Banking business library safety control system based on face recognition and dynamic coded lock

Abstract

The invention discloses a bank business library security control system based on face recognition and a dynamic coded lock, relates to the technical field of bank security, and solves the technical problems that in the prior art, password data are not dynamically updated, are easy to break through and are low in security. According to the method, firstly, a face image is recognized, if face recognition fails, verification is carried out again, and the number of recognition times cannot exceed the preset number of times; if the face recognition succeeds, acquiring an emotion type of the operator from the face image, and if the emotion is normal, performing preliminary verification successfully; after the preliminary verification is successful, inserting a name character of an operator and time for applying to open a service library into the initial password, randomly generating a dynamic password, correcting the dynamic password, and opening the service library if the verification is successful; if not, regenerating the dynamic password; through combination of face recognition and dynamic passwords, a powerful multi-layer protection system is formed, and the security level of service library access is remarkably improved.

Description

Technical Field

[0001] The invention belongs to the technical field of bank security, and in particular is a bank business library security control system based on face recognition and dynamic password lock. Background Art

[0002] The fourth type of bank business warehouse is an intelligent special cabinet used to store bank balance boxes. According to the size of the approved cash inventory, the business warehouse is divided into one, two, three and four categories. The cash inventory of the first type of warehouse is 50 million yuan (inclusive) or more; the cash inventory of the second type of warehouse is 10 million yuan (inclusive) to 50 million yuan; the cash inventory of the third type of warehouse is 800,000 yuan (inclusive) to 10 million yuan; the cash inventory of the fourth type of warehouse is less than 800,000 yuan.

[0003] The invention patent application with publication number CN113674468A discloses a security management system based on a financial venue business library, which includes an input module and a processing module. The input module includes an image unit, a fingerprint unit and a key unit. The input module is used to input user data information, and the processing module is used to receive and process the data information input by the input module.

[0004] The above-mentioned prior art image data, fingerprint data and password data set multiple anti-theft protections for the business library, but the password data is not dynamically updated, which is easy to be hacked and has low security. Therefore, the present invention provides a bank business library security control system based on face recognition and dynamic password lock. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a banking business library security control system based on face recognition and dynamic password lock, which is used to solve the technical problems in the prior art that password data is not dynamically updated, is easily cracked, and has low security.

[0006] To achieve the above-mentioned object, the first aspect of the present invention provides a banking service library security control system based on face recognition and dynamic password lock, including a face recognition module and a dynamic password module;

[0007] Face recognition module: used to identify the identity information of the operator and conduct preliminary verification; the identity information includes the face image to be identified and the emotion to be identified;

[0008] Dynamic password module: After the initial verification is successful, it is used to randomly insert the operator information on the basis of the initial password to generate a dynamic password for final verification; the operator information includes the operator's name and the time of applying to open the business database.

[0009] Preferably, the facial image is a facial image of an operator who opens the service library, and is captured by a camera at the service library.

[0010] Preferably, the identification information of the operator includes:

[0011] Extract the features of the face image to be identified, compare them with the features of the face images in the bank's back-end database, and determine whether there is a matching face image in the bank's back-end database; if not, the face image recognition fails; if yes, the face image recognition succeeds; and,

[0012] The successfully recognized face image is input into the emotion recognition module, which outputs the emotion vector to analyze the operator's emotion type. If the emotion type is positive, the operator is in a normal mood and the initial verification is successful. If the emotion type is negative, the operator is in an abnormal mood and the initial verification fails. The emotion recognition module is constructed by a CNN neural network model.

[0013] The present invention first identifies the face image to determine whether the operator is a bank staff member, and performs the first step of early warning for security control; then extracts emotion-related features from the face image, identifies the emotion vector, and analyzes the current emotional state of the operator, such as extreme tension, anxiety or anger, which may indicate that the ongoing operation is not based on normal intention, or that the operator is coerced; in this case, the system can temporarily prevent the execution of the transaction to avoid potential risks caused by erroneous operations or forced operations under the influence of emotions.

[0014] Preferably, when face verification fails, re-verification will be performed. If the number of failures exceeds a preset number, the bank monitoring terminal will display an abnormality and promptly alarm; and when the emotions are abnormal, the bank monitoring terminal will display an abnormality and promptly alarm.

