Campus card face recognition payment method, system and terminal
By intelligently identifying user identity and transaction type, and dynamically matching payment processes with historical data, the problem of face payment methods lacking flexibility in handling unconventional transactions in the existing technology is solved, and a more efficient and secure payment experience is achieved.
Patent Information
- Application Number
- CN202411901316.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-23
- Publication Date
- 2025-04-15
AI Technical Summary
The facial payment methods in the prior art are difficult to make flexible adjustments based on specific transaction scenarios, resulting in the system that may lack the necessary adaptability and flexibility when handling unconventional or special transactions.
By intelligently identifying user identity and transaction type, and dynamically matching the optimal payment process with historical data, selecting the most appropriate payment method to reduce potential risks and ensure the safety of user funds.
It realizes the selection of corresponding payment methods according to the needs of different transaction types, enhances the applicability and convenience of payment, improves overall service efficiency, optimizes resource allocation, and provides users with a more convenient and efficient payment experience.
Smart Images
Figure CN119831600A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of face payment technology, and in particular to a campus card face recognition payment method, system and terminal. Background Art
[0002] Under the current technological background, facial payment system, as an advanced payment method, has been widely used in various consumption scenarios. The system usually captures the user's facial image through a camera, uses facial recognition technology to authenticate the identity, and then completes the payment operation according to preset payment rules.
[0003] The current facial payment process generally includes key links such as facial image acquisition, feature extraction, identity authentication, payment instruction generation and execution. Although the facial payment system is convenient, the transaction scenarios on campus are diverse and complex. Different transaction types have different requirements for the security and convenience of the payment process. It is difficult to make flexible adjustments according to specific transaction scenarios, resulting in the system may lack the necessary adaptability and flexibility when processing unconventional or special transactions. Summary of the invention
[0004] This application provides a campus card face recognition payment method, system and terminal, which effectively solves the problem that the current face payment method in the prior art is difficult to flexibly adjust according to the specific transaction scenario, resulting in the system may lack the necessary adaptability and flexibility when processing unconventional or special transactions. It can flexibly respond to the needs of different transaction types, match the optimal payment process, and select the most appropriate payment method to reduce potential risks, ensure the security of user funds, enhance the applicability and convenience of payment, make payment services on campus more secure and efficient, improve overall service efficiency, optimize resource allocation, and provide users with a more convenient and efficient payment experience.
[0005] In a first aspect, the present application provides a campus card face recognition payment method, including: generating current transaction information; after generating the current transaction information, obtaining a user's face image; performing recognition processing on the face image to obtain the user's identity information and the campus card associated with the identity information; based on the identity information, obtaining the user's historical transaction information; comparing the current transaction information with the historical transaction information to determine the transaction type of the current transaction information; and determining the payment process of the campus card according to the transaction type.
[0006] Furthermore, the facial image is identified and processed to obtain the user's identity information and the campus card associated with the identity information, including: using a facial recognition model to extract features from the acquired facial image to obtain feature data; preprocessing the feature data; calculating the similarity between the preprocessed feature data and each facial feature in the first database; selecting the identity information corresponding to the facial feature with the highest similarity as the user's identity information; wherein the first database contains the facial features of all users, the identity information corresponding to the facial features, and the campus card associated with the identity information.
[0007] Furthermore, based on the identity information, obtaining the historical transaction information of the user includes: querying the historical transaction information of the identity information in a preset second database; wherein the second database contains the historical transaction information corresponding to all the identity information.
[0008] Furthermore, the transaction information includes: goods, the quantity of each kind of goods, and the amount to be paid; the transaction types include regular transactions and unconventional transactions; the current transaction information is compared with the historical transaction information to determine the transaction type of the current transaction information, including: traversing the goods in the current transaction information, and for the goods currently traversed, executing the following steps until the current transaction information is marked as an unconventional transaction or the traversal is completed: judging whether the goods currently traversed are abnormal based on the historical transaction information, if the goods currently traversed are abnormal, marking the current transaction information as an unconventional transaction; after the traversal is completed, if the current transaction information mark is still not marked as an unconventional transaction, marking the current transaction information as a regular transaction.
