An Automatic Unlocking Verification Method, Device and Storage Medium for a Trunk

Abstract

The present application provides a method, device and storage medium for automatically unlocking and verifying a cash box, which can effectively improve the security of the cash box. The method of the present application includes: obtaining a cash box unlocking request of a user, where the cash box unlocking request includes user biometric features, and the user biometric features include fingerprint, face and iris information; determining the position of the cash box; collecting current environmental parameters of the cash box according to the position of the cash box, and using an artificial large model to determine an adjustment coefficient according to the environmental parameters; adjusting the verification weight of the biometric features according to the adjustment coefficient to obtain a target verification weight; verifying the user biometric features according to the target verification weight, and after determining that the biometric feature verification is passed, obtaining an unlocking instruction for the cash box; sending the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction; when it is determined that the verification is passed, using the unlocking instruction to perform an unlocking process on the cash box and collecting unlocking information during the unlocking process.

Description

Technical Field

[0001] This application relates to the technical field of intelligent locks, and particularly to an automatic unlocking and verification method, device, and storage medium for a money box. Background Art

[0002] With the rapid development of the financial industry and the continuous improvement of security requirements, as a safe storage carrier for cash, valuables, important documents, etc., the security of money boxes has always been the focus of attention of financial institutions. At present, the traditional unlocking methods of money boxes mostly rely on physical keys or simple password verification. These methods have security risks such as being easy to lose, copy, and crack, and obviously cannot meet the high standards of modern financial security requirements.

[0003] Therefore, in order to enhance the security of money boxes, biometric recognition technology has been introduced into intelligent money boxes. By extracting the biometric features of operators, such as fingerprints, faces, irises, etc., due to the uniqueness, non-replicability, and stability of biometric features, when operating the money box, it is possible to determine whether the operator is a legitimate person based on the biometric features of the current operator, thereby effectively improving the security of unlocking the money box.

[0004] Although the adoption of biometric recognition can increase the security of money boxes, some problems are also faced during the process of using biometric recognition. For example, under different environmental conditions, the light intensity may be different, which may affect the recognition accuracy of biometric features. Moreover, the current single biometric verification method has the risk of being forged or attacked, seriously threatening the security of money boxes and thus unable to meet the security requirements of the financial industry. Summary of the Invention

[0005] This application provides an automatic unlocking and verification method, device, and storage medium for a money box to solve the above technical problems.

[0006] In the first aspect of this application, an automatic unlocking and verification method for a money box is provided. The method includes:

[0007] Obtain a money box unlocking request from a user, where the money box unlocking request includes user biometric features, and the user biometric features include fingerprint, face, and iris information;

[0008] Determine the position of the money box;

[0009] Collect the current environmental parameters of the money box according to the position of the money box, and use an artificial large model to determine an adjustment coefficient based on the environmental parameters;

[0010] Adjust the verification weight of the biometric features according to the adjustment coefficient to obtain a target verification weight;

[0011] Verify the user's biometric features according to the target verification weight. After determining that the biometric feature verification is passed, obtain the unlocking instruction for the cash box.

[0012] Send the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction.

[0013] When it is determined that the verification is passed, use the unlocking instruction to unlock the cash box and collect the unlocking information during the unlocking process.

[0014] Encrypt the unlocking information to obtain encrypted unlocking information.

[0015] Upload the encrypted unlocking information to the blockchain system for storage.

[0016] Optionally, collect the current environmental parameters of the cash box according to the location of the cash box, and use the artificial large model to determine the adjustment coefficient according to the environmental parameters, including:

[0017] Collect the light intensity and network intensity of the environment where the cash box is located according to the location of the cash box.

[0018] Use the artificial large model to determine the adjustment coefficient according to the light intensity and network intensity of the environment where the cash box is located.

[0019] Optionally, use the artificial large model to determine the adjustment coefficient according to the light intensity and network intensity of the environment where the cash box is located, including:

[0020] Obtain the historical accuracy rate.

[0021] Input the light intensity, the network intensity, and the location of the cash box into the artificial large model to calculate the environmental risk score.

[0022] Determine the adjustment coefficient according to the environmental risk score and the historical accuracy rate.

[0023] The formula for the adjustment coefficient is:

[0024] W = α * environmental risk score + β * historical accuracy rate, where α and β are the dynamic coefficients output by the artificial large model, and W is the adjustment coefficient for biometric feature verification.

[0025] Optionally, adjust the verification weight of the biometric features according to the adjustment coefficient to obtain the target verification weight, including:

[0026] Adjust the verification weights of the user's fingerprint, face, and iris according to the adjustment coefficient.

[0027] Calculate the verification weights of the user's fingerprint, face, and iris to obtain the target value.

[0028] Determine whether the target value meets a preset value;

[0029] If so, output the target verification weight, where the target verification weight includes the verification weights of the user's fingerprint, face, and iris.

[0030] Optionally, verify the user's biometric features according to the target verification weight. After determining that the biometric feature verification is passed, obtain the unlocking instruction for the money box, including:

[0031] Perform multimodal fusion comparison on the user's biometric features to generate an original similarity score;

[0032] Calculate a target threshold according to the target verification weight;

[0033] The formula for calculating the target threshold: S = S0 × T, where S is the target threshold, S0 is the original similarity score, and T is the target verification weight;

[0034] Determine a preset threshold;

[0035] When the target threshold is greater than the preset threshold, determine that the biometric feature verification is passed, and obtain the unlocking instruction for the money box.

[0036] Optionally, after determining that the verification is passed, use the unlocking instruction to perform an unlocking process on the money box, and collect the unlocking information during the unlocking process, including:

[0037] After determining that the verification is passed, send the unlocking instruction to the money box so that the money box performs an unlocking process according to the unlocking instruction;

[0038] After the money box is opened, collect the operator ID, timestamp, and geographical location information during the unlocking process.