[0015] Preferably, based on the emotion vector, analyzing the emotion type of the operator includes:

[0016] The score difference PFC between positive emotions and negative emotions is calculated by the formula PFC=ΣQi-ΣWj; wherein Qi is the score of each emotion type in positive emotions, Wj is the score of each emotion type in negative emotions, i is the positive emotion type, i=0, 1, ..., N; j is the negative emotion type, j=0, 1, ..., M; M, N are positive integers;

[0017] If the score difference is 0, there is no positive or negative emotion expression, and the operator's emotion is normal;

[0018] If the score difference is greater than 0, it indicates positive emotion and the operator’s emotion is normal;

[0019] If the score difference is less than 0, it indicates negative emotion and the operator’s emotion is abnormal.

[0020] The present invention achieves a comprehensive and integrated assessment of an individual's emotional state by simultaneously considering the scores of multiple positive and negative emotion types, thereby avoiding the one-sidedness that may be caused by a single emotion indicator; analyzing the differences between emotion types facilitates an intuitive understanding of whether the operator's emotional tendency is positive or negative, thereby achieving preliminary verification.

[0021] Preferably, randomly inserting operator information on the basis of the initial password to generate a dynamic password includes:

[0022] The operator's name characters and the time of applying to open the business library are randomly inserted into the initial password and converted into binary to obtain a binary password; the binary password is divided into several groups according to the preset number of digits of the password to obtain several binary groups, and each group is converted into decimal to obtain a dynamic password; wherein, each application to open the business library will regenerate a dynamic password.

[0023] Each dynamic password of the present invention is embedded with the operator's name and the time of applying to open the business library, which means that the validity period of the password is strictly limited. Once a specific time point has passed, the password will become invalid. This mechanism can effectively prevent password reuse or delay attacks, ensuring that each access is immediate and secure; even if an attacker obtains a one-time password information through phishing or other means, he cannot use it at other times or impersonate other operators, thereby improving the security level of system access.

[0024] Preferably, the order of the name characters in the English alphabet is extracted, and the order is converted into binary to obtain the binary form of the name characters.

[0025] Preferably, the binary password is divided into several groups according to the number of bits set by the password, including:

[0026] The number of bits in the binary password is compared with the number of bits in the initial password. If the ratio is not equal to 1, it means that there are redundant or missing bits in the binary code. The redundant parts are deleted, and the missing parts are filled with 0.

[0027] The present invention divides the binary password into several binary groups by comparing the ratio between the number of bits of the binary password and the number of bits of the initial password, directly deletes the redundant binary bits or fills the missing parts with 0, thereby ensuring that all generated binary passwords meet the format requirements of the initial password. The system can process these passwords in a more unified and standardized manner, reducing processing errors caused by different formats.

[0028] Preferably, if the decimal value of the binary group is greater than 10, the decimal value is rounded from the last digit to the first digit until one digit is retained.

[0029] The present invention limits the decimal value to a single-digit range by rounding off from the last digit, thereby ensuring consistency and standardization with the initial password; at the same time, further obfuscation processing is performed on the password, which increases the randomness and unpredictability of the password to a certain extent. Therefore, even similar original information may produce different password results after such processing, further improving the security of the password system.

[0030] Preferably, the control method of the banking service library security control system based on face recognition and dynamic password lock includes:

[0031] Step 1: Collect the operator's face image;

[0032] Step 2: Compare the face image with the face image in the bank's back-end database to determine whether the face image exists in the bank's back-end database; if yes, the face recognition is successful and proceed to step 3; if not, the face recognition fails and returns to step 1. When the number of face recognition failures exceeds the preset number, the bank monitoring terminal displays an abnormality and issues an alarm in time;

[0033] Step 3: After successful face recognition, the operator's emotion type is obtained from the face image. If the emotion is normal, the preliminary verification is successful and the process goes to step 4. If the emotion is abnormal, the bank monitoring terminal will display the abnormality and issue an alarm in time.