[0009] Furthermore, it is determined whether the currently traversed commodity is abnormal based on the historical transaction information, including determining whether the currently traversed commodity appears in the historical purchase information: if the currently traversed commodity appears in the historical purchase information, the average purchase quantity and standard deviation of the currently traversed commodity in the historical transaction information are calculated, and the Z-score method is used to determine whether the quantity of the currently traversed commodity in the current transaction information is abnormal; if the currently traversed commodity does not appear in the historical purchase information, the user's purchase intention is calculated, and based on the user's purchase intention, it is determined whether the price of the currently traversed commodity in the current transaction information is abnormal; it is determined whether at least one of the quantity and price of the currently traversed commodity is abnormal, and if so, the current transaction information is marked as an irregular transaction.
[0010] Furthermore, the average purchase quantity and standard deviation of the currently traversed commodity in the historical transaction information are calculated, and the Z-score method is used to determine whether the number of the currently traversed commodity in the current transaction information is abnormal, including: calculating the average purchase quantity of the currently traversed commodity in the historical transaction information: In the formula, μ represents the average purchase amount, n1 represents the number of transactions of the product in the historical transaction information, Represents the purchase amount of the i1th transaction; calculate the standard deviation of the currently traversed commodity in the historical transaction information: Where σ represents the standard deviation; calculate the Z-score: In the formula, Z represents Z-score, x new Represents the purchase quantity of the current commodity in the current transaction information; determines whether |Z|> the preset threshold, and if so, determines that the quantity of the currently traversed commodities in the current transaction information is abnormal.
[0011] Furthermore, the user's purchase intention is calculated, and based on the user's purchase intention, whether the price of the currently traversed commodity in the current transaction information is abnormal is determined, including: obtaining commodities similar to the current commodity in the current merchant, marking them as similar commodities, and obtaining the prices of the similar commodities; identifying similar commodities from the user's historical transaction information, and marking them as reference commodities; sorting the current commodity and similar commodities in order of price from low to high, and calculating the price positioning of the reference commodity based on the sorting: Where PPI1 represents the price positioning of the reference product, n2 represents the number of identified reference products, represents the serial number of the i2th identified reference product in the sorting, and m represents the number of types of the current product and similar products; determines whether PPI1 is greater than 0.5, and if so, marks the user's purchase intention for the current product as a high-quality purchase, and if not, marks the user's purchase intention for the product as a high-cost-performance purchase; based on the sorting, calculates the price positioning of the current product: In the formula, PPI2 represents the current price positioning of the product, r new Represents the serial number of the current product in the sorting; if the user's purchase intention for the product is high-quality purchase: determine whether PPI2 is greater than 0.5. If PPI2 is not greater than 0.5, then determine that the price of the currently traversed product in the current transaction information is abnormal; if the user's purchase intention for the product is high-cost-performance purchase: determine whether PPI2 is not greater than 0.5. If PPI2 is greater than 0.5, then determine that the price of the currently traversed product in the current transaction information is abnormal.
[0012] Furthermore, the payment process of the campus card is determined according to the transaction type, including: if the current transaction information is marked as a regular transaction, the amount to be paid is directly deducted from the balance of the campus card; if the current transaction information is marked as an unconventional transaction, a button is generated for the user to click to confirm, and the click event of the button is monitored; it is determined whether the click event is triggered, and if triggered, the amount to be paid is deducted from the balance of the campus card.
[0013] On the second aspect, the present application provides a campus card face recognition payment system, which includes: a transaction information generation module, a face image acquisition module, an identity information and campus card recognition module, a historical transaction information acquisition module, a transaction information matching module, and a payment process determination module.
[0014] The transaction information generation module is used to generate current transaction information; the face image acquisition module is used to acquire the user's face image after generating the current transaction information; the identity information and campus card recognition module is used to identify the face image to obtain the user's identity information and the campus card associated with the identity information; the historical transaction information acquisition module is used to acquire the user's historical transaction information based on the identity information; the transaction information matching module is used to compare the current transaction information with the historical transaction information to determine the transaction type of the current transaction information; the payment process determination module is used to determine the payment process of the campus card according to the transaction type.
[0015] In a third aspect, the present application provides a terminal comprising a memory and a processor; the memory is used to store a computer program; and the processor is used to implement the steps of the campus card face recognition payment method as described in the first aspect when executing the computer program.