[0039] Optionally, after sending the unlocking instruction to the money box so that the money box verifies the unlocking instruction, the method further includes:

[0040] When it is determined that the verification fails, send an alarm message and collect the information indicating that the verification fails;

[0041] Encrypt the information indicating that the verification fails and upload it to the blockchain system for storage.

[0042] The second aspect of the present application provides a money box automatic unlocking verification device, and the device includes:

[0043] A first acquisition unit, configured to acquire a money box unlocking request of a user, where the money box unlocking request includes user biometric features, and the user biometric features include fingerprint, face, and iris information;

[0044] A first determination unit for determining the position of the cash box;

[0045] A second determination unit for collecting the current environmental parameters of the cash box according to the position of the cash box, and determining an adjustment coefficient according to the environmental parameters by using an artificial large model;

[0046] An adjustment unit for adjusting the verification weight of the biometric feature according to the adjustment coefficient to obtain a target verification weight;

[0047] A second acquisition unit for verifying the user biometric feature according to the target verification weight, and after determining that the biometric feature verification is passed, acquiring an unlocking instruction for the cash box;

[0048] A verification unit for sending the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction;

[0049] A collection unit for, when determining that the verification is passed, performing an unlocking process on the cash box by using the unlocking instruction and collecting unlocking information during the unlocking process;

[0050] A third acquisition unit for encrypting the unlocking information to obtain encrypted unlocking information;

[0051] An upload unit for uploading the encrypted unlocking information to a blockchain system for storage.

[0052] The third aspect of the present application provides a cash box automatic unlocking verification device, and the device includes:

[0053] A processor, a storage, an input / output unit, and a bus;

[0054] The processor is connected to the storage, the input / output unit, and the bus;

[0055] The storage stores a program, and the processor calls the program to execute the method of the first aspect and any optional method in the first aspect.

[0056] The fourth aspect of the present application provides a computer-readable storage medium, and a program is stored on the computer-readable storage medium, and when the program is executed on a computer, it executes the method of the first aspect and any optional method in the first aspect.

[0057] It can be seen from the above technical solutions that the present application has the following advantages:

[0058] 1. The present application combines user biometric features such as fingerprint, face, and iris information for verification, significantly improving the accuracy and security of unlocking verification, and effectively preventing the risks of forgery or misrecognition that may exist in a single biometric feature.

[0059] 2. The present application adjusts the verification weight of biometrics according to the environmental parameters of the location where the cash box is located, enabling the verification process to adapt to different environmental conditions, further enhancing the robustness and security of the verification. Moreover, during the unlocking process, biometric verification is not only performed on the user side, but also the unlocking instruction is verified on the cash box side, forming a dual security protection to ensure the safety and reliability of the unlocking operation.

[0060] 3. By collecting the unlocking information during the unlocking process and encrypting the unlocking information, the security of the data during transmission and storage is ensured, preventing data leakage and illegal access. After encrypting the unlocking information, it is uploaded to the blockchain system for storage, providing strong data support for subsequent auditing, traceability, and management.

[0061] 4. The present application realizes an automated processing flow from user requests to cash box unlocking, reducing manual intervention and improving the unlocking efficiency.

[0062] 5. Through the unlocking information stored in the blockchain system, managers can remotely monitor the unlocking status of the cash box, facilitating centralized management and exception handling. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0064] Figure 1 It is a schematic flowchart of an embodiment of the cash box automatic unlocking verification method of the present application;

[0065] Figure 2 It is another schematic diagram of an embodiment of the cash box automatic unlocking verification method of the present application;

[0066] Figure 3 It is another schematic diagram of an embodiment of the cash box automatic unlocking verification method of the present application;

[0067] Figure 4 It is another schematic diagram of an embodiment of the cash box automatic unlocking verification method of the present application;

[0068] Figure 5 It is another schematic diagram of an embodiment of the cash box automatic unlocking verification method of the present application;

[0069] Figure 6 It is a schematic diagram of an embodiment of the cash box automatic unlocking verification device of the present application;

[0070] Figure 7Schematic diagram of another embodiment of the cash box automatic unlocking verification device of the present application. Detailed implementation manners

[0071] The present application provides a method, device and storage medium for automatically unlocking and verifying a cash box, improving the security of cash box use, effectively solving many problems existing in traditional cash box unlocking methods, and providing a more secure, reliable and convenient cash box management solution for financial institutions.

[0072] Please refer to Figure 1 , an embodiment of a method for automatically unlocking and verifying a cash box provided by the first aspect of the present application, the embodiment includes:

[0073] 101. Obtain a cash box unlocking request of a user, where the cash box unlocking request includes user biometric features, and the user biometric features include fingerprint, face and iris information;

[0074] 102. Determine the position of the cash box;

[0075] 103. Collect the current environmental parameters of the cash box according to the position of the cash box, and use an artificial large model to determine an adjustment coefficient according to the environmental parameters;

[0076] 104. Adjust the verification weight of the biometric feature according to the adjustment coefficient to obtain a target verification weight;

[0077] 105. Verify the user biometric feature according to the target verification weight. After determining that the biometric feature verification is passed, obtain the unlocking instruction of the cash box;

[0078] 106. Send the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction;

[0079] 107. When it is determined that the verification is passed, use the unlocking instruction to unlock the cash box and collect the unlocking information during the unlocking process;

[0080] 108. Encrypt the unlocking information to obtain encrypted unlocking information;

[0081] 109. Upload the encrypted unlocking information to a blockchain system for storage.

