Method and device for processing abnormal behavior order of selling cabinet, equipment and medium

Abstract

The embodiment of the invention provides a processing method and device for an abnormal behavior order of a selling cabinet, equipment and a medium. The method comprises the steps that target data of a user in the using process of the selling cabinet are acquired; the target data comprises at least one of user behavior data, order payment data and commodity change data; according to the target data, determining whether the order is an abnormal order through an abnormal order identification model; and if the order is an abnormal order, performing management and control processing. By acquiring the user behavior data, the order payment data and the commodity change data, the abnormal condition of the order can be analyzed from multiple dimensions, misjudgment possibly caused by a single data source is avoided, the target data is automatically analyzed through the abnormal order recognition model, the abnormal order can be recognized, more efficient abnormal order processing is achieved, and the user experience is improved. The order of the user in the using process of the selling cabinet can be judged in real time, and the abnormal order can be processed in time when found.

Description

Technical Field

[0001] The present invention relates to the field of intelligent identification technology, and in particular to a method for processing abnormal behavior orders of a vending cabinet, a device for processing abnormal behavior orders of a vending cabinet, a vending equipment and a computer-readable storage medium. Background Art

[0002] In the modern e-commerce era and during the use of unmanned vending machines, abnormal orders caused by uncivilized behavior have become a problem that cannot be ignored. Abnormal orders will cause serious economic losses and reputation losses to merchants. Abnormal orders usually include: orders with unuploaded goods and uncivilized shopping behavior types. When shopping, users may use false identity information and payment information to scan the code to open the door, and then refund or simply not pay after receiving the goods, thereby defrauding goods and funds. This will not only cause economic losses to merchants, but also affect the legitimate rights and interests of other consumers.

[0003] The existing technology for judging abnormal orders is manual, and it is necessary to check the monitoring images generated by the purchase order to verify whether the customer of the order has experienced mistaken deductions or theft. This is inefficient, and manual review requires time, which leads to delays. It is impossible to monitor and process abnormal orders in real time, resulting in losses when abnormal orders are discovered. Summary of the invention

[0004] In view of the above problems, embodiments of the present invention are proposed to provide a method for processing abnormal behavior orders of a vending cabinet, a device for processing abnormal behavior orders of a vending cabinet, a vending equipment and a computer-readable storage medium that overcome the above problems or at least partially solve the above problems.

[0005] In order to solve the above problem, a first aspect of an embodiment of the present invention provides a method for processing abnormal behavior orders of a vending cabinet, the method comprising:

[0006] Acquire target data of the user during use of the vending machine; the target data includes at least one of user behavior data, order payment data, and commodity change data;

[0007] According to the target data, determining whether the order is an abnormal order through an abnormal order identification model;

[0008] If the order is an abnormal order, control and management will be carried out.

[0009] Optionally, the obtaining of target data of the user during use of the vending cabinet includes:

[0010] Acquire video data collected by an image acquisition module provided at the vending cabinet;

[0011] Determining user behavior data based on the video data;

[0012] Obtaining order information, and extracting order payment data according to the order information;

[0013] Obtain commodity change data detected by a sensor disposed in the vending cabinet.

[0014] Optionally, determining whether the order is an abnormal order by using an abnormal order identification model according to the target data includes:

[0015] According to the target data, determining whether the order is an abnormal order through an abnormal order identification model, and if it is determined that the order is an abnormal order, determining the abnormality type;

[0016] If the order is an abnormal order, control and management are performed, including:

[0017] If the order is an abnormal order, control and management will be performed according to the abnormal type.

[0018] Optionally, the abnormality types include: uncivilized shopping, abnormal transactions, and abnormal commodities;

[0019] If the order is an abnormal order, control and management are performed according to the abnormal type, including:

[0020] If the abnormal type of the abnormal order is uncivilized shopping, the vending cabinet is controlled to be locked and a credit penalty is imposed on the user;

[0021] If the abnormal type of the abnormal order is the transaction abnormality, an abnormal behavior label is marked on the user;

[0022] If the abnormal type of the abnormal order is that the product is abnormal, an abnormal notification is sent to the operator of the vending cabinet.