[0016] In a fourth aspect, the present application provides a readable storage medium, which stores computer program instructions. When the computer program instructions are read and executed by a processor, the steps of the campus card face recognition payment method described in the first aspect are executed.
[0017] The technical solution provided by this application has at least the following technical effects or advantages:
[0018] Through intelligent identification of user identity and transaction type, and combining historical data to dynamically match the optimal payment process, the technical means effectively solves the problem that the current facial payment method in the existing technology is difficult to flexibly adjust according to the specific transaction scenario, resulting in the system may lack the necessary adaptability and flexibility when processing unconventional or special transactions. The corresponding payment method can be selected according to the needs of different transaction types, thereby reducing potential risks, ensuring the security of user funds, enhancing the applicability and convenience of payment, making payment services on campus more secure and efficient, improving overall service efficiency, optimizing resource allocation, and providing users with a more convenient and efficient payment experience. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 This is a flow chart of the campus card face recognition payment method in Example 1 of the present application;
[0020] Figure 2 For this application Figure 1A flowchart of comparing current transaction information with historical transaction information to determine the transaction type of the current transaction information;
[0021] Figure 3 For this application Figure 2 A flowchart of traversing the commodities in the current transaction information;
[0022] Figure 4 This is a module diagram of the campus card face recognition payment system in Example 2 of this application. DETAILED DESCRIPTION
[0023] In order to solve the problems raised by the background technology, this application uses technical means to intelligently identify user identity and transaction type, and dynamically match the optimal payment process with historical data. It can select the corresponding payment method according to the needs of different transaction types, thereby reducing potential risks, ensuring the security of user funds, and enhancing the applicability and convenience of payment, making payment services on campus more secure and efficient, improving overall service efficiency, optimizing resource allocation, and providing users with a more convenient and efficient payment experience.
[0024] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.
[0025] Embodiment 1:
[0026] like Figure 1-Figure 3 As shown, this embodiment provides a campus card face recognition payment method, including:
[0027] S100. Generate current transaction information.
[0028] In S100, after the user completes the purchase of goods or services, the store manager will enter the purchased goods or services into the computer system, and the system will automatically generate transaction information based on the product information entered by the manager.
[0029] S200. After generating the current transaction information, obtain the user's facial image.
[0030] After the transaction information is generated, the user is ready to pay to complete the order. At this time, the system will start a specific face acquisition device to scan the user's face, such as FaceScan-2000. FaceScan-2000 uses advanced camera technology and image processing algorithms to capture the user's facial image in a very short time. When the user stands in front of the device, the device will emit soft infrared light to illuminate the user's face, and quickly take a clear face photo through the camera, providing reliable data support for subsequent identity verification and payment processes.
[0031] S300. Perform recognition processing on the facial image to obtain the user's identity information and the campus card associated with the identity information.
[0032] By using advanced face recognition technology, the user's face image obtained in step S200 is analyzed and matched with the information in the relevant database, so that the user's identity information can be quickly and accurately identified and the campus card information held by the user can be automatically associated.
[0033] S400. Based on the identity information, obtain the user's historical transaction information.
[0034] When the user's identity information is accurately identified by the system, the system immediately accesses the associated database to query the user's past transaction information on campus, thereby understanding the user's consumption habits, preferences, historical transaction types and other information, providing a reference for merchants in subsequent transaction processing.
[0035] S500. Compare the current transaction information with the historical transaction information to determine the transaction type of the current transaction information.
[0036] A detailed comparison of the current transaction information generated in step S100 with the user's historical transaction information obtained in step S400 can intelligently identify and determine the type of this transaction, such as whether it is a regular purchase, which helps to understand the user's consumption behavior and provide an accurate type basis for subsequent payment processes.
[0037] S600. Determine the payment process of the campus card according to the transaction type.
[0038] According to the transaction type determined in step S500, the corresponding campus card payment process can be quickly selected and started. The corresponding payment method can be selected according to the needs of different transaction types, thereby reducing potential risks, ensuring the security of user funds, enhancing the applicability and convenience of payment, making payment services on campus more secure and efficient, improving overall service efficiency, optimizing resource allocation, and providing users (students and faculty) with a more convenient and efficient payment experience.