[0082] In an embodiment of the present application, first, a cash box unlocking request of the user is obtained, and the cash box unlocking request includes user biometrics, and the user biometrics include fingerprint, face and iris information; then the location of the cash box is determined, and then the current environmental parameters of the cash box are collected according to the location of the cash box, and an adjustment coefficient is determined according to the environmental parameters using an artificial large model, and the verification weight of the biometric is adjusted according to the adjustment coefficient to obtain a target verification weight; then the user biometric is verified according to the target verification weight, and after determining that the biometric verification has passed, the unlocking instruction of the cash box is obtained, and the unlocking instruction is sent to the cash box, so that the cash box verifies the unlocking instruction; when it is determined that the verification has passed, the unlocking instruction is used to unlock the cash box, and the unlocking information in the unlocking process is collected, and after the unlocking information is encrypted to obtain the unlocking encrypted information, the unlocking encrypted information is finally uploaded to the blockchain system for storage.

[0083] In step 101, a cash box unlocking request of a user is obtained. The cash box unlocking request includes the user's biometric features, which include fingerprint, face and iris information. Specifically, when the user needs to open the cash box, he will interact with the corresponding operation terminal device. This terminal device can be a fixed device installed at the cash box storage location, or it can be a specific application on the user's own mobile device, which is not specifically limited here. When the user initiates the cash box unlocking operation on the terminal device, the cash box unlocking request is initiated.

[0084] When the user chooses to unlock the door, the user places his finger on the corresponding fingerprint recognition sensor. The sensor detects the fingerprint's lines, ridges, and valleys, and converts them into digital signals to form fingerprint image data. The camera on the terminal device starts working and takes a picture of the user's face. The captured facial image will be transmitted to the device's image processing module, which uses algorithms to detect and locate key facial feature points, such as the position and shape of the eyes, nose, mouth, eyebrows, etc., and extracts facial feature information. The iris recognition device on the terminal device also starts working. The device contains a near-infrared light source and a high-resolution camera. The light emitted by the near-infrared light source illuminates the user's eyes, making the iris present a clear texture pattern, and the camera captures the iris image. Then, the device pre-processes the iris image, including denoising, normalization, and other operations, and then uses a specific algorithm to extract the unique texture features of the iris to form iris information.

[0085] The unlocking request initiated by the user and the collected fingerprint, face and iris biometric information are then packaged and sent to the server via a local area network, wide area network or mobile network.

[0086] In step 102, the position of the cash box is determined. It should be noted that when the cash box is manufactured or configured, a GPS positioning device is installed to ensure accurate position determination. So that when the server communicates with the GPS positioning device on the cash box, it can obtain the position information sent by the GPS positioning device. For GPS or Beidou positioning modules, the server will receive longitude and latitude coordinate data, and then the server will parse and process the received longitude and latitude coordinate data to determine the specific geographical location where the cash box is currently located. After determining the position information of the cash box, step 103 is executed.

[0087] In step 103, the current environmental parameters of the cash box are collected according to its position, and an adjustment coefficient is determined by using an artificial large model based on the environmental parameters. Specifically, a variety of environmental sensors are installed around the cash box or inside the cash box itself. The temperature sensor is used to measure the environmental temperature and can sense the temperature change of the surrounding environment in real time; the light sensor uses a photoresistor or a photodiode to detect the environmental light intensity; the pressure sensor can measure the air pressure value of the current environment. These sensors transmit the collected environmental parameter data to the data acquisition module in the form of electrical signals or digital signals.

[0088] The data acquisition module processes the signals transmitted by each sensor, converts them into a unified digital format, and sends the environmental parameter data to the server side of the system through a wireless network. The server side integrates and stores the received environmental parameter data to form a set of parameters of the environment where the current cash box is located.

[0089] It should be noted that the artificial large model is a deep neural network model, and this deep neural network model is trained with a large amount of environmental parameter data and corresponding biometric verification situation data. The training data includes indicators such as the accuracy rate and false recognition rate of fingerprint, face, and iris biometric recognition under different combinations of environmental parameters. After training, the deep neural network model can learn the relationship between environmental parameters and biometric verification.

[0090] After the deep neural network model is trained, the server inputs the integrated environmental parameter data into the trained deep neural network model. The deep neural network model outputs an adjustment coefficient according to the input environmental parameters through internal algorithms and weight calculations. This adjustment coefficient reflects the influence degree of the current environment on biometric verification. For example, in a high-temperature and high-humidity environment, it may affect the accuracy of fingerprint recognition, and the deep neural network model will output a corresponding adjustment coefficient to reduce the verification weight of fingerprint biometrics.

[0091] In step 104, the verification weights of biometric features are adjusted according to the adjustment coefficient to obtain the target verification weights. Specifically, in the initial settings, initial verification weights have been pre-set for the three biometric features of fingerprint, face, and iris respectively. These initial weights are determined based on the general reliability of biometric features and application experience in different scenarios. For example, the initial weight of fingerprint may be set to 0.4, the initial weight of face to 0.3, and the initial weight of iris to 0.3. The sum of these weights is 1, representing the total proportion of the three biometric features in the comprehensive verification.

[0092] After the system receives the adjustment coefficient output by the deep neural network model in step 103, it adjusts the initial weights of each biometric feature according to a specific weight adjustment algorithm. The adjustment is carried out through a multiplication operation algorithm, that is, the target verification weight of each biometric feature is equal to its initial weight multiplied by the adjustment coefficient. For example, if the adjustment coefficient is 0.8, then the target verification weight of fingerprint becomes 0.4 * 0.8 = 0.32, the target verification weight of face becomes 0.3 * 0.8 = 0.24, and the target verification weight of iris becomes 0.3 * 0.8 = 0.24. Then, the system normalizes the adjusted weights to ensure that the sum of the target verification weights of the three biometric features is still 1, so as to ensure the rationality and accuracy of the verification.

