Multi-user unlocking control method, device and storage medium for an intelligent box

Abstract

The present application discloses a multi-user unlocking control method, device and storage medium for an intelligent box, which is used to improve the security of the intelligent box. Obtain unlocking requests from two management terminals; check the terminal tag information; when the tag is compliant, obtain biometric information, fusion number information and a password to be verified; determine the corresponding artificial intelligence feature fusion model and fusion materials according to the fusion number information of the management terminal; perform feature fusion processing on the fusion materials and the corresponding biometric information through the artificial intelligence feature fusion model; input the feature fusion data into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data; screen the artificial intelligence feature discrimination models according to the discrimination distribution data, and generate a numbered password according to the digital tags of the screened artificial intelligence feature discrimination models; verify the target password, numbered password and password to be verified, and when the verification is passed, perform an unlocking process on the intelligent box.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of intelligent safes, and in particular, to a multi-user unlocking control method, device, and storage medium for an intelligent safe. Background Art

[0002] Nowadays, in response to the storage requirements of important items, more and more intelligent safes have been developed. Especially in the financial field, there are not only intelligent safes but also storage repositories dedicated to storing intelligent safes.

[0003] In order to increase the safety factor of intelligent safes in the financial field and prevent the loss of items in the intelligent safes, the unlocking conditions are usually controlled by the administrator. Such unlocking conditions are usually items that are easy to lose, such as passwords and keys. To ensure the security of the items in the intelligent safe, multiple management terminals are usually selected for unlocking cooperation. It is necessary to initiate unlocking requests from multiple management terminals simultaneously within the same time period and verify the unlocking conditions. The unlocking conditions are converted from simple unlocking tools such as passwords and keys to biotags with stronger discrimination, and there are verification means for the cooperation of multiple types of unlocking tools.

[0004] Nowadays, in the field of unlocking based on biometric recognition, especially face recognition, fingerprint recognition, and pupil recognition are more common, and the traditional password is also added to form a composite unlocking rule to increase security. In the prior art, multiple administrators control the management terminals to send unlocking requests to the intelligent safe terminal within the same time period. After the intelligent safe terminal determines that the control terminal meets the unlocking verification conditions, biometric collection instructions and password collection instructions are sent to multiple administrators. After the intelligent safe terminal obtains the biometric information and the password to be verified, it starts to identify, compares it with the biometric stored internally, and compares the password to be verified with the password in the internal storage repository. After the comparison is completed, it is determined whether to unlock the intelligent safe according to the comparison result.

[0005] However, nowadays, with the continuous rise of AI artificial intelligence and network technology, the security of biometrics and passwords has gradually declined. Because nowadays, the images synthesized by AI artificial intelligence are getting closer and closer to real scenarios, and even biometrics can be transformed in videos. Moreover, passwords are at risk of being stolen through various channels. This leads to the situation that when the intelligent safe terminal obtains biometric information, it may obtain biometrics generated by AI, while the password is stolen data. When a non-manager operates on the management terminal and conducts operations on the acquisition device of the intelligent safe or the management terminal through AI artificial intelligence, AI biometric images and AI biometric videos can be generated. Since the intelligent safe terminal is usually in a single-machine state during the non-unlocking stage and its AI recognition ability cannot be updated in real time, there is a possibility that a non-manager can unlock the intelligent safe terminal through AI artificial intelligence, resulting in a reduction in the security of the intelligent safe. Summary of the Invention

[0006] The present application discloses a multi-user unlocking control method, device, and storage medium for an intelligent safe, which are used to improve the security of the intelligent safe.

[0007] The first aspect of the present application discloses a multi-user unlocking control method for an intelligent safe, including:

[0008] Obtain unlocking requests respectively sent by two management terminals, where the unlocking requests include the terminal tag information of the two management terminals respectively;

[0009] Verify the terminal tag information of the two management terminals respectively;

[0010] When the tag verifications are all in line, obtain the biometric information, fusion number information, and password to be verified respectively sent by the two management terminals;

[0011] Determine the corresponding artificial intelligence feature fusion model and fusion materials from the box body database according to the fusion number information of each management terminal. There are also several artificial intelligence feature discrimination models set in the box body database. Each artificial intelligence feature discrimination model has a different digital tag, and each artificial intelligence feature discrimination model is an artificial intelligence model that has completed different degrees of pre-training on various biometric data;

[0012] Perform feature fusion processing on the fusion materials and the corresponding biometric information through the artificial intelligence feature fusion model to generate feature fusion data;

[0013] Input the feature fusion data of each management terminal into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data;

[0014] Screen the artificial intelligence feature discrimination model according to the probability distribution in the discrimination distribution data, and generate a numbered password based on the digital label of the screened artificial intelligence feature discrimination model;

[0015] Verify the target password, numbered password, and password to be verified in the box database, and when the verification passes, perform an unlocking process on the intelligent box.

[0016] Optionally, before the step of obtaining the unlocking requests sent by the two management terminals respectively, the multi-user unlocking control method further includes:

[0017] Receive takeover requests sent by the two management terminals, where the takeover requests include the takeover terminal information of the two management terminals;

[0018] Obtain the real-time positioning information of the box, and use the real-time positioning information of the box to verify the inbound and outbound records with the repository server, and obtain the secret key;

[0019] When the inbound and outbound records meet the conditions, obtain the encrypted takeover end verification information from the box database, and the encrypted takeover end verification information is the takeover terminal information pre-entered into the box database by the previous repository server;

[0020] Decrypt the encrypted takeover end verification information with the secret key, and verify the takeover terminal information of the two management terminals with the decrypted takeover end verification information;

[0021] When the verification passes, determine that the two management terminals have the identity of new managers, generate a biometric collection instruction for each of the two management terminals, and send the corresponding biometric collection instructions to the two management terminals respectively;

[0022] Obtain the first biometric information sent by the two management terminals, and the first biometric information is biometric information of the image acquisition type;

[0023] Determine several artificial intelligence feature discrimination models and several artificial intelligence feature fusion models from the box database, and select corresponding types of fusion materials for the two management terminals respectively from the fusion material library in the box database;

[0024] Input the first biometric information and the corresponding fusion material into the artificial intelligence feature fusion model, add the features of the fusion material to the first biometric information, and generate feature fusion information for the two management terminals respectively;

[0025] Input the two pieces of feature fusion information into several artificial intelligence feature discrimination models for discrimination, and generate a first discrimination result and a second discrimination result;

[0026] Determine the digital tags corresponding to the eligible artificial intelligence feature discrimination models respectively according to the probability distributions in the first discrimination result and the second discrimination result, and generate two numbered passwords based on the two groups of digital tags. The two numbered passwords are used as the target passwords for unlocking the intelligent safe box.

[0027] Use the fusion numbered information of the two fusion materials and the two numbered passwords to generate corresponding takeover success messages for the two management terminals respectively, and send the takeover success messages to the two management terminals respectively.

[0028] Optionally, after the step of inputting the two feature fusion information into a number of artificial intelligence feature discrimination models respectively for discrimination to generate the first discrimination result and the second discrimination result, and before the step of determining the digital tags corresponding to the eligible artificial intelligence feature discrimination models respectively according to the probability distributions in the first discrimination result and the second discrimination result and generating two numbered passwords based on the two groups of digital tags, the multi-user unlocking control method further includes:

[0029] Input the first biometric information sent by the two management terminals into a number of artificial intelligence feature discrimination models respectively for discrimination to generate the third discrimination result and the fourth discrimination result;

[0030] Conduct discrimination probability analysis on the first discrimination result, the second discrimination result, the third discrimination result and the fourth discrimination result;

[0031] When the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are the same, it is determined that the fusion material meets the standard;

[0032] When the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are different, re-select new fusion materials and continue with feature fusion and model discrimination.