[0039] In S300, the facial image is recognized and processed to obtain the user's identity information and the campus card associated with the identity information, including:
[0040] S310. Use a face recognition model to extract features from the acquired face image to obtain feature data.
[0041] The face recognition model is used to deeply analyze the face image. The model will focus on extracting two major types of features: one is geometric features, which include the overall contour of the face, the shape and size of the facial features (such as eyes, nose, and mouth), and their relative position relationship; the other is texture and depth features, which covers the subtle texture, wrinkles, spots on the skin surface, and depth information of various parts of the face. These features can be extracted through deep learning technology (such as convolutional neural networks), which can capture more subtle and unique details of the face, thereby improving recognition accuracy.
[0042] S320. Preprocess the feature data.
[0043] Preprocessing includes normalizing the feature data to ensure that the numerical ranges of different features are consistent, which is convenient for subsequent comparison and analysis, and then performing noise reduction. The algorithm removes noise and redundant information in the feature data and retains key valid features. Feature enhancement can also be performed to strengthen certain key features to improve their recognition during the comparison process.
[0044] S330. Calculate the similarity between the preprocessed feature data and each facial feature in the first database; select the identity information corresponding to the facial feature with the highest similarity as the user's identity information; wherein the first database contains the facial features of all users, the identity information corresponding to the facial features, and the campus card associated with the identity information.
[0045] The first database is a comprehensive data storage system that systematically integrates all users' facial feature data, their closely corresponding identity information, and the campus card information associated with this information. In this system, facial feature data, identity information, and campus card information can be flexibly stored separately or in combination according to actual needs, and distributed in one or more specialized databases.
[0046] The preprocessed feature data is carefully compared with the first database to evaluate the similarity between the preprocessed feature data and each facial feature stored in the database, such as using Euclidean distance or cosine similarity. When the Euclidean distance calculation method is used, the preprocessed feature data and the facial feature data in the database are regarded as vectors in the feature space. By calculating the straight-line distance between the two vectors, a value can be obtained. The smaller the value, the higher the similarity between the two. The cosine similarity calculation method calculates the cosine value of the angle between the preprocessed feature data vector and the facial feature data vector in the database. The range of this cosine value is between -1 and 1. When the cosine value is close to 1, it indicates that the directions of the two vectors are very close, that is, the similarity is very high. After completing the similarity calculation, the identity information corresponding to the facial feature with the highest similarity is selected as the identity information of the current user.
[0047] Acquiring the historical transaction information of the user based on the identity information in S400 includes: querying the historical transaction information of the identity information in a preset second database; wherein the second database contains the historical transaction information corresponding to all the identity information.
[0048] Based on the confirmed user identity information, the user's historical transaction information is further obtained. This step is achieved by querying the preset second database, which is a database that specifically stores all user identity information and its corresponding historical transaction information. After obtaining the user's identity information, the content in this identity information will be used as a query condition, such as the user's unique identification information, and submitted to the second database for retrieval. The database system can quickly locate the historical transaction records that match the identity information in the massive data.
[0049] In S500, the transaction information includes: commodities, quantity of each commodity, and amount to be paid; the transaction type includes regular transaction and irregular transaction.
[0050] Goods can be physical items, such as books and stationery, or services, such as catering and printing; the quantity of each commodity represents the specific quantity selected by the students; the amount to be paid represents the total fee that the user needs to pay, and the amount to be paid will be calculated based on the quantity, price and possible promotions or discounts of the goods purchased by the user. If the merchant has promotions or students enjoy specific discounts, the amount to be paid will be adjusted accordingly to reflect the discounted total price.
[0051] In S500, the current transaction information is compared with the historical transaction information to determine the transaction type of the current transaction information, including:
[0052] S510. Traverse the commodities in the current transaction information, and for the commodities currently traversed, perform the following steps until the current transaction information is marked as an unconventional transaction or the traversal is completed:
[0053] Determine whether the currently traversed commodity is abnormal based on historical transaction information. If the currently traversed commodity is abnormal, mark the current transaction information as an irregular transaction.
[0054] S520. After the traversal is completed, if the current transaction information is still not marked as an unconventional transaction, the current transaction information is marked as a regular transaction.