[0093] In step 105, the user's biometric features are verified according to the target verification weights. After determining that the biometric feature verification is passed, the unlocking instruction for the cash box is obtained. Specifically, when verifying fingerprint information, the system compares the fingerprint image data collected by the user with the fingerprint template pre-stored in the database. The comparison algorithm calculates the similarity score between the fingerprint image and the template by calculating the matching degree of the positions and directions of minutiae points in the fingerprint image, such as the starting point, ending point, and bifurcation point of the fingerprint pattern, with the minutiae points in the template, and obtains a similarity score.

[0094] When verifying face features, the extracted user face feature information is matched with the face template stored in the database. The face recognition algorithm analyzes the positions, shapes, and relative relationships of face feature points, calculates feature values such as the distances and angles between face features, and compares them with the feature values in the template to obtain a face matching score.

[0095] When performing iris verification, a special iris recognition algorithm is used to compare the collected iris feature information with the stored iris template. The algorithm analyzes features such as the texture details and color distribution of the iris, calculates the similarity between iris features, and thus obtains an iris similarity score.

[0096] The system calculates the weighted verification scores of fingerprints, faces, and irises based on the obtained target verification weights. That is, multiply the verification score of each biometric by its corresponding target verification weight, and then add the three weighted scores to obtain a comprehensive verification score. For example, if the fingerprint verification score is 80 points, the face verification score is 70 points, the iris verification score is 85 points, and the target verification weights are 0.32, 0.24, and 0.24 respectively, then the comprehensive verification score is 80 * 0.32 + 70 * 0.24 + 85 * 0.24 = 25.6 + 16.8 + 20.4 = 62.8 points.

[0097] After that, the system compares the calculated comprehensive verification score with a pre-set verification threshold. If the comprehensive verification score reaches or exceeds the verification threshold (such as 60 points), it is determined that the biometric verification is passed. At this time, the system obtains the unlocking instruction of the cash box from the instruction source, and the unlocking instruction is used to control the unlocking mechanism of the cash box to open.

[0098] In step 106, the unlocking instruction is sent to the cash box so that the cash box verifies the unlocking instruction. Specifically, the server side of the system sends the obtained unlocking instruction to the cash box through wireless communication. During the sending process, in order to ensure the security and accuracy of the instruction, the instruction will be encrypted and a check code will be added.

[0099] Among them, the cash box is equipped with a corresponding wireless communication module for receiving the unlocking instruction sent by the server. When the cash box receives the unlocking instruction, it first decrypts the instruction, then extracts the check code in the instruction, and verifies the instruction content according to the pre-set verification algorithm. The verification content includes the integrity, legality, and validity of the instruction. When the verification passes, step 107 is executed. If the verification fails, an alarm message is sent and the information of the failed verification is collected, and the information of the failed verification is encrypted and uploaded to the blockchain system for storage for later viewing.

[0100] In step 107, when it is determined that the verification is passed, the cash box is unlocked using the unlocking instruction, and the unlocking information during the unlocking process is collected. Specifically, if the verification result of the unlocking instruction by the cash box is passed, the unlocking mechanism of the cash box executes the unlocking operation according to the unlocking instruction.

[0101] And during the unlocking process, the unlocking information during the unlocking process is collected through sensors, and these sensors transmit the collected unlocking information to the data recording module of the cash box in the form of digital signals. The data recording module integrates and stores these information, and sends the unlocking information to the server side of the system through the wireless communication network for subsequent processing and storage.

[0102] Specifically, after the money box is opened, unlocking information during the unlocking process is collected. The unlocking information includes the operator ID, timestamp, geographical location information, etc., which are not specifically limited herein.

[0103] In step 108, the unlocking information is encrypted to obtain encrypted unlocking information. Specifically, the system selects a symmetric encryption algorithm according to the sensitivity and security requirements of the unlocking information. At this time, the system generates an encryption key, and then uses this key to encrypt the unlocking information, converting the plaintext unlocking information into ciphertext form. It should be noted that during the encryption process, further processing can also be performed on the encrypted ciphertext, such as adding a digital signature, etc., to enhance the security and traceability of the information. The finally obtained ciphertext is the encrypted unlocking information.

[0104] In step 109, the encrypted unlocking information is uploaded to the blockchain system for storage. Specifically, the system encapsulates the encrypted unlocking information into a transaction record that conforms to the data format of the blockchain system. This transaction record contains information such as the content of the encrypted unlocking information, timestamp, sender, etc. Then, the system sends the transaction record to the nodes in the blockchain network.

[0105] The nodes in the blockchain network will verify the transaction record, including verifying whether the format of the transaction record is correct, whether the signature is valid, etc. After being verified and passed by the consensus algorithm of the nodes, the transaction record is packaged into a new block and added to the blockchain. When this block is added to the blockchain, the encrypted unlocking information is permanently stored, and due to the immutable characteristic of the blockchain, the security and traceability of the unlocking information are ensured. Subsequently, through the query interface of the blockchain system, specific query conditions can be used to query and verify this encrypted unlocking information.

[0106] This application adjusts the verification weight of biometric features according to the environmental parameters of the location of the money box, enabling the verification process to adapt to different environmental conditions, further enhancing the robustness and security of the verification. Moreover, during the unlocking process, biometric verification is not only performed on the user side, but also the unlocking instruction is verified on the money box side, forming a dual security protection to ensure the security and reliability of the unlocking operation, thereby meeting the security requirements of the financial industry.

[0107] Please refer to Figure 2 , according to some embodiments of the present invention, in step 103, collecting the current environmental parameters of the money box according to the location of the money box, and using an artificial large model to determine the adjustment coefficient according to the environmental parameters may specifically include, but is not limited to, the following:

[0108] 201. Collect the light intensity and network intensity of the environment where the money box is located according to the location of the money box;

[0109] 202. Use an artificial large model to determine an adjustment coefficient based on the light intensity and network strength of the environment where the cash box is located.