[0033] Optionally, the unlocking request includes the current location information of each of the two management terminals;

[0034] After the step of obtaining the unlocking requests sent by the two management terminals respectively, and before the step of verifying the terminal tag information of the two management terminals respectively, the multi-user unlocking control method further includes:

[0035] Obtain the real-time positioning information of the box body, and determine the location information and link establishment information of the current storage server from the box body database;

[0036] When the real-time positioning information of the box body meets the location condition with the location information of the current storage server, establish a link with the storage server through the link establishment information and obtain the inbound and outbound records;

[0037] When the inbound and outbound records show that the intelligent cash box has not been shipped out, obtain the unlocking position information reserved by the two managers from the box database;

[0038] Analyze the current position information of each of the two management terminals and the unlocking position information reserved by the two managers;

[0039] When the position analysis result meets the conditions, determine that the intelligent cash box preliminarily meets the unlocking verification conditions.

[0040] Optionally, after verifying the target password, numbered password, and password to be verified in the box database and performing the unlocking process on the intelligent cash box when the verification passes, the multi-user unlocking control method further includes:

[0041] Receive the handover requests sent by each of the two management terminals. The handover requests include the encrypted takeover end verification information of the two new management terminals, and the handover requests also include the location information and link establishment information of the new repository server;

[0042] Store the encrypted takeover end verification information, the location information of the new repository server, and the link establishment information in the box database;

[0043] Receive the outbound notice sent by the repository server and perform the handover block processing on the intelligent cash box.

[0044] Optionally, the steps of obtaining the real-time positioning information of the box, verifying the inbound and outbound records with the repository server through the real-time positioning information of the box, and obtaining the key include:

[0045] Obtain the real-time positioning information of the box, and perform position analysis on the real-time positioning information of the box and the location information of the repository server;

[0046] When the position analysis is in line, establish an interactive channel with the repository server through the link establishment information of the repository server;

[0047] Send the inbound and outbound records and the key acquisition request to the repository server;

[0048] Receive the inbound and outbound records and the key sent by the repository server, and verify the inbound and outbound records.

[0049] The second aspect of the present application discloses a multi-user unlocking control device for an intelligent cash box, including:

[0050] The first acquisition unit is used to acquire the unlocking requests sent by each of the two management terminals. The unlocking requests include the terminal tag information of each of the two management terminals;

[0051] The verification unit is used to verify the terminal tag information of each of the two management terminals;

[0052] A second acquisition unit, configured to acquire biometric information, fusion number information, and passwords to be verified sent by two management terminals respectively when the label verifications are all compliant;

[0053] A first determination unit, configured to determine corresponding artificial intelligence feature fusion models and fusion materials from a box database according to the fusion number information of each management terminal. A number of artificial intelligence feature discrimination models are also set in the box database. Each artificial intelligence feature discrimination model has a different digital label, and each artificial intelligence feature discrimination model is an artificial intelligence model that has completed different degrees of pre-training on various biometric data;

[0054] A first generation unit, configured to perform feature fusion processing on the fusion materials and corresponding biometric information through an artificial intelligence feature fusion model to generate feature fusion data;

[0055] A second generation unit, configured to input the feature fusion data of each management terminal into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data;

[0056] A third generation unit, configured to screen the artificial intelligence feature discrimination models according to the probability distribution in the discrimination distribution data, and generate a numbered password according to the digital label of the screened artificial intelligence feature discrimination model;

[0057] A verification unit, configured to verify the target password, numbered password, and password to be verified in the box database, and perform an unlocking process on the intelligent box when the verification is passed.

[0058] Optionally, before the first acquisition unit, the multi-user unlocking control device further includes:

[0059] A first receiving unit, configured to receive takeover requests sent by two management terminals, where the takeover requests include takeover terminal information of the two management terminals;

[0060] A third acquisition unit, configured to acquire box real-time positioning information, and check the inbound and outbound records with a repository server through the box real-time positioning information, and acquire a key;

[0061] A fourth acquisition unit, configured to acquire encrypted takeover end verification information from the box database when the inbound and outbound records meet the conditions, where the encrypted takeover end verification information is the takeover terminal information pre-input into the box database by the previous repository server;

[0062] A decryption unit, configured to decrypt the encrypted takeover end verification information with the key, and verify the takeover terminal information of the two management terminals with the decrypted takeover end verification information;

[0063] A second determination unit, configured to determine that the two management terminals have the identity of new managers when the verification is passed, generate a biometric collection instruction for each of the two management terminals, and send the corresponding biometric collection instructions to the two management terminals respectively;

[0064] A fifth acquisition unit, configured to acquire the first biometric information sent by the two management terminals, where the first biometric information is biometric information of the image acquisition type;

[0065] A third determination unit, configured to determine a plurality of artificial intelligence feature discrimination models and a plurality of artificial intelligence feature fusion models from the box database, and respectively select corresponding types of fusion materials for the two management terminals from the fusion material library of the box database;

[0066] A fusion unit, configured to input the first biometric information and the corresponding fusion material into the artificial intelligence feature fusion model, add the features of the fusion material to the first biometric information, and generate feature fusion information for the two management terminals respectively;

[0067] A fourth generation unit, configured to input the two pieces of feature fusion information into a plurality of artificial intelligence feature discrimination models for discrimination respectively, and generate a first discrimination result and a second discrimination result;

[0068] A fifth generation unit, configured to respectively determine the digital labels corresponding to the artificial intelligence feature discrimination models that meet the conditions according to the probability distributions in the first discrimination result and the second discrimination result, and generate two numbered passwords according to the two sets of digital labels, where the two numbered passwords are used as the target passwords for unlocking the intelligent safe box;

[0069] A sending unit, configured to generate corresponding takeover success information for the two management terminals respectively using the fusion number information of the two fusion materials and the two numbered passwords, and send the takeover success information to the two management terminals respectively.

[0070] Optionally, after the fourth generation unit and before the fifth generation unit, the multi-user unlocking control device further includes:

[0071] A sixth generation unit, configured to input the first biometric information sent by the two management terminals into a plurality of artificial intelligence feature discrimination models for discrimination respectively, and generate a third discrimination result and a fourth discrimination result;

[0072] A first analysis unit, configured to perform discrimination probability analysis on the first discrimination result, the second discrimination result, the third discrimination result, and the fourth discrimination result;

[0073] A fifth determination unit, configured to determine that the fusion material meets the standard when the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are the same;

[0074] A selection unit, configured to re-select new fusion materials to continue feature fusion and model discrimination when the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are different.

[0075] Optionally, the unlocking request includes the current location information of each of the two management terminals;

[0076] After the first acquisition unit and before the verification unit, the multi-user unlocking control device further includes:

[0077] A sixth acquisition unit, configured to acquire the real-time positioning information of the box body, and determine the location information and link establishment information of the current repository server from the box body database;

[0078] A seventh acquisition unit, configured to establish a link with the repository server through the link establishment information and acquire the inbound and outbound records when the real-time positioning information of the box body meets the location condition with the location information of the current repository server;

[0079] An eighth acquisition unit, configured to acquire the unlocking location information reserved by the two managers from the box body database when the inbound and outbound records show that the intelligent box has not been shipped out;

[0080] A second analysis unit, configured to analyze the current location information of each of the two management terminals and the unlocking location information reserved by the two managers;

[0081] A fourth determination unit, configured to determine that the intelligent box preliminarily meets the unlocking verification condition when the location analysis result meets the conditions.