[0055] S500 is mainly dedicated to determining the transaction type of the current transaction information. The transaction type is mainly divided into two categories: regular transactions and unconventional transactions. The current transaction information is comprehensively and carefully compared with the user's historical transaction information. The comparison process starts from step S510, and each commodity in the current transaction information is examined one by one. For each commodity, it is judged whether it is abnormal based on historical information. The regularity of the transaction can be further evaluated, and the current transaction information can be marked as an unconventional transaction, and the subsequent traversal process can be stopped.
[0056] S520 is performed after the traversal of all commodities is completed. If the current transaction information is not marked as an unconventional transaction during the traversal process, it can be reasonably inferred that the commodities in this transaction are commodities that the user regularly purchases, and the price and quantity of the commodities also meet the user's purchasing intention, so the current transaction information is marked as a regular transaction.
[0057] Through such a comparison process, it is possible to accurately distinguish between regular transactions and irregular transactions, providing a favorable reference for subsequent payment process selection.
[0058] In S510, judging whether the currently traversed commodity is abnormal according to the historical transaction information includes:
[0059] S511. For the currently traversed product, determine whether it appears in the historical purchase information:
[0060] If the currently traversed product appears in the historical purchase information, calculate the average purchase quantity and standard deviation of the currently traversed product in the historical transaction information, and use the Z-score method to determine whether the quantity of the currently traversed product in the current transaction information is abnormal.
[0061] If the currently traversed commodity does not appear in the historical purchase information, the user's purchase intention is calculated, and based on the user's purchase intention, it is determined whether the price of the currently traversed commodity in the current transaction information is abnormal.
[0062] S512. Determine whether at least one of the quantity and price of the currently traversed goods is abnormal. If so, mark the current transaction information as an abnormal transaction.
[0063] In S511-S512, by combining historical transaction information and user behavior patterns, abnormal transactions caused by misoperation, fraudulent behavior or system failure can be more effectively identified and marked. Specifically, it is determined whether the quantity and price of goods purchased by the user are abnormal based on the historical transaction information.
[0064] In S511, the average purchase quantity and standard deviation of the currently traversed commodity in the historical transaction information are calculated, and the Z-score method is used to determine whether the quantity of the currently traversed commodity in the current transaction information is abnormal, including:
[0065] Sa1. Calculate the average purchase volume of the currently traversed commodity in the historical transaction information: In the formula, μ represents the average purchase amount, n1 represents the number of transactions of the product in the historical transaction information, Represents the purchase amount of the i1th transaction.
[0066] Sa2. Calculate the standard deviation of the currently traversed commodity in the historical transaction information: In the formula, σ represents the standard deviation.
[0067] Sa3. Calculate Z-score: In the formula, Z represents Z-score, x new Represents the purchase amount of the current product in the current transaction information.
[0068] Sa4. Determine whether |Z|>preset threshold, if yes, determine that the number of currently traversed commodities in the current transaction information is abnormal.
[0069] In Sa1-Sa4, the purchase volume of each transaction of the currently traversed commodity is counted in the historical transaction information, and the average of these purchase volumes is calculated as the normal purchase level benchmark for the commodity. At the same time, the standard deviation is calculated to evaluate the discreteness of the purchase volume data, that is, the normal range of purchase volume fluctuations. The purchase volume of the commodity in the current transaction is obtained, and the Z-score formula is used to calculate the difference between the current purchase volume and the historical average purchase volume. The larger the Z-score value, the greater the difference between the current purchase volume and the historical average level. By presetting a threshold, if the value of |Z| is greater than the preset threshold, it means that the purchase volume deviates significantly from the historical average level, and it can be judged that the purchase volume of the commodity in the current transaction information is abnormal.
[0070] For example, in the past year, a student's average weekly purchase of a certain fruit is 2, and the standard deviation of the weekly purchase has been calculated from historical data, with a standard deviation of 1. In a certain week, the student suddenly buys 6 apples, then The preset threshold value |Z|>2 indicates anomaly. Since Z=4>2, the apple purchase volume this week is considered abnormal.
[0071] The value of the preset threshold may refer to the average purchase amount, for example, the value of the preset threshold may be the value of the average purchase amount.