[0110] In the embodiment of the present application, the light intensity and network strength of the environment where the cash box is located are collected according to the position of the cash box. Specifically, the position of the cash box has been determined in the aforementioned step 102. Based on this accurate position information, the environmental parameters at this position can be collected specifically.

[0111] Among them, a photoresistor type and a photodiode type are installed on the cash box itself or its surrounding environment. The resistance value of the photoresistor will change with the change of light intensity. The stronger the light, the lower the resistance value; the photodiode can convert the optical signal into an electrical signal, and the intensity of the electrical signal output is related to the light intensity. These sensors are reasonably arranged at positions where the light conditions of the environment where the cash box is located can be accurately sensed, such as installed on the outer surface of the cash box or on the fixed brackets around the cash box, and no specific limitation is made here.

[0112] In practical applications, the light intensity sensor monitors the light conditions in the environment in real time and outputs the detected light intensity information in the form of electrical signals. These electrical signals are then transmitted to the data acquisition module. The data acquisition module processes the signals and converts them into light intensity data in digital format. Then, these digital data are sent to the server side of the system through a wired or wireless network for storage and subsequent processing.

[0113] When collecting the network strength, since a network connection module is equipped in the cash box, this network connection module can detect the network signal strength of the current environment in real time. For the Wi-Fi network, the Wi-Fi module will scan the surrounding Wi-Fi hotspots and obtain the signal strength information of each hotspot, usually expressed in decibels milliwatt. For the mobile network, the communication module will detect the strength of the base station signal and convert it into a corresponding numerical representation.

[0114] The network connection module sorts and packages the detected network strength information, and then sends these data to the server side of the system through its own communication function. The server side receives and stores these network strength data for subsequent analysis together with the light intensity data.

[0115] The deep neural network model has been trained with a large amount of data. These training data include the performance data of the biometric recognition system under different combinations of light intensity and network strength, such as the accuracy rate, false recognition rate, recognition time, etc. of fingerprint, face, and iris recognition, as well as the associated data of the corresponding environmental parameters and biometric verification situations. Through these trainings, the deep neural network model has learned the potential relationships and patterns between light intensity, network strength, and biometric verification.

[0116] Afterwards, the collected light intensity data and network strength data of the environment where the cash box is located are input into the deep neural network model. The deep neural network model first preprocesses these input data, and then, according to its internal algorithms and weights, performs layer-by-layer calculations and analyses on the input data.

[0117] By analyzing the input light intensity and network strength data, and combining the knowledge and patterns learned during its training process, the deep neural network model calculates an adjustment coefficient. This adjustment coefficient reflects the degree of influence of the current environmental light intensity and network strength on biometric verification.

[0118] For example, if the light intensity is low, it may affect the accuracy of face recognition. At the same time, a weak network strength may cause delays or incompleteness in the transmission of biometric data, thus affecting the verification process. The deep neural network model comprehensively considers these factors and outputs a suitable adjustment coefficient for subsequent adjustment of the biometric verification weight.

[0119] If the light intensity is extremely low and the network strength is also very weak, the deep neural network model may output a relatively large adjustment coefficient to significantly reduce the weight of biometric verification, or increase the weight of other verification methods to ensure the reliability and security of the verification process.

[0120] After calculating the adjustment coefficient, the deep neural network model outputs it to the system. The system receives this adjustment coefficient and applies it to the subsequent adjustment of the biometric verification weight to optimize the process and results of biometric verification according to the current environmental conditions.

[0121] Please refer to Figure 3 , according to some embodiments of the present invention, the specific steps of using the artificial large model to determine the adjustment coefficient based on the light intensity and network strength of the environment where the cash box is located may include, but are not limited to, the following:

[0122] 301. Obtain the historical accuracy rate;

[0123] 302. Input the light intensity, the network strength, and the location of the cash box into the artificial large model to calculate the environmental risk score;

[0124] 303. Determine the adjustment coefficient according to the environmental risk score and the historical accuracy rate;

[0125] The formula for the adjustment coefficient is:

[0126] W = α * environmental risk score + β * historical accuracy rate, where α and β are dynamic coefficients output by the artificial large model, and W is the adjustment coefficient for biometric verification.

[0127] In the embodiments of the present application, when using a deep neural network model to determine an adjustment coefficient based on the light intensity and network strength of the environment where the money box is located, the historical accuracy rate is first obtained. The data of the historical accuracy rate mainly comes from the past biometric verification records of the system. These records are stored in the system database and cover the detailed situations of biometric verification, including fingerprint, face, and iris information verification, for different users at different times and different money box positions.

[0128] To ensure the relevance and effectiveness of the data, the system will screen the data according to certain rules. For example, it will select the verification records within the past one month or three months, or select the verification records in areas close to the current money box position.

[0129] Statistical analysis is performed on the selected verification records to count the number of successful verifications and the total number of verifications, and then the historical accuracy rate is calculated through the formula "historical accuracy rate = number of successful verifications / total number of verifications". For example, within the past month, a total of 100 biometric verifications were conducted near a certain money box position, and 90 of them were successfully verified. Then the historical accuracy rate of this money box position is 90%.

[0130] The calculated historical accuracy rate will be stored in the system database, and as new verification records are continuously generated, the historical accuracy rate will be updated regularly to reflect the latest verification situation.

[0131] Furthermore, the light intensity, network strength, and the position of the money box are input into the deep neural network model to calculate the environmental risk score. Specifically, the light intensity data and network strength data of the environment where the money box is located that have been collected are input into a pre-trained deep neural network model. The deep neural network model has learned the relationships between a large number of different environmental parameters, such as light intensity, network strength, and money box position, and environmental risk during the training stage. The model will perform a series of calculations and analyses on the input data, such as weighted summation and non-linear transformation through multiple layers of neurons in the neural network.

