns and rules of illegal access attempts, so as to optimize security strategies, update verification algorithms or strengthen network protection measures in a targeted manner.

[0046] Embodiment 1:

[0047] Step 1: Use a dedicated biometric feature collection terminal device that integrates a high-precision fingerprint recognition module, a high-definition camera, an iris scanner, and voiceprint information;

[0048] For fingerprint information, the extracted fingerprint feature vector contains information such as the coordinates, directions, and curvatures of multiple feature points. This information is stored in a specific data structure to facilitate subsequent processing and transmission;

[0049] After facial images are collected, the system automatically performs image preprocessing, graying the color images to reduce the amount of data while highlighting the facial texture features; then Gaussian filtering is used for noise reduction to remove noise interference in the image; then, through the face detection and positioning algorithm based on deep learning, the face area is accurately identified and its position and size are determined;

[0050] When collecting iris information, the device emits near-infrared light to illuminate the eyes and uses iris recognition technology to obtain the texture features of the iris, including detailed information such as wrinkles and spots on the iris and pigment distribution characteristics, and encodes these features;

[0051] During the voiceprint information collection process, a specific voice command is recorded into the microphone, and the voiceprint's characteristic parameters such as frequency, amplitude, harmonic structure, etc. are extracted through voice signal processing technology.

[0052] Step 2: The system generates a 128-bit (16-byte) random binary sequence as the encryption key of the fingerprint, which is recorded as FingerprintKey, the facial feature is recorded as FaceKey, the iris feature is recorded as IrisKey, and the voiceprint feature is recorded as VoiceKey. For example, the fingerprint feature data is FingerprintData, and it is converted into a format suitable for AES encryption, such as a binary sequence;

[0053] If the length of FingerprintData is not an integer multiple of 128 bits, use PKCS#7 padding to fill it so that its length becomes an integer multiple of 128 bits. (PKCS#7 padding rule) Assume that the length of the last data block (packet) is n bytes (n < 16), then fill 16-n bytes after this data block, and the value of each byte is 16-n. If the length of the last data block is 10 words (n = 10), it needs to fill 6 bytes, and the value of each byte is, so the length of the filled block reaches 16 bytes, which meets the requirements of the AES encryption packet length. Initial round key addition;

[0054] For the first FingerprintBlocks1 in FingerprintBlocks, perform a bitwise XOR operation with the subkey of the first round (obtained by expanding FingerprintKey) to obtain:

[0055] TempFingerprintBlock1=FingerprintBlock1^RoundKey1;

[0056] Byte replacement: Use each byte in TempFingerprintBlock1 as the S-box index of AES, and replace the original byte with the corresponding byte in the S-box to obtain:

[0057] SubstitutedFingerprintBlock1

[0058] Row shift: Perform row shift operation on SubstitutedFingerprintBlock1 to change the order of bytes and obtain ShiftedFingerprintBlock1;

[0059] Column confusion: Perform column confusion on ShiftedFingerprintBlock1 so that the bytes affect each other and get MixedFingerprintBlock1;

[0060] The round key is added to XOR the MixedFingerprintBlock1 with the subkey corresponding to the round to obtain a new intermediate result TempFingerprintBlock2;

[0061] The next round of encryption is performed on TempFingerprintBlock2 according to the above byte replacement, row shift, column confusion and round key addition operations. This is repeated for 9 rounds. The last round is byte replacement (SubBytes) and row shift (ShiftRows) operations:

[0062] Get FinalFingerprintBlock.

[0063] Perform the AddRoundKey operation to obtain the final encrypted fingerprint data block:

[0064] EncryptedFingerprintBlock=FinalFingerprintBlock^RoundKey10;

[0065] All blocks in EncryptedFingerprintBlocks are concatenated in the original order to obtain the final encrypted fingerprint information EncryptedFingerprintData;

[0066] FaceData, IrisData and VoiceData are also encrypted using AES to obtain: EncryptedFaceData, EncryptedIrisData and EncryptedVoiceData;

[0067] Taking the encrypted fingerprint information EncryptedFingerprintData as an example, it is divided into multiple data segments, which are recorded as Segment1, Segment2, and Segment3 respectively, and the segment is encoded in hexadecimal and set to 01, 02, and 03. This can conveniently represent a larger range of segment number values ​​and can more compactly represent information during data storage and transmission.