[0072] In S511, the user's purchase intention is calculated, and based on the user's purchase intention, whether the price of the currently traversed commodity in the current transaction information is abnormal is determined, including:
[0073] Sb1. Obtain products similar to the current product from the current merchant, mark them as similar products, and obtain the prices of similar products. Specifically, this can be done by presetting a category table containing all the merchant's products, matching the current product with the categories in the category table according to the defined matching rules, and matching it to the corresponding category, as well as products belonging to the same category (similar products).
[0074] Sb2. Identify similar products from the user's historical transaction information and mark them as reference products.
[0075] Sb3. Sort the current product and similar products in order of price from low to high, and calculate the price positioning of the reference product based on the sorting: Where PPI1 represents the price positioning of the reference product, n2 represents the number of identified reference products, represents the serial number of the i2th identified reference product in the sorting, and m represents the number of types of the current product and similar products.
[0076] Sb4. Determine whether PPI1 is greater than 0.5. If so, mark the user's purchase intention for the current product as high-quality purchase. If not, mark the user's purchase intention for the product as high-cost-performance purchase.
[0077] Sb5. Based on the ranking, calculate the price positioning of the current product: In the formula, PPI2 represents the current price positioning of the product, r new Represents the sequence number of the current product in the sorting.
[0078] If the user's purchase intention for the product is high-quality purchase: determine whether PPI2 is greater than 0.5. If PPI2 is not greater than 0.5, determine that the price of the currently traversed product in the current transaction information is abnormal.
[0079] If the user's purchase intention for the product is to purchase with high cost performance: determine whether PPI2 is not greater than 0.5. If PPI2 is greater than 0.5, determine that the price of the currently traversed product in the current transaction information is abnormal.
[0080] In Sb1-Sb5, similar products to the current product are searched in the current merchant, and the prices of similar products are obtained. These similar products are then identified from the user's historical transaction information and marked as reference products. The number of reference products can be a specified number, preferably 1-3 reference products. The current product and similar products are sorted from low to high according to price. Based on this sorting, the price positioning of the reference product is calculated to reflect the index of the position of the reference product in the price sorting. According to the value of PPI1, the user's purchase intention for the current product can be judged. If PPI1 is greater than 0.5, it indicates that the user prefers similar products with lower prices, and the user's purchase intention for the current product is marked as high-quality purchase. If PPI1 is not greater than 0.5, it indicates that the user prefers similar products with higher prices, and the user's purchase intention for the current product is marked as high-cost-effective purchase. Finally, the position of the current product in the price sorting is calculated. For users with high-quality purchase intention, if PPI2 is not greater than 0.5, it means that the price is abnormal. For users with high-cost-effective purchase intention, if PPI2 is greater than 0.5, it means that the price is abnormal.
[0081] For example, the prices of six types of apples, A, B, C, D, E, and F, are obtained from the current merchant, and they are sorted from low to high by price: ￥11.2, ￥17.51, ￥22.42, ￥36.23, ￥40.00, and ￥48.00. In the user's historical transaction information, apples E and F are often purchased, so these two apples are identified as reference products, and the price positioning of reference products E and F is calculated. Since PPI1 is greater than 0.5, it indicates that the user's purchase intention for the current product is high quality. If the current product is apple type B, its price positioning is It is not greater than 0.5, which means that the price of this apple is abnormal for the user.
[0082] In S600, the payment process of the campus card is determined according to the transaction type, including:
[0083] If the current transaction information is marked as a regular transaction, the amount to be paid will be directly deducted from the balance of the campus card.
[0084] If the current transaction information is marked as an unconventional transaction, a button is generated for the user to click to confirm, and the click event of the button is monitored; it is determined whether the click event is triggered, and if triggered, the amount to be paid is deducted from the balance of the campus card.
[0085] The payment process of the campus card is intelligently determined according to the transaction type. After the order of the current transaction information is generated, the user's identity information is obtained through face recognition, and the associated database is accessed through the identity information, and the associated campus card operation authority is obtained. The transaction type of this transaction information is determined based on historical transaction information. For regular transactions, the amount to be paid is directly deducted from the balance of the campus card without the user's reconfirmation. For transactions marked as unconventional, the system will generate a confirmation button and listen to the user's click event. Only when the user explicitly clicks to confirm will the system deduct money from the campus card.