[0082] Optionally, after the verification unit, the multi-user unlocking control device further includes:

[0083] A second receiving unit, configured to receive the handover requests sent by each of the two management terminals, where the handover request includes the encrypted takeover end verification information of the two new management terminals, and the handover request further includes the location information and link establishment information of the new repository server;

[0084] A storage unit, configured to store the encrypted takeover end verification information, the location information of the new repository server, and the link establishment information into the box body database;

[0085] A third receiving unit, configured to receive the outbound notice sent by the repository server and perform handover lockdown processing on the intelligent box.

[0086] Optionally, the third acquisition unit includes:

[0087] Acquire the real-time positioning information of the box body, and perform location analysis on the real-time positioning information of the box body and the location information of the repository server;

[0088] When the location analysis is in line, an interaction channel is established between the information and the repository server through the link of the repository server;

[0089] Send the incoming and outgoing record and the key acquisition request to the repository server;

[0090] Receive the incoming and outgoing record and the key sent by the repository server, and check the incoming and outgoing record.

[0091] The third aspect of this application provides a multi-user unlocking control device for an intelligent cashbox, including:

[0092] A processor, a memory, an input / output unit, and a bus;

[0093] The processor is connected to the memory, the input / output unit, and the bus;

[0094] The memory stores a program, and the processor calls the program to execute the multi-user unlocking control method as in the first aspect and any optional multi-user unlocking control method of the first aspect.

[0095] The fourth aspect of this application provides a computer-readable storage medium, on which a program is stored, and when the program is executed on a computer, it executes the multi-user unlocking control method as in the first aspect and any optional multi-user unlocking control method of the first aspect.

[0096] It can be seen from the above technical solutions that the embodiments of this application have the following advantages:

[0097] In this application, first, unlocking requests of multiple management terminals need to be obtained within the same time period. If there are insufficient management terminals, unlocking will not be directly permitted. Next, the unlocking permissions of the management terminals themselves are detected. When all management terminals have unlocking permissions, biometric information, combined number information, and passwords to be verified are collected through the management terminals. Among them, the combined number information and passwords to be verified are auxiliary unlocking data generated when binding the management terminal and the manager. When verifying, the manager needs to input the corresponding combined number information and passwords to be verified. First, the collected biometric information is subjected to routine identification and verification. If the routine verification fails, there is no need to continue. At this time, the corresponding combined materials are determined in the box database through the combined number information of each management terminal. The combined materials are subjected to feature extraction through a preset artificial intelligence feature fusion model and fused into the biometric information according to the preset channel ratio and rules (the fusion rules for different biometric types are different) to generate feature fusion data. Next, the feature fusion data of each management terminal is input into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data. Then, the artificial intelligence feature discrimination models are screened according to the probability distribution in the discrimination distribution data, and numbered passwords are generated according to the digital tags of the screened artificial intelligence feature discrimination models. Finally, the target password, numbered password, and password to be verified in the box database are verified. When the verification passes, the intelligent safe is unlocked.

[0098] Due to the differences in the parameters and structures of each artificial intelligence feature discrimination model, the analysis results of each artificial intelligence feature discrimination model for the same feature fusion data are different. The target password is the biometric image or video truly collected during the binding process of the manager. First, the combined materials are selected to fuse the biometric image or video to a certain extent, and then all artificial intelligence feature discrimination models are used to distinguish between true and false, obtaining a set of biometric models judged as true and a set of biometric models judged as false. Then, numbered passwords are generated according to the digital tags of one type of biometric model. Each manager has a unique numbered password, and all the numbered passwords are integrated to form the target password. The numbered password and the combined number information corresponding to the combined materials are sent to the corresponding management terminal for the manager to save. In the unlocking process, if the collected biometric information is real, after feature fusion and discrimination, the set of biometric models judged as true and the set of biometric models judged as false are the same as those during binding, that is, the generated numbered passwords are also the same. On the contrary, if the collected biometric information itself is AI-generated, after fusing the biometric information with the combined materials, the set of biometric models judged as true and the set of biometric models judged as false are different from those during binding, and the numbered passwords are also different, improving the security of the intelligent safe. Description of the Drawings

[0099] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0100] Figure 1 Schematic diagram of an embodiment of the multi-user unlocking control method for the intelligent cash box of the present application;

[0101] Figure 2 Schematic diagram of an embodiment of the takeover method for the intelligent cash box of the present application;

[0102] Figure 3 Schematic diagram of an embodiment of the determination method for the fusion materials of the present application;

[0103] Figure 4 Schematic diagram of an embodiment of the method for verifying the unlocking environment of the present application;

[0104] Figure 5 Schematic diagram of an embodiment of the method for transferring the intelligent cash box of the present application;

[0105] Figure 6 Schematic diagram of an embodiment of the method for interacting with the repository server of the present application;

[0106] Figure 7 Schematic diagram of an embodiment of the multi-user unlocking control device for the intelligent cash box of the present application;

[0107] Figure 8 Schematic diagram of another embodiment of the multi-user unlocking control device for the intelligent cash box of the present application. Detailed implementation manners

[0108] In the following description, for the purpose of illustration rather than limitation, specific details such as specific system architectures and technologies are presented to thoroughly understand the embodiments of the present application. However, those skilled in the art should understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0109] It should be understood that when used in the specification of the present application and the appended claims, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0110] It should also be understood that the term "and / or" as used in the specification and appended claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0111] As used in the specification and appended claims of this application, the term "if" may be construed, depending on the context, as "when", "once", "in response to determining", or "in response to detecting". Similarly, the phrases "if determined" or "if [the described condition or event] is detected" may be construed, depending on the context, as meaning "once determined", "in response to determining", "once [the described condition or event] is detected", or "in response to detecting [the described condition or event]".

[0112] In addition, in the description of the specification and appended claims of this application, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be construed as indicating or implying relative importance.

[0113] Reference to "one embodiment" or "some embodiments" or the like described in the specification of this application means that a particular feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of this application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "comprising", "including", "having", and their variants all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0114] In the prior art, multiple administrators control the management terminal to send an unlocking request to the intelligent cash box terminal during the same time period. After the intelligent cash box terminal determines that the control terminal meets the unlocking verification conditions, it sends a biometric collection instruction and a password collection instruction to multiple administrators. After the intelligent cash box terminal obtains the biometric information and the password to be verified, it starts to identify, judges whether to compare it with the biometric stored internally, compares the password to be verified with the password in the internal repository, and after the comparison is completed, determines whether to unlock the intelligent cash box according to the comparison result.

[0115] However, nowadays, with the continuous rise of AI artificial intelligence and network technology, the security of biometrics and passwords has gradually declined. Because nowadays, the images synthesized by AI artificial intelligence are getting closer and closer to real scenarios, and even biometrics can be transformed in videos. Moreover, passwords are at risk of being stolen in various ways. This leads to the situation that when the intelligent safe terminal obtains biometric information, it may obtain biometrics generated by AI, while the password is the stolen data. When a non-manager operates on the management terminal and conducts operations on the acquisition device of the intelligent safe or the management terminal through AI artificial intelligence, AI biometric images and AI biometric videos can be generated. Since the intelligent safe terminal is usually in a single-machine state during the non-unlocking stage and its AI recognition ability cannot be updated at all times, there is a possibility that a non-manager can unlock the intelligent safe terminal through AI artificial intelligence, resulting in a reduction in the security of the intelligent safe.

[0116] Based on this, the present application discloses a multi-user unlocking control method, device and storage medium for an intelligent safe, which are used to improve the security of the intelligent safe.

[0117] Next, the technical solutions in the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0118] The method of the present application can be applied to a server, a device, a terminal or other devices with logical processing capabilities. In this regard, the present application makes no limitation. For the convenience of description, the following description will be made taking the execution entity as a terminal as an example.