[0023] Optionally, the determining whether the order is an abnormal order by using an abnormal order identification model according to the target data further includes:

[0024] If the abnormal order identification model determines that the order is an abnormal order, the target data corresponding to the order is sent to the operator of the vending machine so that the operation and maintenance personnel can reconfirm the abnormal condition of the order.

[0025] Optionally, the abnormal order recognition model can be trained in the following way:

[0026] Acquire target data of multiple users in the process of using the vending machine to create sample data; the sample data includes user behavior data, order payment data, and commodity change data;

[0027] Annotating the sample data to obtain annotated sample data;

[0028] Inputting the sample data and the annotations corresponding to the sample data into a deep learning model, training the deep learning model, and adjusting model parameters according to the training results;

[0029] After the deep learning model training is completed, the abnormal order recognition model is obtained.

[0030] Optionally, the method further comprises:

[0031] Acquire normal behavior data collected by multiple image acquisition modules provided in the vending cabinet; the normal behavior data includes the number of goods taken, the speed of taking goods, and the length of time the goods are taken;

[0032] The normal behavior data is compared with the user behavior data of the current user during the use of the vending machine, and whether the order is an abnormal order is determined based on the comparison result.

[0033] According to a second aspect of the present invention, a device for processing abnormal behavior orders of a vending cabinet is provided, the device comprising:

[0034] A data acquisition module, used to acquire target data of the user during the use of the vending cabinet; the target data includes at least one of user behavior data, order payment data, and commodity change data;

[0035] A first order determination module, configured to determine whether the order is an abnormal order according to the target data by using an abnormal order identification model;

[0036] The order processing module is used to perform control and processing if the order is an abnormal order.

[0037] Optionally, the data acquisition module includes:

[0038] A video acquisition submodule, used to acquire video data acquired by an image acquisition module provided at the vending cabinet;

[0039] A behavior data determination submodule, which determines user behavior data based on the video data;

[0040] A payment data acquisition submodule is used to acquire order information and to acquire order payment data according to the order information;

[0041] The commodity data acquisition submodule is used to acquire commodity change data detected by the sensor installed in the vending cabinet.

[0042] Optionally, the first order determination module includes:

[0043] An abnormality type determination submodule is used to determine whether the order is an abnormal order according to the target data through an abnormal order recognition model, and if the order is determined to be an abnormal order, determine the abnormality type;

[0044] The order processing module comprises:

[0045] The control and processing submodule is used to perform control and processing according to the exception type if the order is an abnormal order.

[0046] Optionally, the abnormality types include: uncivilized shopping, abnormal transactions, and abnormal commodities; the control and processing submodule includes:

[0047] The abnormal order processing unit is used to control the vending cabinet to lock the cabinet and impose a credit penalty on the user if the abnormal type of the abnormal order is the uncivilized shopping; if the abnormal type of the abnormal order is the transaction abnormality, mark the user with an abnormal behavior label; if the abnormal type of the abnormal order is the commodity abnormality, send an abnormal notification to the operator of the vending cabinet.

[0048] Optionally, the first order determination module further includes:

[0049] The order abnormality confirmation submodule is used to send the target data corresponding to the order to the operator of the sales cabinet if the abnormal order identification model determines that the order is an abnormal order, so that the operation and maintenance personnel can reconfirm the abnormal condition of the order.

[0050] Optionally, the abnormal order recognition model can be trained by the following modules:

[0051] A sample data acquisition module, used to acquire target data of multiple users during the use of the vending cabinet to create sample data; the sample data includes user behavior data, order payment data, and commodity change data;

[0052] A sample data labeling module, used to label the sample data to obtain labeled sample data;

[0053] A model training module, used to input the sample data and the annotations corresponding to the sample data into a deep learning model, train the deep learning model, and adjust the model parameters according to the training results;

[0054] The model determination module is used to obtain the abnormal order recognition model after the deep learning model training is completed.

[0055] Optionally, the device further comprises:

[0056] The second order confirmation module obtains normal behavior data collected by multiple image acquisition modules installed in the vending machine; the normal behavior data includes the number of goods picked up, the speed of goods picked up, and the length of time the goods are kept for picking up; the normal behavior data is compared with the user behavior data of the current user during the use of the vending machine, and whether the order is an abnormal order is determined based on the comparison result.