[0086] For example, the prices of six types of apples, A, B, C, D, E, and F, in the current merchants are: ￥11.2, ￥17.51, ￥22.42, ￥36.23, ￥40.00, and ￥48.00. The user's historical transaction information often shows the purchase of two types of fruits, A and B. If there is only type B fruit in the current transaction information, the transaction type is judged as a regular purchase based on the historical transaction information, and the purchase price is directly deducted from the campus card without the user's reconfirmation. This design can effectively improve payment efficiency and reduce waiting time in crowded places such as campuses. However, if the user lacks knowledge of various apples and mistakenly selects type E apples, the transaction type is judged as an unconventional purchase based on the historical transaction information, and a button for the user to confirm will be generated. Only when the user explicitly clicks to confirm will the system deduct money from the campus card. This effectively avoids the situation of erroneous payment caused by misidentification or misoperation, and ensures the security and accuracy of payment. This intelligent payment process design not only improves payment efficiency but also ensures the security of users' funds.
[0087] Embodiment 2:
[0088] like Figure 4 As shown, this embodiment provides a campus card face recognition payment system, which includes: a transaction information generation module, a face image acquisition module, an identity information and campus card recognition module, a historical transaction information acquisition module, a transaction information matching module, and a payment process determination module.
[0089] The transaction information generation module is used to generate current transaction information; the face image acquisition module is used to acquire the user's face image after generating the current transaction information; the identity information and campus card recognition module is used to identify the face image to obtain the user's identity information and the campus card associated with the identity information; the historical transaction information acquisition module is used to acquire the user's historical transaction information based on the identity information; the transaction information matching module is used to compare the current transaction information with the historical transaction information to determine the transaction type of the current transaction information; the payment process determination module is used to determine the payment process of the campus card according to the transaction type.
[0090] This embodiment has all the advantages of the campus card face recognition payment method in Embodiment 1, and can automatically implement the steps of the campus card face recognition payment method.
[0091] Embodiment three:
[0092] This embodiment provides a terminal, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement the steps of the campus card face recognition payment method in Example 1 when executing the computer program.
[0093] Embodiment 4:
[0094] This embodiment provides a readable storage medium, in which computer program instructions are stored. When the computer program instructions are read and executed by a processor, the steps of the campus card face recognition payment method in Embodiment 1 are executed.
[0095] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0096] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1The steps of the functions specified in a box or multiple boxes. Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they know the basic creative concept. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.
Claims
1. A campus card face recognition payment method, characterized in that: include: Generate current transaction information; After generating the current transaction information, obtaining the user's face image; Performing recognition processing on the facial image to obtain the user's identity information and the campus card associated with the identity information; Based on the identity information, obtaining historical transaction information of the user; Compare current transaction information with historical transaction information to determine the transaction type of the current transaction information; The payment process of the campus card is determined according to the transaction type.
2. The campus card face recognition payment method as claimed in claim 1, characterized in that: Performing recognition processing on the face image to obtain the user's identity information and the campus card associated with the identity information includes: A face recognition model is used to extract features from the acquired face image to obtain feature data; Preprocessing the feature data; Calculate the similarity between the preprocessed feature data and each facial feature in the first database; select the identity information corresponding to the facial feature with the highest similarity as the user's identity information; wherein the first database contains the facial features of all users, the identity information corresponding to the facial features, and the campus card associated with the identity information.
3. The campus card face recognition payment method as claimed in claim 1, characterized in that: Based on the identity information, obtaining the historical transaction information of the user includes: Query the historical transaction information of the identity information in a preset second database; wherein the second database contains the historical transaction information corresponding to all the identity information.
4. The campus card face recognition payment method as claimed in claim 1, characterized in that: The transaction information includes: commodities, quantity of each commodity, and amount to be paid; the transaction types include regular transactions and irregular transactions; Compare the current transaction information with the historical transaction information to determine the transaction type of the current transaction information, including: Traverse the commodities in the current transaction information, and for the currently traversed commodities, perform the following steps until the current transaction information is marked as an irregular transaction or the traversal is completed: determine whether the currently traversed commodity is abnormal based on historical transaction information, and if the currently traversed commodity is abnormal, mark the current transaction information as an irregular transaction; After the traversal is completed, if the current transaction information is still not marked as an unconventional transaction, the current transaction information is marked as a regular transaction.