[0068] Step 3: When storing the encrypted and segmented data segments, a certain allocation strategy is used to disperse the data segments into multiple storage nodes. A hash calculation is performed on the tag information of each biometric information segment to obtain a hash value. The storage node where the data segment is stored is determined based on the modulo operation result of the hash value and the number of storage nodes. This ensures that the data is relatively evenly distributed on each storage node to avoid data skew.

[0069] The system will create multiple identical copies of each data segment on different storage nodes. For example, if the number of copies is set to 3, when a data segment is stored in a storage node, the system will also store the same data segment on the other two storage nodes.

[0070] Step 4: When the enterprise information server receives the verification request, it first parses the tag information, which contains key content such as the segment number, the type of biometric feature, and the association information with other segments. Then, based on a specific search algorithm, such as a search algorithm based on a distributed hash table (DHT), the enterprise information server uses the segment tag information as a hash key for hash calculation, and quickly locates the corresponding storage node in the distributed object storage system through this calculation result.

[0071] Due to the characteristics of the distributed hash table, each storage node can participate in the data location process according to its rules without relying on the central index server. This not only improves the search efficiency, but also enhances the scalability and reliability of the system. Once the storage node is determined, the selected storage node will receive instructions from the enterprise information server and accurately search for the corresponding biometric information segment in the local storage based on the received tag information;

[0072] After finding the corresponding biometric information segment, the storage node decrypts it using the pre-stored decryption key. Finally, the storage node verifies the decrypted segment using the local biometric verification algorithm.

[0073] Step 5: After each storage node completes the local verification of the corresponding biometric information segment, it will immediately send the verification result back to the enterprise information server for comprehensive consideration using a specific fusion algorithm. This fusion algorithm is usually a fusion algorithm based on credibility weighting, which will pre-assign different weights to different biometric features based on their accuracy and stability in the overall verification.

[0074] For example, iris recognition may be given a relatively high weight, such as 0.4, due to its extremely high accuracy and good stability; fingerprint recognition is also highly accurate but less stable, so its weight is set to 0.3; facial recognition is easily affected by the environment and has a weight of 0.2; voiceprint recognition is relatively weak, so its weight is set to 0.1;

[0075] When the verification process involves three storage nodes verifying different biometric features, node A passes the fingerprint verification result (its fingerprint matching value is 0.8 after conversion), node B passes the face verification result (the facial similarity score is 0.7 after conversion), and node C passes the iris verification result (the iris matching degree is extremely high, which is 0.95 after conversion). According to the above weight calculation, the final verification score is:

[0076] 0.3×0.8+0.2×0.7+0.4×0.95=0.82;

[0077] The final verification score is then compared with the set threshold. If the final verification score is higher than the set threshold (for example, the threshold is set to 0.75), the user is judged to be a legitimate user, and the enterprise information server allows the user to access the enterprise information server, and the user can smoothly carry out subsequent processes such as data query and business operations. On the contrary, if the verification score does not reach the threshold, such as in the above example, if the threshold is set to 0.85, the user will be judged as an illegal user, and the enterprise information server will deny access and activate the corresponding security alarm mechanism, while recording the relevant verification failure information in detail.

[0078] Embodiment 2:

[0079] When collecting biometrics, the semiconductor fingerprint sensor collects fingerprints and extracts feature point information into a list after processing. The high-definition network camera takes facial images and processes them to obtain feature vectors. Professional iris collection equipment combines multi-spectral imaging to process iris images and extracts feature codes. High-sensitivity microphones record voice samples to extract features and combine them into matrices. In the encryption stage, a 256-bit key is obtained by hashing the specific biometric fragment with SHA-256. After combining serialized data, it is encrypted with AES-256 and a confusion factor is introduced. Storage uses a combination of local cache (first-in-first-out, keeping the last three successful data) and private cloud storage (stored by ID classification and redundant in two locations) and ensures synchronization. Verification first reads encrypted data from the local cache. If there is no data, it is downloaded from the cloud and decrypted by generating a key based on the biometric fragment. Fingerprints pass when the score calculated by the elastic matching algorithm exceeds 80 points, faces pass when the similarity probability exceeds 0.7 when the support vector machine classifier is used for comparison, irises pass when the Hamming distance algorithm is used for comparison less than 0.3, and voiceprints pass when the matching cost function value of the dynamic time warping algorithm is less than the threshold. The judgment is based on weighted voting, with weights of 0.4 for iris, 0.3 for fingerprint, 0.2 for face, and 0.1 for voiceprint. If the weighted sum exceeds 0.6, it is legal, otherwise it is illegal. If it is illegal, a local prompt will be given and detailed information will be uploaded to the security management server to analyze potential threats in order to adjust the policy optimization algorithm.