[0057] According to a third aspect of the present invention, a vending device is provided, comprising: a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein when the computer program is executed by the processor, the steps of a method for processing abnormal behavior orders of a vending cabinet as described above are implemented.

[0058] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of a method for processing abnormal behavior orders of a vending cabinet as described above.

[0059] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:

[0060] The embodiment of the present invention provides a method and device for processing abnormal behavior orders of a vending cabinet, the method comprising: obtaining target data of a user during the use of the vending cabinet; the target data comprising at least one of user behavior data, order payment data, and commodity change data; determining whether the order is an abnormal order through an abnormal order recognition model according to the target data; if the order is an abnormal order, performing management and control processing. By obtaining user behavior data, order payment data, and commodity change data, and automatically analyzing the target data through an abnormal order recognition model, abnormal orders can be identified to achieve more efficient abnormal order processing, and the orders of users during the use of the vending cabinet can be judged in real time, and abnormal orders can be processed in a timely manner when they are found. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a flowchart of the steps of a method for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention;

[0062] Figure 2 It is a schematic diagram of an interface calling flow of a method for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention;

[0063] Figure 3 It is a flowchart of the steps of another method for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention;

[0064] Figure 4 It is a schematic diagram of a model training process of a method for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention;

[0065] Figure 5 It is a structural block diagram of a device for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0066] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0067] The existing technology for judging abnormal orders is manual and requires checking the surveillance images generated by the purchase order to verify whether the customer of the order has made any mistaken deductions or thefts. This is inefficient and affects user experience, causing economic losses.

[0068] One of the core concepts of the embodiment of the present invention is to obtain the target data of the user during the use of the vending machine; the target data includes at least one of user behavior data, order payment data, and commodity change data; according to the target data, determine whether the order is an abnormal order through the abnormal order recognition model; if the order is an abnormal order, perform management and control. By obtaining user behavior data, order payment data, and commodity change data, and automatically analyzing the target data through the abnormal order recognition model, abnormal orders can be identified to achieve more efficient abnormal order processing, thereby improving user experience.

[0069] Reference Figure 1 , shows a flowchart of a method for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention, and the method may specifically include the following steps:

[0070] Step 101, obtaining target data of the user during the use of the vending cabinet; the target data includes at least one of user behavior data, order payment data, and commodity change data;

[0071] A vending machine is a device used to automatically sell goods, usually used in the retail industry. They can provide 24-hour uninterrupted service, reduce labor costs, and improve sales efficiency. There are many types of vending machines, which can be divided into several categories according to different goods and application scenarios: automatic vending machines, smart vending machines, unmanned retail cabinets, customized vending cabinets, etc., which are not limited in the embodiments of the present invention.

[0072] User behavior data refers to various behavior records generated by users when using products, services or systems. The infrared sensor installed in the vending machine can be used to transmit and receive infrared rays to detect the movement trajectory of objects. For example, the opening and closing state of the cabinet door, the range of movement of the product position, or the frequency of repeated movement of the user's arm to take the product will cause abnormal changes in the infrared sensing value. It is also possible to install surveillance cameras inside or around the unmanned vending machine, and debug the corresponding angles and key detection areas in the early stage, such as: product grids, door lock areas, surveillance cameras, etc. Start the surveillance recording from the time the door is opened until the cloud storage video ends when the door is closed. Or use the acceleration sensor installed in the vending machine to measure the acceleration of the human body in three directions, and calculate the movement trajectory and posture recognition of the human body by detecting the change in acceleration. The above data acquisition methods can be used to obtain user behavior data.

[0073] Order payment data refers to payment-related data records generated when users complete purchases. This includes the categories, quantities, prices, and times of purchases. This sample data can be used to analyze user consumption, user history behavior information, monitor whether new users' purchase behavior is abnormal, and whether old users' purchase habits, bills, and payment information have changed significantly. And the payment status of users' purchases.