[0132] After being processed by the model's calculations, an environmental risk score will be output. This score is a quantitative value used to represent the degree of risk that the current environment where the money box is located may pose to biometric verification. The higher the score, the greater the adverse impact of the environment on biometric verification. For example, in a remote area with extremely low light intensity and very weak network strength, the environmental risk score may be relatively high; while in an indoor environment with sufficient light and stable network, the environmental risk score may be relatively low.

[0133] Next, an adjustment coefficient is determined based on the environmental risk score and historical accuracy. Specifically, during the process of calculating the environmental risk score by the deep neural network model, two dynamic coefficients α and β will be output. These two coefficients are dynamically generated by the model based on data such as the input light intensity, network strength, and safe box position, combined with the knowledge it has learned. The dynamic coefficients reflect the importance of the environmental risk score and historical accuracy when determining the adjustment coefficient.

[0134] For example, when the environmental conditions are poor, such as both the light and network conditions are bad, the value of α may be relatively large, meaning that the environmental risk score accounts for a relatively high proportion in the calculation of the adjustment coefficient; while when the environmental conditions are good and the historical accuracy fluctuates greatly, the value of β may be relatively large, indicating that the historical accuracy is more important in the calculation of the adjustment coefficient.

[0135] Next, the adjustment coefficient W is calculated according to the given formula "W = α * (environmental risk score) + β * historical accuracy". For example, assume the environmental risk score is 0.8, the historical accuracy is 0.9, α = 0.6, and β = 0.4. Then the adjustment coefficient W = 0.6 * 0.8 + 0.4 * 0.9 = 0.48 + 0.36 = 0.84.

[0136] The calculated adjustment coefficient W will be used to adjust the verification weights of biometric features to ensure more reasonable biometric verification and improve the accuracy and reliability of verification based on different environmental conditions and historical verification situations.

[0137] Please refer to Figure 4 , according to some embodiments of the present invention, adjusting the verification weights of biometric features according to the adjustment coefficient in step 104 to obtain the target verification weights may specifically include, but are not limited to, the following:

[0138] 401. Adjust the verification weights of the user's fingerprint, face, and iris according to the adjustment coefficient;

[0139] 402. Calculate the verification weights of the user's fingerprint, face, and iris to obtain a target value;

[0140] 403. Determine whether the target value meets a preset value;

[0141] 404. If so, output the target verification weights, and the target verification weights include the verification weights of the user's fingerprint, face, and iris.

[0142] In the embodiments of the present application, the verification weights of the user's fingerprint, face, and iris are adjusted according to the adjustment coefficient. Specifically, in the overall solution, initial verification weights are set for the three biometric features of the user's fingerprint, face, and iris respectively. These initial weights are determined comprehensively based on various factors such as the general reliability of biometric features, the maturity of recognition technology, and the requirements of the application scenario. For example, it may be initially set that the verification weight of the fingerprint is 0.4, the verification weight of the face is 0.3, and the verification weight of the iris is 0.3, and their sum is 1 to ensure a reasonable weight distribution of each biometric feature during comprehensive verification.

[0143] In the foregoing steps, the adjustment coefficient has been calculated. This adjustment coefficient reflects the influence degree of the current environmental light intensity, network intensity, and historical verification situation on biometric feature verification. According to the adjustment coefficient, the system will make corresponding adjustments to the initial verification weights of each biometric feature.

[0144] The system uses multiplication operation to adjust the weights. That is, the new weight of each biometric feature is equal to its initial weight multiplied by the adjustment coefficient. For example, if the adjustment coefficient is 0.84, then the new weight of the fingerprint will become 0.4×0.84 = 0.336, the new weight of the face will become 0.3×0.84 = 0.252, and the new weight of the iris will become 0.3×0.84 = 0.252. Through such calculations, the dynamic adjustment of the biometric feature verification weights according to the environment and historical situations is realized to adapt to different verification conditions.

[0145] After that, the verification weights of the user's fingerprint, face, and iris are calculated to obtain the target value. Specifically, after obtaining the adjusted verification weights of the fingerprint, face, and iris, further calculations need to be performed on these weights to obtain a comprehensive target value.

[0146] Suppose the adjusted verification weight of the fingerprint is 0.336, the verification weight of the face is 0.252, and the verification weight of the iris is 0.252. The system will assign corresponding weighting factors to them according to the importance and mutual relationship of these three biometric features in the actual verification process (these weighting factors can also be dynamically adjusted according to the actual situation). For example, the weighting factors are a, b, and c respectively, and a + b + c = 1. Then the calculation formula for the target value N is: N = a×0.336 + b×0.252 + c×0.252. Through this weighted summation calculation, the verification weights of the three biometric features are combined to obtain a target value that can reflect the overall verification weight situation.

[0147] Next, it is further determined whether the target value meets the preset value. Herein, the preset value is a reference standard preset in advance according to various requirements such as the security and accuracy of biometric verification. For example, in some scenarios with high security requirements, the preset value may be set relatively high to ensure that verification is passed only when the biometric verification is very accurate; while in some scenarios with relatively high convenience requirements, the preset value may be appropriately reduced.

[0148] The calculated target value is compared with the preset value, and the system will make a judgment based on the comparison result. If the target value is greater than or equal to the preset value, it is considered that the preset condition is met; if the target value is less than the preset value, it is considered that the preset condition is not met. This judgment process is to ensure that the final obtained biometric verification weight combination is reasonable and meets the system requirements.

[0149] When the judgment result is that the target value meets the preset value, the system will perform an output operation. This means that the currently adjusted biometric verification weight combination is reasonable and meets the verification criteria set by the system, and can be used for subsequent biometric verification processes.