[0080] Embodiment three:

[0081] Basic biometric feature collection is performed, including the use of capacitive fingerprint sensors to collect fingerprints and save them as bitmaps, webcams to collect facial images and save them as PNGs, near-infrared devices to collect iris images for simple encoding, and microphones to collect voiceprints for basic feature extraction and save them as vectors; a 128-bit key is then generated by concatenating the system timestamp and user ID with a hash, and the feature data is concatenated and stored in the local computer user name folder after multiple byte-by-byte XOR encryption; the encrypted data is read during verification, and the key is generated and decrypted using the same method, and simple methods such as pixel ratio, feature point distance, texture features, and frequency amplitude comparison are used to verify fingerprints, faces, irises, and voiceprints; finally, a simple voting mechanism is used to determine if more than three pass the test and the test is legal; otherwise, it is illegal and a log containing time, results, and basic data information is recorded. The lack of in-depth records is not conducive to analysis and investigation.

[0082] It should be noted that in the comparative examples, in the process of identity verification, in the collection of biometric features, the comparative example 1 uses a simple device that uses an ordinary fingerprint recognition module and a low-resolution camera without an iris scanning device and a voiceprint collection function to collect biometric features, and the comparative example 2 uses a device that integrates multiple modules, but the accuracy of each module is low. The examples 1, 2, and 3 are compared, as shown in Table 1:

[0083] Table 1: Biological collection feature accuracy table

[0084]

[0085] As shown in Table 1, Example 1 shows a high accuracy rate in all aspects of biometric feature collection, indicating that the technology and equipment it adopts are relatively advanced and reliable. Example 2 has a good performance in most feature collection, but is slightly inferior to Example 1 in fingerprint and facial feature collection. The accuracy rate of biometric feature collection in Example 3 is generally low, and the technology may need further improvement. Comparative Examples 1 and 2 perform poorly in various aspects of biometric feature collection, especially Comparative Example 2, which is at the lowest level in almost all biometric feature collection, and may need to re-evaluate and improve its biometric feature collection technology and methods.

[0086] It can be seen that this method can accurately collect the customer's biometrics, thereby increasing the success rate of verification.

[0087] It should be noted that, in the comparative embodiment, in the process of verifying the identity, in the case of processing the verification request, the storage point verification request time and the response comparison of the comparative example 1 and the comparative example 2 are shown in Table 2:

[0088] Table 2: Comparison of verification request time and success rate

[0089]

[0090] As shown in Table 2, in terms of the time for searching storage nodes and the comparison of relevant data for verification request processing, Example 1 shows significant advantages. In terms of the time for searching storage nodes, Example 1 only takes 1 second, which is the same as 1 second in Example 2, but compared with 10 seconds in Example 3 and 10 seconds in Comparative Example 1 and 20 seconds in Comparative Example 2, it has obvious speed advantages, which means that Example 1 can achieve fast response in data storage and retrieval, avoid long waiting times for users, and ensure efficient operation of the system.

[0091] Secondly, in terms of decryption success rate, Example 1 reaches 98%, which is the same as 98% of Example 2, but much higher than 70% of Example 3, 70% of Comparative Example 1, and 50% of Comparative Example 2. The high decryption success rate shows that the encryption algorithm and key management mechanism adopted in Example 1 are more reliable, can provide strong protection for the security of data during transmission and storage, and greatly reduce the risk of data leakage.

[0092] Finally, in terms of feature matching accuracy (taking fingerprints as an example), Example 1 is as high as 95%, which is higher than 80% of Example 2, 70% of Example 3, 60% of Comparative Example 1, and 30% of Comparative Example 2. This reflects the high accuracy of Example 1 in the biometric feature recognition algorithm, which can more accurately perform fingerprint matching, thereby effectively preventing illegal access in high security demand scenarios and ensuring system security.

[0093] In summary, Example 1 performs excellently in multiple key indicators of verification request processing and has considerable advantages over other examples and comparison examples.