[0074] Product change data refers to the record of changes in the properties of products at different time points or in different states. The pressure sensor installed on the sales cabinet monitors whether the pressure change on the cabinet shelf is abnormal. If the pressure value suddenly increases or decreases, the product has changed, and it is possible to judge that someone has placed the product maliciously or there are foreign objects. If the user frequently takes out and puts back the product in a short period of time (such as within 1 minute), or the gravity sensor detects abnormal weight changes at the corresponding position, the product change data of the product taken will be recorded.

[0075] In this embodiment, user behavior data includes the user's interaction with the vending machine, such as opening the cabinet door, selecting goods, closing the cabinet door, etc. The user's behavior is detected by sensors installed on the vending machine (such as infrared sensors, cameras, pressure sensors, etc.), or the user's movements are captured by a camera for image recognition and behavior analysis.

[0076] Order payment data includes the user's payment information when purchasing goods, such as payment method, payment amount, payment time, etc. The vending machine can integrate multiple payment methods (such as WeChat Pay, Alipay, bank card, etc.) and obtain payment data through the payment system's API; if the vending machine has an independent payment terminal (such as a barcode scanner, POS machine, etc.), the payment information can be recorded through the payment terminal; if the user uses an e-wallet to pay, the payment data can be obtained by connecting with the e-wallet system.

[0077] Product change data includes changes in inventory, types and locations of products in the sales cabinet. The entry and exit of products are monitored in real time through labels or barcode scanners to record changes in products. A weight sensor is installed in each grid of the sales cabinet to detect changes in weight to determine whether products are taken out or put in. The image of the products in the cabinet is captured by a camera, and image recognition technology is used to analyze changes in products. The inventory information of products can also be updated in real time by connecting to the backend inventory management system.

[0078] The acquired target data is stored in a local server or cloud database to ensure data security and accessibility. Through the above methods, the target data of users in the process of using the vending machine can be effectively acquired, and corresponding analysis and application can be carried out.

[0079] Step 102, determining whether the order is an abnormal order through an abnormal order recognition model according to the target data;

[0080] The abnormal order recognition model is a machine learning or data analysis model for detecting and identifying abnormal orders, and can be obtained by training a neural network model or a machine learning model, which is not limited in the embodiment of the present invention.

[0081] Abnormal orders refer to orders that are inconsistent with normal order behavior and may involve fraud, malicious behavior, operational errors or other abnormal situations. Orders that have not been uploaded and uncivilized shopping behaviors are identified as abnormal orders. Uncivilized shopping behaviors include: blocking the camera, putting in goods or other foreign objects, stealing, drinking, stealing, and violently destroying vending machines; unlisted goods and unidentified goods are judged as abnormal orders. The reason may be that the consumer took the goods in the cabinet that have no product model, or the operator did not set the inventory of the goods before loading, and the goods that were inconsistent with the product model library were put into the cabinet, resulting in the system failing to fail to retrieve the goods.

[0082] In this embodiment, based on the acquired user behavior data, order payment data, and commodity change data, these data are preprocessed and features are extracted, and the extracted features are combined into a feature vector as the input of the model. The currently obtained feature vector is input into the abnormal order recognition model, and the model outputs the probability or classification result of whether the order is an abnormal order. The identified abnormal order is marked as "abnormal" and the relevant information (such as the cause of the abnormality, time, user information, etc.) is recorded.

[0083] Step 103: If the order is an abnormal order, control and management are performed.

[0084] Control and management processing means taking a series of measures to manage and control abnormal orders after identifying them, so as to reduce risks, minimize losses and optimize operational efficiency.

[0085] In this embodiment, if the order is identified as an abnormal order, it needs to be managed and processed immediately to prevent further risks or losses. The identified abnormal order is marked as "abnormal", and relevant information (such as order ID, abnormal type, timestamp, etc.) is recorded, and user behavior data, order payment data, commodity change data, etc. are recorded to facilitate subsequent analysis and processing. The system can automatically reject the payment request and prompt the user that the payment failed or limit the user's operating authority to prohibit further operations. It can effectively deal with abnormal orders, ensure the security and stability of the vending cabinet system, and improve the user experience.

[0086] This embodiment can analyze order anomalies from multiple dimensions by acquiring user behavior data, order payment data, and product change data, avoiding misjudgments that may be caused by a single data source. By automatically analyzing target data through an abnormal order recognition model, abnormal orders can be identified to achieve more efficient abnormal order processing.