[0034] Step 4: After the initial verification is successful, insert the operator's name and the time of applying to open the business database into the initial password to randomly generate a dynamic password;

[0035] Step 5: The operator enters the dynamic password. If the password verification is correct, the business library will be opened; if the password verification is incorrect, return to step 4.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] 1. The present invention uses facial images to perform initial identity verification on operators, and then performs emotion recognition; this preliminarily ensures that only authorized personnel can access the business database, effectively preventing unauthorized access or fraud; further, through emotional reflection, it can be determined whether the current operator is under external coercion, etc., to avoid potential risks; through the combination of the two, a more comprehensive initial security protection system is built for the bank business database;

[0038] 2. After the initial verification is successful, the present invention verifies the dynamic password; by randomly inserting the operator's name and the time of applying to open the business library into the initial password, the initial password is made more unique; and the initial password after random insertion is converted between binary and decimal to obtain a dynamic password; each time the dynamic password is generated, it contains the operator's personalized information (name) and instant information (application time), so that each access password is unique, which greatly increases the difficulty of password cracking or reuse and enhances system security;

[0039] 3. The present invention compares the face image with the face image in the bank's back-end database to determine whether the face image exists in the bank's back-end database; if not, the face recognition fails and re-verification is required. When the number of face recognition failures exceeds the preset number, the bank monitoring terminal displays an abnormality and promptly alarms; if yes, the face recognition is successful, and the operator's emotion type is obtained from the face image. If the emotion is normal, the preliminary verification is successful; after the preliminary verification is successful, the operator's name characters and the time of applying to open the business library are inserted into the initial password, and a dynamic password is randomly generated; the operator enters the dynamic password, and the business library is opened if the password verification is correct; if the password verification is incorrect, the dynamic password is regenerated; through the multi-layer verification strategy, the protection capability of the bank system is significantly enhanced to ensure the security of the business library. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0042] Figure 2 This is a schematic diagram of the identity authentication process of the present invention;

[0043] Figure 3 This is a schematic diagram of the dynamic password generation process of the present invention;

[0044] Figure 4 The figure is a flow chart of the method for unlocking the service library of the present invention. DETAILED DESCRIPTION

[0045] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] See also Figure 1 , the first aspect of the present invention provides a banking service library security control system based on face recognition and dynamic password lock, including a face recognition module and a dynamic password module;

[0047] The face recognition module identifies the identity information of the operator and performs preliminary verification;

[0048] The face image is the face image of the operator who opens the service library and is collected by a camera at the service library; the identity information includes the face image to be identified and the emotion to be identified;

[0049] See also Figure 2 Specifically, the face image to be identified is compared with the face images in the bank's back-end database to determine whether there is a matching face image in the bank's back-end database; if not, the face image recognition fails and is re-verified. When the number of face recognition failures exceeds the preset number, the bank's monitoring terminal displays an abnormality and issues an alarm in time; the preset number is generally 3; if yes, the face image recognition is successful.

[0050] It should be noted that the bank's back-end database refers to the key data information system used within the bank to store and manage a large number of bank employees and customers.

[0051] For example, when an operator stands in front of a business database equipped with a high-resolution camera, the system first needs to confirm the operator's identity through facial recognition technology. The camera captures the operator's facial image. The image is processed using a pre-trained facial recognition algorithm, such as the deep learning-based FaceNet model, to extract facial features. The extracted features are compared with the facial features of operators registered in the database to confirm the operator's identity.

[0052] The successfully recognized face image is input into the emotion assessment module, and the emotion assessment module outputs the emotion vector; wherein, the emotion recognition module is constructed by the CNN neural network.

[0053] It should be noted that the emotion vector = [the score of each positive emotion type, the score of each negative emotion type].

[0054] For example: extract several facial images and different emotion types from historical data as input data of the CNN neural network model, and use the scores of each emotion type in the facial images as output data of the CNN neural network model; train the CNN neural network through the input data and output data to obtain an emotion assessment model; among them, emotion types are divided into two categories, namely positive emotions and negative emotions.

[0055] Analyze operators’ emotional types;

[0056] Specifically, the emotional score difference PFC between positive emotions and negative emotions is calculated by the formula PFC=ΣQi-ΣWj; wherein Qi is the score of each emotional type in positive emotions, Wj is the score of each emotional type in negative emotions, i is the positive emotional type, i=0, 1, ..., N; j is the negative emotional type, j=0, 1, ..., M; M, N are positive integers;

[0057] If the score difference is 0, there is no positive or negative emotion expression, and the operator's emotion is normal;

[0058] If the score difference is greater than 0, it indicates positive emotion and the operator’s emotion is normal;

[0059] If the score difference is less than 0, it indicates negative emotion and the operator’s emotion is abnormal.