[0094] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A verification method for providing information verification for an enterprise information server, characterized in that: The following steps are involved: S1. Multimodal biometric collection and preliminary processing. When a user needs to access the enterprise information server, the system immediately starts the multimodal biometric collection process. Through special biometric collection hardware or terminal devices with corresponding functions, the system simultaneously collects multiple biometric data of the user, including fingerprints, facial images, irises, and voiceprint information; S2, biometric information encryption and segmentation marking, encrypt each pre-processed biometric information separately, use the AES encryption algorithm to generate an independent encryption key for each biometric, then segment the encrypted biometric information, divide the biometric into multiple data segments, and add unique marking information to each segment, which includes the segment number, biometric type, and association information with other segments, so that it can be accurately reassembled and identified later; S3, distributed object storage and verification request generation, the segmented and marked biometric information is stored in the distributed object storage system, which is composed of multiple storage nodes, which are distributed in different network areas or geographical locations within the enterprise. Each node stores part of the biometric information segment, and according to the redundancy strategy of distributed storage, the reliability and availability of data are ensured. When the user completes the storage of the biometric information, a verification request is initiated to the enterprise information server. The request contains the user's identity, the biometric type used in this verification plan, and the corresponding segmentation tag information; S4, storage node information retrieval and segment verification. After receiving the verification request, the enterprise information server searches for the corresponding storage node in the distributed object storage system according to the tag information in the request. Each selected storage node obtains the corresponding biometric information segment from the local storage according to the received tag information, decrypts it using the pre-stored decryption key, and verifies the decrypted segment using the local biometric verification algorithm. S5. Integration of multi-node verification results and final judgment. Each storage node transmits the verification results back to the enterprise information server. The enterprise information server uses a specific fusion algorithm to comprehensively consider the verification results of multiple nodes and ultimately decide whether to allow them to access the enterprise information server.

2. A verification method for providing information verification for an enterprise information server according to claim 1, characterized in that: In step S1, the multimodal biometric feature uses a high-precision fingerprint recognition algorithm to extract detailed feature points such as ridge endpoints and bifurcation points from the collected fingerprint information, and converts them into digital feature vectors in a specific format; Perform image preprocessing operations on facial images, including grayscale, noise reduction, face detection and positioning, and then use deep convolutional neural networks to extract key facial features, such as facial contours and facial texture; For iris information, iris recognition technology is used to obtain and encode the iris texture features, pigment distribution features, etc. The voiceprint information is extracted through speech signal processing technology to extract characteristic parameters such as frequency, amplitude, and harmonic structure.

3. A verification method for providing information verification for an enterprise information server according to claim 1, characterized in that: In step S2, a unified segmentation marking standard is formulated to clarify the coding rules of the segment number, the identification method of the biometric type to which it belongs, and the format and content of the associated information. The segment number is encoded in hexadecimal and the biometric type is represented by a specific two-letter code; At the same time, the scalability of segmented tags is taken into consideration to adapt to possible future updates of biometric information or the addition of new biometric features. Specific tag bits or fields are reserved to indicate the version number of biometric information or the identification of new feature types, ensuring that the existing storage and verification mechanisms can be smoothly transitioned when the system is upgraded or expanded.

4. A verification method for providing information verification for an enterprise information server according to claim 1, characterized in that: In step S3, the distributed object storage is mainly composed of storage nodes, metadata servers, access nodes and other parts. The storage nodes are used for actual data storage, the metadata servers manage the metadata of the objects, such as object size, creation time and other information, and the access nodes are responsible for receiving client requests and coordinating the storage and reading of data between storage nodes.

5. A verification method for providing information verification for an enterprise information server according to claim 1, characterized in that: In the step S4, a search algorithm based on a distributed hash table DHT is adopted, and the segmented tag information is used as a hash key to quickly locate the node storing the corresponding biometric information segment through hash calculation.

6. A verification method for providing information verification for an enterprise information server according to claim 1, characterized in that: In step S5, a specific fusion algorithm is used, which is a fusion algorithm based on credibility weighting, to comprehensively consider the verification results of multiple nodes, and assign different weights according to the accuracy and stability of different biometric features in the overall verification; If the final verification score is higher than the set threshold, the user is deemed to be a legitimate user and is allowed to access the enterprise information server; On the contrary, if the verification score does not reach the threshold, the user is judged as an illegal user, access is denied, and the corresponding security alarm mechanism is activated to record relevant verification failure information, such as verification time, type of biometrics used, verification results of each node, etc., for subsequent security analysis and tracking.

Images