[0087] Reference Figure 2 , showing a schematic diagram of an interface calling flow of another method for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention;

[0088] In this embodiment, the user uses WeChat or Alipay applet to scan the QR code of the container. WeChat or Alipay payment score applet determines whether the credit score meets the requirements. For example, the credit score must be greater than 550 to authorize the cabinet door to be opened. The merchant server determines whether the door is allowed to be opened based on the judgment result. The server calls the interface to issue the door opening instruction. The sales cabinet hardware receives the development instruction device to execute the door opening successfully and sends the door opening notification. The merchant server system creates an order (starting surveillance recording from opening the door until closing the door and storing the video in the cloud) and the merchant server receives a successful callback. At this time, it starts to obtain the user behavior sample database and analyze and process the sample data. If there are abnormal changes, management measures are taken according to risk control. User shopping behavior in the cabinet, such as taking goods and other operations. The user closes the door to confirm the order callback notification. The sales cabinet hardware notifies the server of the order information. The merchant server receives the order successfully and the callback completes the automatic deduction. The payment success callback notification displays the order and payment information in the applet.

[0089] Reference Figure 3 , shows a flowchart of another method for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention, and the method may specifically include the following steps:

[0090] Step 201, obtaining target data of the user during the use of the vending cabinet; the target data includes at least one of user behavior data, order payment data, and commodity change data;

[0091] In this embodiment, user behavior data includes the user's interaction with the vending machine, such as opening the cabinet door, selecting goods, closing the cabinet door, etc. The user's behavior is detected by sensors installed on the vending machine (such as infrared sensors, cameras, pressure sensors, etc.), or the user's movements are captured by a camera for image recognition and behavior analysis.

[0092] Order payment data includes the user's payment information when purchasing goods, such as payment method, payment amount, payment time, etc. The vending machine can integrate multiple payment methods (such as WeChat Pay, Alipay, bank card, etc.) and obtain payment data through the payment system's API; if the vending machine has an independent payment terminal (such as a barcode scanner, POS machine, etc.), the payment information can be recorded through the payment terminal; if the user uses an e-wallet to pay, the payment data can be obtained by connecting with the e-wallet system.

[0093] Product change data includes changes in inventory, types and locations of products in the sales cabinet. The entry and exit of products are monitored in real time through labels or barcode scanners to record changes in products. A weight sensor is installed in each grid of the sales cabinet to detect changes in weight to determine whether products are taken out or put in. The image of the products in the cabinet is captured by a camera, and image recognition technology is used to analyze changes in products. The inventory information of products can also be updated in real time by connecting to the backend inventory management system.

[0094] The acquired target data is stored in a local server or cloud database to ensure data security and accessibility. Through the above methods, the target data of users in the process of using the vending machine can be effectively acquired, and corresponding analysis and application can be carried out.

[0095] In some embodiments, step 201 includes the following sub-steps:

[0096] Sub-step S11, obtaining video data collected by an image collection module provided at the vending cabinet; and determining user behavior data based on the video data;

[0097] Image acquisition module is a hardware or software component used to capture and record images. It is widely used in various fields, such as security monitoring, industrial detection, medical imaging, autonomous driving, virtual reality, etc. The core function of the image acquisition module is to obtain image data through a camera or other optical device and transmit it to a computer or other processing device for further analysis or storage.

[0098] Video data refers to a continuous sequence of images captured by a video acquisition device (such as a camera), which is usually stored and processed in digital form. Video data has a wide range of applications in many fields, including security monitoring, industrial inspection, and autonomous driving.

[0099] In this embodiment, the user behavior data is determined through the video data collected by the image acquisition module (such as a camera) installed in the sales cabinet. Cameras are installed at key locations of the sales cabinet (such as near the cabinet door, the product display area, etc.) to ensure that the user's complete operation process can be captured. The camera collects video data in real time and transmits the video data to the background server or local storage device, stores the video data in the local server or the cloud, and extracts the user behavior data. Based on the video data collected by the image acquisition module of the sales cabinet, the user behavior data is accurately extracted.