[0060] The present invention evaluates the operator's emotion type through the emotion score difference; the operator's emotion type is mainly closely related to the score of the positive emotion type and the score of the negative emotion type; if the comprehensive score of the positive emotion is greater than the comprehensive score of the negative emotion, it indicates that the operator's emotion tends to be positive, that is, the emotion score difference is greater than 0; if the comprehensive score of the positive emotion is less than the comprehensive score of the negative emotion, it indicates that the operator's emotion tends to be negative, that is, the emotion score difference is less than 0.

[0061] If the emotion type is positive, the operator is in a normal mood and the initial verification is successful; if the emotion type is negative, the operator is in an abnormal mood and the initial verification fails, then the bank monitoring terminal displays an abnormality and issues an alarm in a timely manner;

[0062] It should be noted that when the bank system detects an abnormal situation, the alarm will trigger a covert warning mechanism, such as a silent alarm and a covert information transmission system, which will not disturb potential criminals and ensure the safety of on-site personnel while waiting for external support to arrive.

[0063] After the identity information is successfully verified, the bank's back-end database generates a dynamic password for final verification.

[0064] See also Figure 3Specifically, the operator's name characters and the time of applying to open the business library are inserted into the initial password and converted into binary to obtain a binary password; wherein, the order of the name characters in the English alphabet is extracted, and the order is converted into binary to obtain the binary form of the name characters; the time of applying to open the business library does not include the year;

[0065] The number of bits of the binary password is processed by ratio with the number of bits of the initial password, and the binary password is divided into several binary groups according to the ratio to obtain several binary groups, and each binary group is converted into decimal to obtain a dynamic password; wherein, if the ratio is not equal to 1, it means that there are redundant parts or missing parts in the binary code bits; wherein, the redundant parts are deleted; and the missing parts are supplemented with 0.

[0066] It should be noted that each time you apply to open a business library, a dynamic password will be regenerated.

[0067] For example: Assume that on May 1, 2024, Zhang San applies to open a business database, and his identity information has been successfully verified; the bank's back-end database extracts the operator's name and application time; the initial password of the business database is a six-digit character consisting of numbers and letters; the operator's name and application time are randomly inserted into the initial password; if the initial password is "1a3k25", after random insertion, it is assumed that "15a31k2zhangsan5" is obtained;

[0068] Now convert the operator's name "Zhang San" into the pinyin "zhangsan", and extract the order of the name pinyin in the order of English letters, and get "26,8,1,14,7,19,1,14" and convert it into binary "11010, 1000, 1, 1110, 111, 10011, 1110"; convert the application time "5.1" into binary "101, 1"; if the binary of the initial password "1a3k25" is "1111101110101"; convert the randomly inserted password into binary form, and perform ratio processing on the number of binary digits and the set number of initial password digits, divide the binary password into several groups, and then convert it into decimal to obtain a dynamic password.

[0069] If the number of binary numbers converted to decimal is greater than 10, then rounding is used to round the decimal value from the last digit to the first digit until one digit is retained.

[0070] Among them, if the highest digit is 9 and then increases by 1 to 10, it is rounded up again until it becomes a single digit.

[0071] See also Figure 4 The second aspect of the present invention provides a control method for a banking service library security control system based on face recognition and dynamic password lock, including:

[0072] Step 1: Collect the operator's face image;

[0073] Step 2: Compare the face image with the face image in the bank's back-end database to determine whether the face image exists in the bank's back-end database; if yes, the face recognition is successful and proceed to step 3; if not, the face recognition fails and returns to step 1. When the number of face recognition failures exceeds the preset number, the bank monitoring terminal displays an abnormality and issues an alarm in time;

[0074] Step 3: After successful face recognition, the face image is input into the emotion assessment model to obtain the emotion vector, and the emotion score difference is calculated to determine whether the emotion is normal; if yes, the preliminary verification is successful and the process goes to step 4; if no, the bank monitoring terminal displays an abnormality and issues an alarm in time;

[0075] Step 4: After the initial verification is successful, insert the operator's name and the time of applying to open the business database into the initial password to randomly generate a dynamic password;

[0076] Step 5: The operator enters the dynamic password. If the password verification is correct, the business library will be opened; if the password verification is incorrect, return to step 4.

[0077] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0078] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A banking database security control system based on face recognition and dynamic password lock, characterized in that: Including face recognition module and dynamic password module; Face recognition module: used to identify the identity information of the operator and conduct preliminary verification; the identity information includes the face image to be identified and the emotion to be identified; Dynamic password module: After the initial verification is successful, it is used to randomly insert the operator information on the basis of the initial password to generate a dynamic password for final verification; the operator information includes the operator's name and the time of applying to open the business database.