[0119] Please refer to Figure 1 , an embodiment of a multi-user unlocking control method for an intelligent safe provided by the present application includes:

[0120] 101. Obtain the unlocking requests sent by two management terminals respectively, where the unlocking requests include the terminal tag information of the two management terminals respectively;

[0121] In this embodiment, the intelligent safe terminal can obtain the unlocking requests sent by more than two management terminals respectively, and the unlocking requests need to be transmitted to the intelligent safe terminal through the communication channel set by the repository server.

[0122] 102. Check the terminal tag information of the two management terminals respectively;

[0123] The intelligent box terminal first verifies the terminal label information in the request. When the intelligent box terminal is stored in the repository, it is bound to the corresponding management terminal of the administrator, and the identity data of the management terminal is reserved.

[0124] 103. When the label inspections are all compliant, obtain the biometric information, fusion number information, and password to be verified sent by each of the two management terminals.

[0125] When the label inspections are all compliant, that is, when the intelligent box terminal determines that the identity of the management terminal is compliant, it sends a notice of unlocking data transmission to the management terminal, controlling the person of the management terminal to input the corresponding biometric characteristics, fusion number information, and password to be verified. Among them, during the binding process of the intelligent box terminal and the management terminal, the administrator has stored the biometric characteristics of the self-selected type of the administrator in the intelligent box. And after the intelligent box terminal screens the fusion materials and performs feature fusion processing according to the biometric information, it generates a numbered password based on the fused data, and then transmits the numbered password and the fusion number information corresponding to the fusion materials to the administrator corresponding to the management terminal. Therefore, when unlocking, the controller of the management terminal needs to provide these three data.

[0126] After receiving these three data, first detect the biometric information. If the biometric information does not match the reserved biometric information, there is no need to perform subsequent steps. The embodiments of the present application discuss the situation where the biometric information matches the reserved biometric information, and further check whether there is AI processing for the biometric information.

[0127] 104. Determine the corresponding artificial intelligence feature fusion model and fusion materials from the box body database according to the fusion number information of each management terminal. There are also several artificial intelligence feature discrimination models set in the box body database. Each artificial intelligence feature discrimination model has a different digital label, and each artificial intelligence feature discrimination model is an artificial intelligence model that has completed different degrees of pre-training on various biometric data.

[0128] The intelligent box terminal first determines the corresponding artificial intelligence feature fusion model and fusion materials from the box body database according to the fusion number information of each management terminal. The type of the fusion materials of the fusion number information is the same as the biometric information, and the determined artificial intelligence feature fusion model is also specifically loaded into the intelligent box for this type of biometric characteristics. For example: the biometric information of the user is face recognition, and the fusion number information determines the corresponding face simulation materials from the box body database, and then selects the artificial intelligence feature fusion model of the corresponding face features from the box body database.

[0129] If the biometric information of the user does not match the fusion materials of the fusion number information, subsequent identification cannot unlock the lock either.

[0130] 105. The fusion material and the corresponding biometric information are subjected to feature fusion processing through an artificial intelligence feature fusion model to generate feature fusion data.

[0131] The terminal subjects the fusion material and the corresponding biometric information to feature fusion processing through an artificial intelligence feature fusion model to generate feature fusion data. In this embodiment, the artificial intelligence feature fusion model first extracts features from the fusion material and then performs feature fusion according to the characteristics of the biometric feature. For example, if it is face recognition, the face edge, neck area, etc. in the biometric information are first determined for feature fusion to a preset degree. If the transmitted biometric information has already been processed by AI, the feature fusion of the intelligent lockbox will increase the trace of this AI processing.

[0132] 106. The feature fusion data of each management terminal is input into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data.

[0133] After generating the feature fusion feature data, the intelligent lockbox terminal inputs the feature fusion data of each management terminal into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data. The feature fusion data corresponding to the first management terminal is input into several artificial intelligence feature discrimination models stored in the box body database for detection to obtain the analysis data of each artificial intelligence feature discrimination model. The same applies to the second management terminal, and the analysis data of each artificial intelligence feature discrimination model is also obtained.

[0134] 107. The artificial intelligence feature discrimination models are screened according to the probability distribution in the discrimination distribution data, and an identification password is generated according to the digital labels of the screened artificial intelligence feature discrimination models.

[0135] The terminal screens the artificial intelligence feature discrimination models according to the probability distribution in the discrimination distribution data, and then generates an identification password according to the digital labels of the screened artificial intelligence feature discrimination models. For example, the intelligent lockbox terminal first selects N feature discrimination models determined as "real image", and then sorts the digital labels corresponding to these N feature discrimination models to generate an identification password. In addition to sorting, other methods can also be used to generate the identification password.

[0136] Because the cabinet database stores AI feature discrimination models with different training effects, if the real biometric information input by the same manager during binding and current unlocking is used, the two biometric information will be added with the same fusion material features for feature fusion to the same degree, and the discrimination results obtained by inputting the same AI feature discrimination model will be the same. Anyway, if there are traces of AI processing, some AI feature discrimination models will have different discrimination results, which will cause slight differences in subsequent number passwords and lead to subsequent unlocking failures. Using multiple models with different training degrees and even different hierarchical structures can ensure that at least one model has a distinguishing effect on the traces of AI processing.

[0137] 108. Verify the target password, number password and password to be verified in the box database. When the verification is passed, the smart cash box is unlocked.

[0138] The smart cash box terminal compares the target password pre-stored in the box database, the calculated numbered password, and the password to be verified entered by the user. If they are the same, it means that the biometric information is consistent and there is no trace of AI processing.

[0139] In this embodiment, it is first necessary to obtain unlocking requests from multiple management terminals in the same time period. If there are not enough management terminals, the unlocking will not be granted directly. Next, the unlocking authority of the management terminal itself is detected. When all management terminals have unlocking authority, biometric information, fusion number information and password to be verified are collected through the management terminal, wherein the fusion number information and password to be verified are auxiliary unlocking data generated when binding the management terminal and the manager, and the manager needs to enter the corresponding fusion number information and password to be verified during verification. First, the collected biometric information is subjected to conventional identification verification. If the conventional verification fails, there is no need to continue. At this time, the corresponding fusion material is determined in the box database through the fusion number information of each management terminal. The fusion material is feature extracted through the preset artificial intelligence feature fusion model, and is fused into the biometric information according to the preset channel ratio and rules (the fusion rules of different biometric types are different) to generate feature fusion data. Next, the feature fusion data of each management terminal is input into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data. Then, the artificial intelligence feature discrimination model is screened according to the probability distribution in the discriminant distribution data, and the numbered password is generated according to the digital label of the screened artificial intelligence feature discrimination model. Finally, the target password, numbered password and password to be verified in the box database are verified. When the verification is passed, the smart cash box is unlocked.

[0140] Due to the differences in the parameters and structures of each artificial intelligence feature discrimination model, the analysis results of each artificial intelligence feature discrimination model for the same feature fusion data are different. The target password is the biological feature image or video that is first collected during the binding process of the administrator. Then, fusion materials are selected to fuse the biological feature image or video to a certain extent. Next, all artificial intelligence feature discrimination models are used to distinguish between true and false. The set of biological feature models determined to be true and the set of biological feature models determined to be false are obtained. Then, a numbered password is generated based on the digital labels of one of the types of biological feature models. Each administrator has a dedicated numbered password, and all the numbered passwords are integrated to form the target password. The numbered password and the fusion number information corresponding to the fusion materials are sent to the corresponding management terminal for the administrator to save. In the unlocking process, if the collected biological feature information is real, after feature fusion and discrimination, the set of biological feature models determined to be true and the set of biological feature models determined to be false are the same as those during binding, that is, the generated numbered passwords are also the same. On the contrary, if the collected biological feature information itself is AI-generated, after fusing the biological features of the fusion materials, the set of biological feature models determined to be true and the set of biological feature models determined to be false are not the same as those during binding, and the numbered passwords are also different, which improves the security of the intelligent safe.