[0150] The target verification weights output by the system include the verification weights of the user's fingerprint, face, and iris. These weights will be applied to the biometric verification module to determine the importance of each biometric in the comprehensive verification when verifying the user's biometric characteristics.

[0151] Please refer to Figure 5 , according to some embodiments of the present invention, in step 105, when verifying the user biometric characteristics according to the target verification weights, after determining that the biometric verification is passed, obtaining the unlocking instruction of the money box specifically may include, but is not limited to, the following:

[0152] 501. Perform multimodal fusion comparison on the user biometric characteristics to generate an original similarity score;

[0153] 502. Calculate a target threshold according to the target verification weights;

[0154] Target threshold calculation formula: S = S0 × T, where S is the target threshold, S0 is the original similarity score, and T is the target verification weight;

[0155] 503. Determine a preset threshold;

[0156] 504. When the target threshold is greater than the preset threshold, determine that the biometric verification is passed, and then obtain the unlocking instruction of the money box.

[0157] In the embodiments of the present application, first, multimodal fusion comparison is performed on the user's biometric features to generate an original similarity score. Specifically, according to the three biometric feature information of fingerprint, face, and iris provided by the user, after collection, these information have been preprocessed respectively. For example, denoising and contrast enhancement of fingerprint images, normalization and feature point detection of face images, segmentation and texture feature extraction of iris images, etc., to improve the accuracy of subsequent comparison.

[0158] Multimodal fusion comparison is to comprehensively analyze and compare the three different types of biometric feature information of fingerprint, face, and iris. The system will adopt a multimodal fusion algorithm, which combines the characteristics and advantages of different biometric features. For example, for fingerprints, the similarity may be calculated by comparing the positions and directions of minutiae; for faces, the relative relationships between facial feature points will be analyzed; for irises, their unique texture patterns will be compared. Then, the comparison results of these three biometric features are fused, and a weighted average or other fusion strategies are adopted to obtain a comprehensive original similarity score.

[0159] For example: Suppose the similarity score obtained from fingerprint comparison is 80 points, the similarity score obtained from face comparison is 70 points, and the similarity score obtained from iris comparison is 85 points. The system calculates the original similarity score through weighted average according to the preset fusion weights (for example, fingerprint weight 0.3, face weight 0.3, iris weight 0.4): Original similarity score = 80×0.3 + 70×0.3 + 85×0.4 = 24 + 21 + 34 = 79 points.

[0160] In the foregoing steps, the target verification weights have been obtained, which include the verification weights of the user's fingerprint, face, and iris. These weights are dynamically adjusted according to the light intensity, network intensity, and historical accuracy, reflecting the importance of each biometric feature under the current verification conditions.

[0161] The calculation of the target threshold is to dynamically determine a reasonable verification standard according to different verification situations. Because in different environments and historical situations, the accuracy requirements for biometric verification may be different, so it is necessary to adjust the threshold according to the target verification weights to ensure the reliability of the verification results.

[0162] The system will further process the original similarity score according to the target verification weights to calculate the target threshold. Target threshold calculation formula: S = S0×T, where S is the target threshold, which is the standard value used to judge whether the biometric verification passes under the current verification conditions, S0 is the original similarity score, and T is the target verification weight.

[0163] Assume the original similarity score S0 = 79 points and the target verification weight T = 0.8. Then, according to the formula S = S0×T, the target threshold S = 79×0.8 = 63.2 points.

[0164] After that, the system will compare the calculated target threshold with the determined preset threshold. If the target threshold is greater than the preset threshold, it indicates that the biometric feature provided by the user has a high enough similarity with the template stored in the system under the current verification conditions, meeting the verification criteria set by the system. Therefore, it can be determined that the biometric verification is passed.

[0165] After the biometric verification is passed, the system will obtain the unlocking instruction for the cash box from the pre-set instruction source. This unlocking instruction is a specific code or signal used to control the unlocking mechanism of the cash box to perform the unlocking operation. The system will send the unlocking instruction to the cash box through a secure communication method to ensure the security and accuracy of the instruction during transmission. If the target threshold is not greater than the preset threshold, the biometric verification fails. The system will not obtain the unlocking instruction and may prompt the user to re-perform the biometric verification or take other measures.

[0166] Please refer to Figure 6 , the second aspect of this application provides a cash box automatic unlocking verification device, including:

[0167] The first acquisition unit 601 is used to acquire the cash box unlocking request of the user. The cash box unlocking request includes the user's biometric feature, and the user's biometric feature includes fingerprint, face, and iris information;

[0168] The first determination unit 602 is used to determine the position of the cash box;

[0169] The second determination unit 603 is used to collect the current environmental parameters of the cash box according to the position of the cash box, and use the artificial large model to determine the adjustment coefficient according to the environmental parameters;

[0170] The adjustment unit 604 is used to adjust the verification weight of the biometric feature according to the adjustment coefficient to obtain the target verification weight;

[0171] The second acquisition unit 605 is used to verify the user's biometric feature according to the target verification weight. After determining that the biometric verification is passed, the unlocking instruction of the cash box is acquired;

[0172] The verification unit 606 is used to send the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction;

[0173] The collection unit 607 is used to perform the unlocking process on the cash box using the unlocking instruction and collect the unlocking information during the unlocking process when it is determined that the verification is passed;

[0174] A third acquisition unit 608, configured to encrypt the unlocking information to obtain encrypted unlocking information;

[0175] An upload unit 609, configured to upload the encrypted unlocking information to a blockchain system for storage.

[0176] Please refer to Figure 7 , this application also provides a cash box automatic unlocking verification device, including:

[0177] A processor 701, a storage 702, an input / output unit 703, and a bus 704;

[0178] The processor 701 is connected to the storage 702, the input / output unit 703, and the bus 704;

[0179] The storage 702 stores a program, and the processor 701 calls the program to execute any of the above methods.