2. The banking business library security control system based on face recognition and dynamic password lock according to claim 1 is characterized in that: The face image is the face image of the operator who opens the service library and is collected by a camera at the service library.

3. The banking business library security control system based on face recognition and dynamic password lock according to claim 1 is characterized in that: The identification information of the operator includes: Extract the features of the face image to be identified, compare them with the features of the face images in the bank's back-end database, and determine whether there is a matching face image in the bank's back-end database; if not, the face image recognition fails; if yes, the face image recognition succeeds; and, The successfully recognized face image is input into the emotion recognition module, which outputs the emotion vector to analyze the operator's emotion type. If the emotion type is positive, the operator is in a normal mood and the initial verification is successful. If the emotion type is negative, the operator is in an abnormal mood and the initial verification fails. The emotion recognition module is constructed by a CNN neural network model.

4. The banking business library security control system based on face recognition and dynamic password lock according to claim 3 is characterized in that: When face verification fails, it will be re-verified. If the number of failures exceeds the preset number, the bank monitoring terminal will display an abnormality and alarm in time; and when the emotions are abnormal, the bank monitoring terminal will display an abnormality and alarm in time.

5. The banking business library security control system based on face recognition and dynamic password lock according to claim 3 is characterized in that: Based on the emotion vector, the analysis of the operator's emotion type includes: The score difference PFC between positive emotions and negative emotions is calculated by the formula PFC=ΣQi-ΣWj; wherein Qi is the score of each emotion type in positive emotions, Wj is the score of each emotion type in negative emotions, i is the positive emotion type, i=0, 1, ..., N; j is the negative emotion type, j=0, 1, ..., M; M, N are positive integers; If the score difference is 0, there is no positive or negative emotion expression, and the operator's emotion is normal; If the score difference is greater than 0, it indicates positive emotion and the operator’s emotion is normal; If the score difference is less than 0, it indicates negative emotion and the operator’s emotion is abnormal.

6. The banking business library security control system based on face recognition and dynamic password lock according to claim 1 is characterized in that: The method of randomly inserting operator information on the basis of the initial password to generate a dynamic password includes: The operator's name characters and the time of applying to open the business library are randomly inserted into the initial password and converted into binary to obtain a binary password; the binary password is divided into several groups according to the preset number of digits of the password to obtain several binary groups, and each group is converted into decimal to obtain a dynamic password; wherein, each application to open the business library will regenerate a dynamic password.

7. The bank service library security control system based on face recognition and dynamic password lock according to claim 6 is characterized in that: The order of the name characters in the English alphabet is extracted, and the order is converted into binary to obtain the binary form of the name characters.

8. The banking database security control system based on face recognition and dynamic password lock according to claim 6 is characterized in that: The binary password is divided into several groups according to the number of bits set in the password, including: The number of bits in the binary password is compared with the number of bits in the initial password. If the ratio is not equal to 1, it means that there are redundant or missing bits in the binary code. The redundant parts are deleted, and the missing parts are filled with 0.

9. The bank service library security control system based on face recognition and dynamic password lock according to claim 6 is characterized in that: If the decimal value of the binary group is greater than 10, the decimal value will be rounded from the last digit to the first digit until one digit is retained.

10. Banking business library security control system based on face recognition and dynamic password lock, characterized in that: The control method of the control system includes: Step 1: Collect the operator's face image; Step 2: Compare the face image with the face image in the bank's back-end database to determine whether the face image exists in the bank's back-end database; if yes, the face recognition is successful and proceed to step 3; if not, the face recognition fails and returns to step 1. When the number of face recognition failures exceeds the preset number, the bank monitoring terminal displays an abnormality and issues an alarm in time; Step 3: After successful face recognition, the operator's emotion type is obtained from the face image. If the emotion is normal, the preliminary verification is successful and the process goes to step 4. If the emotion is abnormal, the bank monitoring terminal will display the abnormality and issue an alarm in time. Step 4: After the initial verification is successful, insert the operator's name and the time of applying to open the business database into the initial password to randomly generate a dynamic password; Step 5: The operator enters the dynamic password. If the password verification is correct, the business library will be opened; if the password verification is incorrect, return to step 4.

Images