[0141] Please refer to Figure 2 , an embodiment of a takeover method for an intelligent safe provided by this application includes:

[0142] 201. Receive takeover requests sent by two management terminals. The takeover requests include takeover terminal information of the two management terminals;

[0143] When the intelligent safe is included in the repository, it is necessary to re-bind the administrator (the corresponding management terminal). Whether it is to replace the administrator or simply change the storage location, it is necessary to re-bind the management terminal.

[0144] When the intelligent safe is transferred to another repository, the administrator controls the management terminal to interact with the intelligent safe through the interaction of the repository.

[0145] 202. Obtain the real-time positioning information of the box body, and check the inbound and outbound records with the repository server through the real-time positioning information of the box body, and obtain the key;

[0146] After the intelligent safe terminal receives the takeover request, it first performs real-time positioning to generate the real-time positioning information of the box body, and determines the nearest repository server to the current location through the real-time positioning information of the box body. Then, it interacts with the repository server through the link information in the box database to obtain and confirm the inbound and outbound records to prevent out-of-library takeover. And by the way, the corresponding key is obtained.

[0147] 203. When the inbound and outbound records meet the conditions, obtain the encrypted takeover terminal verification information from the box database. The encrypted takeover terminal verification information is the takeover terminal information pre-entered into the box database by the previous repository server.

[0148] When the inbound and outbound records meet the conditions, the intelligent cash box terminal obtains the encrypted takeover terminal verification information from the box database. This encrypted takeover terminal verification information is the information entered before the intelligent cash box was shipped out of the previous repository, that is, the information of the next two management terminals (encrypted takeover terminal verification information) is entered. For example, if two management terminals B and C of Bank A are about to take over the intelligent cash box next month, the takeover terminal verification information of management terminal B needs to be encrypted and transmitted to the current management terminal D. After the current management terminal D unlocks the intelligent cash box, the encrypted takeover terminal verification information is stored in the box database of the intelligent cash box. Then, the takeover terminal verification information of management terminal C is encrypted and transmitted to the current management terminal E. After the current management terminal E unlocks the intelligent cash box, the encrypted takeover terminal verification information is stored in the box database of the intelligent cash box.

[0149] 204. Decrypt the encrypted takeover terminal verification information with the key, and verify the takeover terminal information of the two management terminals with the decrypted takeover terminal verification information.

[0150] The terminal decrypts the encrypted takeover terminal verification information with the key transmitted by the repository server. The key is transmitted from the management terminal to the repository server. Next, the intelligent cash box terminal verifies the takeover terminal information of the two management terminals with the decrypted takeover terminal verification information.

[0151] 205. When the verification passes, determine that the two management terminals are the identities of the new managers, generate a biometric collection instruction for each of the two management terminals, and send the corresponding biometric collection instructions to the two management terminals respectively.

[0152] When the verification passes, the intelligent cash box terminal determines that the two management terminals are the identities of the new managers, and then generates a biometric collection instruction for each of the two management terminals, and sends the corresponding biometric collection instructions to the two management terminals respectively. The biometric collection instruction can allow the manager to select the biometric to be input, or allow the current manager to select. It should be noted that the current manager can input at least one biometric.

[0153] 206. Obtain the first biometric information sent by the two management terminals. The first biometric information is biometric information of the image collection type.

[0154] The intelligent box terminal obtains the first biometric information sent by two management terminals. The first biometric information is biometric information of the image acquisition type. In this embodiment, only biometric information of the image type is targeted, such as face images, fingerprint images, and pupil images, etc. These types of biometric features all have one characteristic. If there are traces of AI processing, the face area of the face image is prone to differences from the environmental area, while the areas of fingerprint images and pupil images are relatively small, the collected content is complete, and the area of the AI processing trace is obvious.

[0155] 207. Determine a number of artificial intelligence feature discrimination models and a number of artificial intelligence feature fusion models from the box database, and select corresponding types of fusion materials for the two management terminals respectively from the fusion material library of the box database;

[0156] In this embodiment, the intelligent box terminal determines a number of artificial intelligence feature discrimination models and a number of artificial intelligence feature fusion models from the box database, and selects corresponding types of fusion materials for the two management terminals respectively from the fusion material library of the box database. First, select a fusion material of the same type according to the type of biometric information.

[0157] 208. Input the first biometric information and the corresponding fusion material into the artificial intelligence feature fusion model to add the features of the fusion material to the first biometric information, and generate feature fusion information for the two management terminals respectively;

[0158] The intelligent box terminal inputs the first biometric information and the corresponding fusion material into the artificial intelligence feature fusion model to add the features of the fusion material to the first biometric information, and generate feature fusion information for the two management terminals respectively. Different types of biometric features use different types of artificial intelligence feature fusion models and adopt different feature fusion methods for a certain degree of feature fusion. For example: for face images, use specific face fusion materials to perform feature fusion at the face edge and specific face areas (areas prone to AI processing).

[0159] 209. Input the two pieces of feature fusion information into a number of artificial intelligence feature discrimination models for discrimination respectively, and generate a first discrimination result and a second discrimination result;

[0160] The intelligent box terminal inputs the two feature fusion information into several artificial intelligence feature discrimination models respectively for discrimination, generating a first discrimination result and a second discrimination result. The feature fusion information of management terminal A is input into several artificial intelligence feature discrimination models respectively for discrimination, determining the artificial intelligence feature discrimination models identified as real and those identified as false, and generating the first discrimination result based on this. Then, the feature fusion information of management terminal B is input into several artificial intelligence feature discrimination models respectively for discrimination, determining the artificial intelligence feature discrimination models identified as real and those identified as false, and generating the second discrimination result based on this.

[0161] 210. Respectively determine the digital labels corresponding to the artificial intelligence feature discrimination models that meet the conditions according to the probability distributions in the first discrimination result and the second discrimination result, and generate two numbered passwords based on the two groups of digital labels. The two numbered passwords are used as the target passwords for unlocking the intelligent box.

[0162] The terminal respectively determines the digital labels corresponding to the artificial intelligence feature discrimination models that meet the conditions according to the probability distributions in the first discrimination result and the second discrimination result, and generates two numbered passwords based on the two groups of digital labels. The two numbered passwords are used as the target passwords for unlocking the intelligent box, that is, generate a numbered password for management terminal A through the digital label corresponding to the artificial intelligence feature discrimination model with the result of "real" in the first discrimination result (or generate a numbered password for management terminal A using the digital label corresponding to the artificial intelligence feature discrimination model with the result of "false"). Similarly, generate a numbered password for management terminal B.

[0163] 211. Use the fusion numbering information of the two fusion materials and the two numbered passwords to generate corresponding takeover success information for the two management terminals respectively, and send the takeover success information to the two management terminals respectively.

[0164] The terminal uses the fusion numbering information of the two fusion materials and the two numbered passwords to generate corresponding takeover success information for the two management terminals respectively, and send the takeover success information to the two management terminals respectively.

[0165] The above solution can make the user's numbered password unique. Non - managers need to know the numbered password, the fusion numbering information, and the type of biometric information input by the real manager. Even if they know this information, they cannot pass the same - type detection links of several biometric recognition models because different biometric recognition models have different training degrees for different regions of biometric features. For example, some only perform recognition training on the nose in the face area, some only on the ears, and there are multiple biometric recognition models trained for a certain area, with different training degrees and different specificities for each model, increasing the detection effect of AI processing traces.