[0180] This application also relates to a computer-readable storage medium, on which a program is stored. When the program runs on a computer, the computer is enabled to execute any of the above methods.

[0181] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0182] In several embodiments provided by this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in an electrical, mechanical, or other form.

[0183] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0184] In addition, each functional unit in various embodiments of the present application may be integrated into one processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0185] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present application. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs and other various media that can store program codes.

[0186] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data for analysis, stored data, displayed data, etc.), and signals (including but not limited to signals transmitted between user terminals and other devices, etc.) involved in the present application are all authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions. For example, the relevant information or data such as "human face" and "face" involved in the present disclosure are obtained under full authorization.

Claims

1. An automatic unlocking verification method for a money box, characterized in that, The method includes: Obtaining a request for unlocking a cash box from a user, where the request for unlocking the cash box includes user biometric features, and the user biometric features include fingerprint, face, and iris information; Determining the location of the cash box; Collecting the light intensity and network strength of the environment where the cash box is located according to the location of the cash box; Obtaining the historical accuracy rate; Inputting the light intensity, the network strength, and the location of the cash box into an artificial intelligence model to calculate the environmental risk score; Determining an adjustment coefficient according to the environmental risk score and the historical accuracy rate; The formula for the adjustment coefficient is: W = α * environmental risk score + β * historical accuracy rate, where α and β are dynamic coefficients output by the artificial intelligence model, and W is the adjustment coefficient for biometric verification; Adjusting the verification weight of the biometric features according to the adjustment coefficient to obtain the target verification weight; Verifying the user biometric features according to the target verification weight. After determining that the biometric verification is passed, obtain the unlocking instruction for the cash box; Sending the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction; When it is determined that the verification is passed, using the unlocking instruction to unlock the cash box and collecting the unlocking information during the unlocking process; Performing encryption processing on the unlocking information to obtain encrypted unlocking information; Uploading the encrypted unlocking information to a blockchain system for storage.

2. The automatic unlocking and verification method for the money box according to claim 1, wherein, Adjusting the verification weight of the biometric features according to the adjustment coefficient to obtain the target verification weight, including: Adjusting the verification weights of the user's fingerprint, face, and iris according to the adjustment coefficient; Calculating the verification weights of the user's fingerprint, face, and iris to obtain a target value; Determining whether the target value meets a preset value; If so, outputting the target verification weight, where the target verification weight includes the verification weights of the user's fingerprint, face, and iris.

3. The automatic unlocking and verification method for the money box according to claim 1, wherein Verifying the user biometric features according to the target verification weight. After determining that the biometric verification is passed, obtain the unlocking instruction for the cash box, including: Performing multi-modal fusion comparison on the user biometric features to generate an original similarity score; Calculating a target threshold according to the target verification weight; The formula for the target threshold: S = S0 × T, where S is the target threshold, S0 is the original similarity score, and T is the target verification weight; Determining a preset threshold; When the target threshold is greater than the preset threshold, determining that the biometric verification is passed, and obtaining the unlocking instruction for the cash box.

4. The automatic unlocking verification method for the money box according to claim 1, characterized in that When it is determined that the verification is passed, using the unlocking instruction to unlock the cash box and collecting the unlocking information during the unlocking process, including: When it is determined that the verification is passed, sending the unlocking instruction to the cash box so that the cash box performs unlocking processing according to the unlocking instruction; After the cash box is opened, collecting the operator ID, timestamp, and geographical location information during the unlocking process.

5. The automatic unlocking verification method for the money box according to claim 1, characterized in that, After sending the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction, the method further includes: When it is determined that the verification fails, sending an alarm message and collecting the information indicating that the verification fails; Encrypt the information that fails the verification and upload it to the blockchain system for storage.

6. An automatic unlocking verification device for a money box, characterized in that, The device includes: A first acquisition unit, configured to acquire a request for unlocking a cash box of a user, where the request for unlocking the cash box includes user biometric features, and the user biometric features include fingerprint, face, and iris information; A first determination unit, configured to determine the position of the cash box; A second determination unit, configured to collect the light intensity and network intensity of the environment where the cash box is located according to the position of the cash box; acquire the historical accuracy rate; input the light intensity, the network intensity, and the position of the cash box into an artificial intelligence model to calculate the environmental risk score; determine an adjustment coefficient according to the environmental risk score and the historical accuracy rate; the formula for the adjustment coefficient is: W = α * environmental risk score + β * historical accuracy rate, where α and β are dynamic coefficients output by the artificial intelligence model, and W is the adjustment coefficient for biometric verification; An adjustment unit, configured to adjust the verification weight of the biometric features according to the adjustment coefficient to obtain a target verification weight; A second acquisition unit, configured to verify the user biometric features according to the target verification weight, and after determining that the biometric verification is passed, acquire an unlocking instruction for the cash box; A verification unit, configured to send the unlocking instruction to the cash box so that the cash box verifies the unlocking instruction; A collection unit, configured to, when determining that the verification is passed, perform an unlocking process on the cash box using the unlocking instruction and collect unlocking information during the unlocking process; A third acquisition unit, configured to encrypt the unlocking information to obtain encrypted unlocking information; An upload unit, configured to upload the encrypted unlocking information to the blockchain system for storage.

7. An automatic unlocking verification device for a money box, characterized in that, The device includes: A processor, a storage, an input / output unit, and a bus; The processor is connected to the storage, the input / output unit, and the bus; The storage stores a program, and the processor calls the program to execute the method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, A program is stored on the computer-readable storage medium, and when the program is executed on a computer, it executes the method according to any one of claims 1 to 5.

Images