5. The campus card face recognition payment method as claimed in claim 4, characterized in that: Determine whether the currently traversed product is abnormal based on historical transaction information, including: For the currently traversed product, determine whether it appears in the historical purchase information: If the currently traversed product appears in the historical purchase information, calculate the average purchase quantity and standard deviation of the currently traversed product in the historical transaction information, and use the Z-score method to determine whether the quantity of the currently traversed product in the current transaction information is abnormal; If the currently traversed product does not appear in the historical purchase information, calculate the user's purchase intention, and based on the user's purchase intention, determine whether the price of the currently traversed product in the current transaction information is abnormal; Determine whether there is at least one abnormality in the quantity and price of the currently traversed goods. If so, mark the current transaction information as an abnormal transaction.
6. The campus card face recognition payment method as claimed in claim 5, characterized in that: Calculate the average purchase volume and standard deviation of the currently traversed commodity in the historical transaction information, and use the Z-score method to determine whether the quantity of the currently traversed commodity in the current transaction information is abnormal, including: Calculate the average purchase volume of the currently traversed product in historical transaction information: In the formula, μ represents the average purchase amount, n1 represents the number of transactions of the product in the historical transaction information, represents the purchase amount of the i1th transaction; Calculate the standard deviation of the currently traversed commodity in historical transaction information: In the formula, σ represents the standard deviation; Calculate the Z-score: In the formula, Z represents Z-score, x new Represents the purchase amount of the current product in the current transaction information; Determine whether |Z|>preset threshold, if yes, determine that the number of currently traversed commodities in the current transaction information is abnormal.
7. The campus card face recognition payment method as claimed in claim 5, characterized in that: Calculate the user's purchase intention, and based on the user's purchase intention, determine whether the price of the currently traversed product in the current transaction information is abnormal, including: Get products similar to the current product in the current merchant, mark them as similar products, and get the prices of similar products; Identify similar products from the user's historical transaction information and mark them as reference products; Sort the current product and similar products in order of price from low to high, and calculate the price positioning of the reference product based on the sorting: Where PPI1 represents the price positioning of the reference product, n2 represents the number of identified reference products, represents the serial number of the i2th identified reference product in the sorting, and m represents the number of types of the current product and similar products; Determine whether PPI1 is greater than 0.
5. If so, mark the user's purchase intention for the current product as high-quality purchase. If not, mark the user's purchase intention for the product as high-cost-performance purchase. Based on the ranking, calculate the price positioning of the current product: In the formula, PPI2 represents the current price positioning of the product, r new Represents the sequence number of the current product in the sorting; If the user's purchase intention for the product is high-quality purchase: determine whether PPI2 is greater than 0.
5. If PPI2 is not greater than 0.5, determine that the price of the currently traversed product in the current transaction information is abnormal; If the user's purchase intention for the product is to purchase with high cost performance: determine whether PPI2 is not greater than 0.
5. If PPI2 is greater than 0.5, determine that the price of the currently traversed product in the current transaction information is abnormal.
8. The campus card face recognition payment method as claimed in claim 5, characterized in that: Determining the payment process of the campus card according to the transaction type includes: If the current transaction information is marked as a regular transaction, the amount to be paid will be directly deducted from the balance of the campus card; If the current transaction information is marked as an unconventional transaction, a button is generated for the user to click to confirm, and the click event of the button is monitored; it is determined whether the click event is triggered, and if triggered, the amount to be paid is deducted from the balance of the campus card.
9. A campus card face recognition payment system, characterized in that: It includes: A transaction information generation module, wherein the transaction information generation module is used to generate current transaction information; A face image acquisition module, which is used to acquire a face image of a user after generating current transaction information; An identity information and campus card recognition module, which is used to perform recognition processing on the face image to obtain the user's identity information and the campus card associated with the identity information; A historical transaction information acquisition module, the historical transaction information acquisition module is used to acquire the historical transaction information of the user based on the identity information; A transaction information matching module, which is used to compare current transaction information with historical transaction information to determine the transaction type of the current transaction information; A payment process determination module is used to determine the payment process of the campus card according to the transaction type.
10. A terminal, characterized in that: It includes a memory and a processor; The memory is used to store computer programs; The processor is used to implement the steps of the campus card face recognition payment method as described in any one of claims 1 to 8 when executing the computer program.