[0100] Sub-step S12, obtaining order information, and extracting order payment data according to the order information;

[0101] Order information refers to the detailed record generated when a user purchases goods or services, usually including basic order information, product information, payment information, user information, logistics information, etc.

[0102] In this embodiment, order information related to payment is obtained through the payment module or order management system of the vending machine or through API docking with the payment platform. Fields related to payment are extracted from the order information, such as payment amount, payment method, payment time, payment status, etc. The extracted order payment data is stored in a local database or a cloud database to ensure data security and accessibility. Order payment data can be efficiently extracted from order information and applied to scenarios such as payment reconciliation, payment analysis, and anomaly detection, thereby improving the intelligence level and security of the vending machine system.

[0103] Sub-step S13, obtaining commodity change data detected by a sensor installed in the vending cabinet.

[0104] The sensor includes a pressure sensor, an infrared sensor, a barcode scanner, etc., which is not limited in the embodiment of the present invention.

[0105] In this embodiment, sensors are installed in each grid or commodity display area of ​​the sales cabinet to ensure that changes in commodities can be detected in real time. By detecting the weight changes of commodities in the sales cabinet, the entry and exit of commodities can be judged, the value of the weight sensor can be read in real time, and the weight change (such as when commodities are taken out or put in) can be calculated. The extracted commodity change data is stored in a local database or a cloud database. Commodity change data can be efficiently detected by sensors in the sales cabinet, and applied to scenarios such as inventory management, anomaly detection, and user behavior analysis, thereby improving the intelligence level and user experience of the sales cabinet system.

[0106] Step 202, determining whether the order is an abnormal order through an abnormal order recognition model according to the target data;

[0107] In this embodiment, based on the acquired user behavior data, order payment data, and product change data, these data are preprocessed and features are extracted, and the extracted features are combined into a feature vector as the input of the model. The currently obtained feature vector is input into the abnormal order recognition model, and the model outputs the probability or classification result of whether the order is an abnormal order. The identified abnormal order is marked as "abnormal" and relevant information (such as abnormal reason, time, user information, etc.) is recorded.

[0108] In some embodiments, step 202 includes the following sub-steps:

[0109] Sub-step S21, determining whether the order is an abnormal order through an abnormal order recognition model according to the target data, and determining the abnormality type if the order is determined to be an abnormal order;

[0110] The types of exceptions include: transaction abnormalities: abnormal payment amount (such as the payment amount is 0 yuan or does not match the total price of the goods), abnormal payment method (such as frequent changes in payment methods), abnormal payment time (such as payment time does not match operation time).

[0111] Uncivilized shopping: abnormal user operation time (such as operating late at night), abnormal user operation path (such as frequently opening and closing cabinet doors), and abnormal user operation frequency (such as multiple operations in a short period of time).

[0112] Abnormal goods: abnormal product types (such as purchasing an unusual combination of goods), abnormal product quantities (such as purchasing a large number of goods at one time), abnormal product prices (such as product prices that do not match inventory records).

[0113] In this embodiment, based on the target data, an abnormal order recognition model is used to determine whether the order is an abnormal order. If the order is determined to be an abnormal order, the abnormality type is further determined based on the feature vector of the order combined with the output results of the model.

[0114] In some embodiments, step 202 includes the following sub-steps:

[0115] Sub-step S31, if the abnormal order recognition model determines that the order is an abnormal order, the target data corresponding to the order is sent to the operator of the vending machine so that the operation and maintenance personnel can reconfirm the abnormal status of the order.

[0116] The purpose of reconfirming the abnormal conditions of orders by operation and maintenance personnel is to protect the rights and interests of users. This sales system platform must establish a manual review and confirmation mechanism. Through the manual review process, the user's behavior is managed and secured to ensure the accuracy and reliability of the judgment. It can protect the legitimate rights and interests of consumers and improve consumer trust. The manual verification process may include the review of relevant records, screenshots, videos and other evidence, as well as the inquiry of the account involved and the verification of information to ensure the fairness and accuracy of the handling.