[0166] Please refer to Figure 3 , an embodiment of a method for determining fusion materials provided by this application includes:

[0167] 301. Input the first biometric information sent by two management terminals into a number of artificial intelligence feature discrimination models respectively for discrimination, and generate a third discrimination result and a fourth discrimination result;

[0168] 302. Conduct discrimination probability analysis on the first discrimination result, the second discrimination result, the third discrimination result, and the fourth discrimination result;

[0169] 303. When the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are the same, it is determined that the fusion materials meet the standards;

[0170] 304. When the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are different, re-select new fusion materials and continue with feature fusion and model discrimination.

[0171] In this embodiment, the intelligent safe terminal inputs the first biometric information sent by two management terminals into a number of artificial intelligence feature discrimination models respectively for discrimination, and generates a third discrimination result and a fourth discrimination result. The purpose is to identify real images. Then, conduct discrimination probability analysis on the first discrimination result, the second discrimination result, the third discrimination result, and the fourth discrimination result. When the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are the same, it is determined that the fusion materials meet the standards. If the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are different, re-select new fusion materials and continue with feature fusion and model discrimination. That is, the biometric information integrating the features of the fusion materials and the original biometric information need to have the same detection results on each artificial intelligence feature discrimination model. In this way, the degree of fusion of the fused feature information will not affect the real image, but only play a prominent role in the images that have undergone AI processing, improving the subsequent detection accuracy.

[0172] Please refer to Figure 4 , an embodiment of a method for verifying the unlocking environment provided by this application includes:

[0173] 401. Obtain the real-time positioning information of the box body, and determine the location information and link establishment information of the current storage server from the box body database;

[0174] 402. When the real-time positioning information of the box body meets the location conditions with the location information of the current storage server, establish a link with the storage server through the link establishment information and obtain the inbound and outbound records;

[0175] 403. When the in-and-out records show that the smart cash box has not been out of the warehouse, the unlocking position information reserved by two managers is obtained from the box database;

[0176] 404. Analyze the current location information of the two management terminals and the unlocking location information reserved by the two managers;

[0177] 405. When the location analysis result meets the conditions, it is determined that the smart cash box preliminarily meets the unlocking verification conditions.

[0178] In this embodiment, after obtaining the real-time location information of the box, the smart cash box terminal needs to determine the location information and link establishment information of the current storage library server from the box database. Both of these information are imported into the box storage library by the original storage library server before being transferred out of the previous storage library. When the real-time location information of the box and the location information of the current storage library server meet the location conditions, that is, when the location difference is less than the preset value, the smart cash box terminal can establish a link with the storage library server through the link establishment information and obtain the storage entry and exit records. When the storage entry and exit records show that the smart cash box has not been shipped out, that is, the smart cash box is still in the storage library, indicating that the smart cash box is in a state where the unlocking link can be performed. At this time, the smart cash box terminal obtains the unlocking location information reserved by the two managers from the box database, and then analyzes the current location information of the two management terminals and the unlocking location information reserved by the two managers. When the location analysis result meets the conditions, it is determined that the smart cash box preliminarily meets the unlocking verification conditions. That is, confirm the location of the smart cash box and whether the smart cash box is still inside the storage library, and then confirm whether the location of the management terminal is operated remotely. If both meet the conditions required for the unlocking link, the next step of inspection is performed, which greatly increases security.

[0179] See also Figure 5 The present application provides an embodiment of a method for handing over a smart cash box, including:

[0180] 501. Receive handover requests sent by two management terminals respectively, where the handover requests include encrypted takeover terminal authentication information of the two new management terminals, and the handover requests also include location information and link establishment information of the new repository server;

[0181] 502. storing the encrypted takeover end verification information, the location information of the new repository server and the link establishment information in the cabinet database;

[0182] 503. Receive the outbound notification sent by the storage repository server and perform handover and block processing on the smart cash box.

[0183] In this embodiment, after the intelligent box terminal unlocks, it receives handover requests sent by two management terminals respectively, indicating that handover processing needs to be carried out next. The handover requests include encrypted takeover end verification information of two new management terminals, and the handover requests also include the location information and link establishment information of the new repository server. These information are sent to the management terminal and the repository server by the takeover party within a certain period of time before takeover, and have been verified by the management terminal and the repository server. Next, the intelligent box stores the encrypted takeover end verification information, the location information of the new repository server, and the link establishment information in the box database. Finally, after the intelligent box terminal receives the outbound notice sent by the repository server, the intelligent box terminal performs handover lockdown processing on the intelligent box, that is, the intelligent box receives the takeover request transmitted through a specific channel (the channel set by the repository server). This embodiment, Figure 1 The embodiment shown, Figure 2 The embodiment shown forms a simple closed loop.

[0184] Please refer to Figure 6 , this application provides an embodiment of a method for interacting with a repository server, including:

[0185] 601. Obtain the real-time positioning information of the box, and perform position analysis on the real-time positioning information of the box and the location information of the repository server;

[0186] 602. When the position analysis is in line, establish an interaction channel with the repository server through the link establishment information of the repository server;

[0187] 603. Send the inbound / outbound record and key acquisition request to the repository server;

[0188] 604. Receive the inbound / outbound record and key sent by the repository server, and verify the inbound / outbound record.

[0189] In this embodiment, the terminal obtains the real-time positioning information of the box, and performs position analysis on the real-time positioning information of the box and the location information of the repository server. Only when the position analysis is in line, the intelligent box terminal will establish an interaction channel with the repository server through the link establishment information of the repository server. Next, the intelligent box terminal sends the inbound / outbound record and key acquisition request to the repository server so that the repository server sends the inbound record and key. When the intelligent box terminal receives the inbound / outbound record and key sent by the repository server, it verifies the inbound / outbound record.

[0190] Please refer to Figure 7 , this application provides an embodiment of a multi-user unlocking control device for an intelligent box, including:

[0191] The first receiving unit 701 is configured to receive takeover requests sent by two management terminals, where the takeover requests include takeover terminal information of the two management terminals;

[0192] The third obtaining unit 702 is configured to obtain real-time positioning information of the box body, verify the inbound and outbound records with the repository server through the real-time positioning information of the box body, and obtain a key;

[0193] Optionally, the third obtaining unit 702 includes:

[0194] Obtain the real-time positioning information of the box body, and perform position analysis on the real-time positioning information of the box body and the position information of the repository server;

[0195] When the position analysis is in line, establish an interactive channel with the repository server through the link establishment information of the repository server;

[0196] Send an inbound and outbound record and a key acquisition request to the repository server;

[0197] Receive the inbound and outbound records and the key sent by the repository server, and verify the inbound and outbound records.

[0198] The fourth obtaining unit 703 is configured to, when the inbound and outbound records meet the conditions, obtain encrypted takeover end verification information from the box body database, where the encrypted takeover end verification information is the takeover terminal information pre-entered into the box body database by the previous repository server;

[0199] The decryption unit 704 is configured to decrypt the encrypted takeover end verification information with the key, and verify the takeover terminal information of the two management terminals with the decrypted takeover end verification information;

[0200] The second determination unit 705 is configured to, when the verification is passed, determine that the two management terminals are new manager identities, generate a biometric collection instruction for each of the two management terminals, and send the corresponding biometric collection instruction to the two management terminals respectively;

[0201] The fifth obtaining unit 706 is configured to obtain first biometric information sent by the two management terminals, where the first biometric information is biometric information of the image acquisition type;

[0202] The third determination unit 707 is configured to determine a number of artificial intelligence feature discrimination models and a number of artificial intelligence feature fusion models from the box body database, and respectively select corresponding types of fusion materials for the two management terminals from the fusion material library of the box body database;

[0203] The fusion unit 708 is configured to input the first biometric information and the corresponding fusion material into the artificial intelligence feature fusion model, add the features of the fusion material to the first biometric information, and generate feature fusion information for the two management terminals respectively;

[0204] A fourth generation unit 709, configured to respectively input two pieces of feature fusion information into a plurality of artificial intelligence feature discrimination models for discrimination, and generate a first discrimination result and a second discrimination result;

[0205] A sixth generation unit 710, configured to respectively input the first biometric information sent by two management terminals into a plurality of artificial intelligence feature discrimination models for discrimination, and generate a third discrimination result and a fourth discrimination result;

[0206] A first analysis unit 711, configured to perform discrimination probability analysis on the first discrimination result, the second discrimination result, the third discrimination result, and the fourth discrimination result;

[0207] A fifth determination unit 712, configured to determine that the fusion material meets the standard when the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are the same;

[0208] A selection unit 713, configured to re-select new fusion materials to continue feature fusion and model discrimination when the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are different;

[0209] A fifth generation unit 714, configured to respectively determine digital labels corresponding to artificial intelligence feature discrimination models that meet the conditions according to the probability distributions in the first discrimination result and the second discrimination result, and generate two numbered passwords based on the two groups of digital labels. The two numbered passwords are used as the target passwords for unlocking the intelligent safe;

[0210] A sending unit 715, configured to generate corresponding takeover success messages for two management terminals by using the fusion number information of two pieces of fusion materials and the two numbered passwords, and respectively send the takeover success messages to the two management terminals;

[0211] A first acquisition unit 716, configured to acquire unlocking requests respectively sent by two management terminals, where the unlocking requests include the terminal label information of the two management terminals;

[0212] A sixth acquisition unit 717, configured to acquire real-time positioning information of the box body, and determine the position information and link establishment information of the current repository server from the box body database;

[0213] A seventh acquisition unit 718, configured to establish a link with the repository server through the link establishment information and acquire inbound and outbound records when the real-time positioning information of the box body meets the position condition with the position information of the current repository server;

[0214] An eighth acquisition unit 719, configured to acquire the unlocking position information reserved by two managers from the box body database when the inbound and outbound records show that the intelligent safe has not been shipped out;

[0215] A second analysis unit 720, configured to analyze the current location information of each of the two management terminals and the unlocking location information reserved by the two managers;

[0216] A fourth determination unit 721, configured to determine that the intelligent money box preliminarily meets the unlocking verification condition when the location analysis result meets the conditions;

[0217] An inspection unit 722, configured to inspect the terminal tag information of each of the two management terminals;

[0218] A second acquisition unit 723, configured to acquire the biometric information, fusion number information, and password to be verified sent by each of the two management terminals when the tag inspections are all qualified;

[0219] A first determination unit 724, configured to determine the corresponding artificial intelligence feature fusion model and fusion materials from the box database according to the fusion number information of each management terminal. A number of artificial intelligence feature discrimination models are also set in the box database, and each artificial intelligence feature discrimination model has a different digital label. Each artificial intelligence feature discrimination model is an artificial intelligence model that has completed different degrees of pre-training on multiple biometric data;

[0220] A first generation unit 725, configured to perform feature fusion processing on the fusion materials and the corresponding biometric information through the artificial intelligence feature fusion model to generate feature fusion data;

[0221] A second generation unit 726, configured to input the feature fusion data of each management terminal into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data;

[0222] A third generation unit 727, configured to screen the artificial intelligence feature discrimination models according to the probability distribution in the discrimination distribution data, and generate a numbered password according to the digital label of the screened artificial intelligence feature discrimination model;

[0223] A verification unit 728, configured to verify the target password, numbered password, and password to be verified in the box database, and perform unlocking processing on the intelligent money box when the verification is passed;

[0224] A second receiving unit 729, configured to receive the handover requests sent by each of the two management terminals. The handover requests include the encrypted takeover end verification information of the two new management terminals. The handover requests also include the location information and link establishment information of the new storage repository server;

[0225] A storage unit 730, configured to store the encrypted takeover end verification information, the location information of the new storage repository server, and the link establishment information into the box database;

[0226] A third receiving unit 731, configured to receive the outbound notice sent by the repository server and perform handover and locking processing on the intelligent cash box.

[0227] Please refer to Figure 8 , this application provides a multi-user unlocking control device for an intelligent cash box, including:

[0228] A processor 801, a memory 802, an input / output unit 803, and a bus 804.

[0229] The processor 801 is connected to the memory 802, the input / output unit 803, and the bus 804.

[0230] The memory 802 stores a program, and the processor 801 calls the program to execute the methods as described in Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 and Figure 6 .

[0231] This application provides a computer-readable storage medium, on which a program is stored. When the program is executed on a computer, it executes the methods as described in Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 and Figure 6 .

[0232] 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.

[0233] 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 the units is only a logical function division, and there can 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 is that the couplings, direct couplings, or communication connections shown or discussed with each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0234] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or 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.

[0235] In addition, each functional unit in various embodiments of the present application may be integrated in a processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0236] 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 such an 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. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, read-only memory), random access memories (RAM, random access memory), magnetic disks, or optical discs and other various media that can store program codes.

[0237] 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" involved in this disclosure are obtained under full authorization.

Claims

1. A multi-user unlocking control method for an intelligent box, characterized in that, Including: Obtain the unlocking requests sent by two management terminals respectively, where the unlocking requests include the terminal tag information of the two management terminals respectively; Examine the terminal tag information of the two management terminals respectively; When the tag examinations are all in line, obtain the biometric information, fusion number information, and password to be verified sent by the two management terminals respectively; Determine the corresponding artificial intelligence feature fusion model and fusion materials from the box database according to the fusion number information of each management terminal. There are also several artificial intelligence feature discrimination models set in the box database. Each artificial intelligence feature discrimination model has a different digital tag, and each artificial intelligence feature discrimination model is an artificial intelligence model that has completed different degrees of pre-training on various biometric data; Perform feature fusion processing on the fusion materials and the corresponding biometric information through the artificial intelligence feature fusion model to generate feature fusion data; Input the feature fusion data of each management terminal into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data; Screen the artificial intelligence feature discrimination models according to the probability distribution in the discrimination distribution data, and generate a numbered password according to the digital tags of the screened artificial intelligence feature discrimination models; Verify the target password, the numbered password, and the password to be verified pre-stored in the box database. When the verification passes, perform unlocking processing on the intelligent safe.

2. The multi-user unlocking control method according to claim 1, wherein, Before the step of obtaining the unlocking requests sent by the two management terminals respectively, the multi-user unlocking control method further includes: Receive the takeover requests sent by the two management terminals, where the takeover requests include the takeover terminal information of the two management terminals; Obtain the real-time positioning information of the box, and check the inbound and outbound records through the real-time positioning information of the box to the repository server, and obtain the key; When the inbound and outbound records meet the conditions, obtain the encrypted takeover terminal verification information from the box database. The encrypted takeover terminal verification information is the previous takeover terminal information pre-entered into the box database by the repository server; Decrypt the encrypted takeover terminal verification information through the key, and verify the takeover terminal information of the two management terminals with the decrypted takeover terminal verification information; When the verification passes, determine that the two management terminals are new manager identities, generate a biometric collection instruction for each of the two management terminals, and send the corresponding biometric collection instructions to the two management terminals respectively; Obtain the first biometric information sent by both management terminals. The first biometric information is biometric information of the image collection type; Determine several artificial intelligence feature discrimination models and several artificial intelligence feature fusion models from the box database, and select corresponding types of fusion materials for the two management terminals respectively from the fusion material library of the box database; Input the first biometric information and the corresponding fusion materials into the artificial intelligence feature fusion model, add the features of the corresponding fusion materials to the first biometric information, and generate feature fusion information for the two management terminals respectively; Input the two pieces of feature fusion information into several artificial intelligence feature discrimination models for discrimination respectively to generate a first discrimination result and a second discrimination result; Determine the digital tags corresponding to the eligible artificial intelligence feature discrimination models respectively according to the probability distributions in the first discrimination result and the second discrimination result, and generate two numbered passwords based on the two sets of digital tags. The two numbered passwords are used as the target passwords for unlocking the intelligent safe box. Use the two fusion numbering information corresponding to the respectively selected fusion materials of the corresponding types and the two numbered passwords to generate the corresponding takeover success information for the two management terminals respectively, and send the takeover success information to the two management terminals respectively.

3. The multi-user unlocking control method according to claim 2, characterized in that After the step of respectively inputting the two feature fusion information into a number of artificial intelligence feature discrimination models for discrimination to generate the first discrimination result and the second discrimination result, and before the step of respectively determining the digital tags corresponding to the eligible artificial intelligence feature discrimination models according to the probability distributions in the first discrimination result and the second discrimination result, and generating two numbered passwords based on the two sets of digital tags, the multi-user unlocking control method further includes: Respectively input the first biometric information sent by both management terminals into a number of artificial intelligence feature discrimination models for discrimination to generate the third discrimination result and the fourth discrimination result; Conduct discrimination probability analysis on the first discrimination result, the second discrimination result, the third discrimination result and the fourth discrimination result; When the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are the same, it is determined that the fusion material meets the standard; When the discrimination probability analysis shows that the discrimination results of the first biometric information and its corresponding feature fusion information are different, re-select new fusion materials and continue with feature fusion and model discrimination.

4. The multi-user unlocking control method according to claim 2, characterized in that The unlocking request further includes the current location information of each of the two management terminals; After the step of obtaining the unlocking requests sent by each of the two management terminals, and before the step of verifying the terminal tag information of each of the two management terminals, the multi-user unlocking control method further includes: Obtain the real-time positioning information of the box body, and determine the location information and link establishment information of the current storage server from the box body database; When the real-time positioning information of the box body meets the location condition with the location information of the current storage server, establish a link with the storage server through the intelligent safe box according to the link establishment information and obtain the inbound and outbound records; When the inbound and outbound records show that the intelligent safe box has not been shipped out, obtain the unlocking location information reserved by the two managers from the box body database; Analyze the current location information of each of the two management terminals and the unlocking location information reserved by the two managers; When the location analysis result meets the conditions, it is determined that the intelligent safe box preliminarily meets the unlocking verification conditions.

5. The multi-user unlocking control method according to claim 4, characterized in that, After the step of verifying the target password, the numbered password and the password to be verified in the box body database, and when the verification is passed, performing an unlocking process on the intelligent safe box, the multi-user unlocking control method further includes: Receive the handover requests sent by two management terminals respectively, where the handover requests include the encrypted takeover terminal verification information of two new management terminals, and the handover requests further include the new location information and new link establishment information in the repository server; Store the encrypted takeover terminal verification information, the new location information in the repository server, and the new link establishment information in the box database; Receive the outbound notice sent by the repository server, and perform handover lockdown processing on the intelligent cash box.

6. The multi-user unlocking control method according to claim 5, wherein The steps of obtaining the real-time positioning information of the box, verifying the inbound and outbound records through the real-time positioning information of the box, and obtaining the key include: Obtain the real-time positioning information of the box, and perform position analysis on the real-time positioning information of the box and the location information of the repository server; When the position analysis is in line, use the intelligent cash box to establish an interaction channel with the repository server through the link establishment information of the repository server; Send an inbound and outbound record and a key acquisition request to the repository server; Receive the inbound and outbound record and the key sent by the repository server, and verify the inbound and outbound record.

7. A multi-user unlocking control device for an intelligent box, characterized in that, Include: A first acquisition unit, configured to acquire the unlocking requests sent by two management terminals respectively, where the unlocking requests include the terminal label information of the two management terminals respectively; An inspection unit, configured to inspect the terminal label information of the two management terminals respectively; A second acquisition unit, configured to acquire the biometric information, fusion number information, and password to be verified sent by the two management terminals respectively when the label inspections are all in line; A first determination unit, configured to determine the corresponding artificial intelligence feature fusion model and fusion materials from the box database according to the fusion number information of each management terminal. There are also several artificial intelligence feature discrimination models set in the box database, each artificial intelligence feature discrimination model has a different digital label, and each artificial intelligence feature discrimination model is an artificial intelligence model that has completed different degrees of pre-training on multiple biometric data; A first generation unit, configured to perform feature fusion processing on the fusion materials and the corresponding biometric information through the artificial intelligence feature fusion model to generate feature fusion data; A second generation unit, configured to input the feature fusion data of each management terminal into each artificial intelligence feature discrimination model for detection to generate discrimination distribution data; A third generation unit, configured to screen the artificial intelligence feature discrimination models according to the probability distribution in the discrimination distribution data, and generate a numbered password according to the digital labels of the screened artificial intelligence feature discrimination models; A verification unit, configured to verify the target password, the numbered password, and the password to be verified pre-stored in the box database. When the verification passes, perform unlocking processing on the intelligent cash box.

8. The multi-user unlocking control device according to claim 7, characterized in that The multi-user unlocking control device further includes: A first receiving unit, configured to receive the takeover requests sent by two management terminals, where the takeover requests include the takeover terminal information of the two management terminals; A third acquisition unit, configured to acquire the real-time positioning information of the box, verify the inbound and outbound records through the real-time positioning information of the box, and acquire the key; A fourth acquisition unit, configured to obtain encrypted takeover terminal verification information from the box database when the inbound and outbound record meets the conditions, where the encrypted takeover terminal verification information is the previous takeover terminal information pre-input by the repository server into the box database; A decryption unit, configured to decrypt the encrypted takeover terminal verification information with the key, and verify the takeover terminal information of the two management terminals with the decrypted takeover terminal verification information; A second determination unit, configured to determine that the two management terminals have the identity of new managers when the verification is passed, generate a biometric collection instruction for each of the two management terminals, and send the corresponding biometric collection instructions to the two management terminals respectively; A fifth acquisition unit, configured to obtain first biometric information sent by both of the two management terminals, where the first biometric information is biometric information of the image acquisition type; A third determination unit, configured to determine a plurality of artificial intelligence feature discrimination models and a plurality of artificial intelligence feature fusion models from the box database, and respectively select corresponding types of fusion materials for the two management terminals from the fusion material library of the box database; A fusion unit, configured to input the first biometric information and the corresponding fusion material into the artificial intelligence feature fusion model, add the features of the corresponding fusion material to the first biometric information, and generate feature fusion information for the two management terminals respectively; A fourth generation unit, configured to input the two pieces of feature fusion information into a plurality of artificial intelligence feature discrimination models for discrimination respectively, and generate a first discrimination result and a second discrimination result; A fifth generation unit, configured to respectively determine digital labels corresponding to the artificial intelligence feature discrimination models that meet the conditions according to the probability distributions in the first discrimination result and the second discrimination result, and generate two numbered passwords according to the two groups of digital labels, where the two numbered passwords are used as the target passwords for unlocking the intelligent safe; A sending unit, configured to generate corresponding takeover success information for the two management terminals respectively by using the two fusion number information corresponding to the respectively selected corresponding types of fusion materials and the two numbered passwords, and send the takeover success information to the two management terminals respectively.

9. A multi-user unlocking control device for an intelligent box, characterized in that, Comprising: A processor, a memory, an input / output unit, and a bus; The processor is connected to the memory, the input / output unit, and the bus; The memory stores a program, and the processor calls the program to execute the multi-user unlocking control method according to any one of claims 1 to 6.

10. 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 multi-user unlocking control method according to any one of claims 1 to 6.

Images