[0117] In this embodiment, the identified abnormal order is marked as "abnormal", and relevant information (such as order ID, abnormal type, timestamp, etc.) is recorded, and the target data corresponding to the abnormal order is sent to the operator of the vending cabinet. After receiving the notification, the operator checks the target data to understand the abnormality of the order. The operator determines whether the order is indeed abnormal based on the target data. For example, if the payment amount is 0 yuan, but the user actually takes the goods, it may be a payment system failure. If the user frequently opens and closes the cabinet door but does not take the goods, it may be a malicious operation. The operator confirms the abnormality of the order and records the review results.

[0118] Step 203, obtaining normal behavior data collected by multiple image acquisition modules installed in the vending machine; the normal behavior data includes the number of goods picked up, the speed of goods picked up, and the length of time the goods are kept for picking up; comparing the normal behavior data with the user behavior data of the current user during the use of the vending machine, and determining whether the order is an abnormal order based on the comparison result.

[0119] Normal behavior data refers to the expected or normal behavior patterns exhibited by users when using products, services or systems. Normal behavior data includes: number of items taken: the number of items taken by users during the operation; speed of item taking: the speed at which users take items (such as the number of items taken per second); and duration of stay when taking items: the length of time users stay when taking items.

[0120] In this embodiment, multiple cameras are installed at key locations of the sales cabinet (such as near the cabinet door, product display area, etc.) to ensure that the user's complete operation process can be captured. The user's normal behavior data is collected in real time through the image acquisition module and stored in the database, and features are extracted from the normal behavior data. The current user's behavior data is collected in real time through the image acquisition module, and the current user's behavior data is compared with the reference range of normal behavior data. For example, if the number of products taken by the current user exceeds the normal range, it is determined to be abnormal. The identified abnormal order is marked as "abnormal" and relevant information (such as abnormal type, timestamp, etc.) is recorded.

[0121] Through the anomaly detection algorithm, a large amount of normal purchase user behavior data is collected through user shopping videos detected by the camera in the sales cabinet, and the statistics of the mean of behavioral characteristics such as the number of items taken, the normal average speed of taking, and the length of time the arm stays in the cabinet after opening the door are calculated. When user behavior data appears, it is matched and compared with the statistics of the mean of this behavioral characteristic. If the number of items taken far exceeds the normal shopping average, or the length of stay is more than twice the normal average, the current user's order may be judged as an abnormal order.

[0122] Step 204: If the order is an abnormal order, control processing is performed.

[0123] In this embodiment, if the order is identified as an abnormal order, it needs to be managed and processed immediately to prevent further risks or losses. The identified abnormal order is marked as "abnormal", and relevant information (such as order ID, abnormal type, timestamp, etc.) is recorded, and user behavior data, order payment data, commodity change data, etc. are recorded to facilitate subsequent analysis and processing. The system can automatically reject the payment request and prompt the user that the payment failed or limit the user's operating authority to prohibit further operations. It can effectively deal with abnormal orders, ensure the security and stability of the vending cabinet system, and improve the user experience.

[0124] In some embodiments, step 204 includes the following sub-steps:

[0125] Sub-step S41, if the order is an abnormal order, control and processing are performed according to the abnormal type.

[0126] In this embodiment, if an order is determined to be an abnormal order, targeted management and control is required according to the type of abnormality to ensure the security and stability of the vending machine system.

[0127] In some embodiments, the abnormality types include: uncivilized shopping, abnormal transactions, and abnormal commodities; the step S41 includes the following sub-steps:

[0128] Sub-step S411, if the abnormal type of the abnormal order is the uncivilized shopping, the vending cabinet is controlled to be locked and a credit penalty is imposed on the user; if the abnormal type of the abnormal order is the transaction abnormality, the user is marked with an abnormal behavior label; if the abnormal type of the abnormal order is the product abnormality, an abnormal notification is sent to the operator of the vending cabinet.

[0129] According to the abnormal type of abnormal orders, they are divided into several categories: Uncivilized shopping: such as users maliciously damaging cabinet doors, frequently opening and closing cabinet doors, and successfully paying without taking away goods. Transaction abnormalities: such as abnormal payment amount, abnormal payment method, payment failure, etc. Product abnormalities: such as abnormal product quantity, abnormal product type, abnormal product inventory, etc.

[0130] In this embodiment, different control measures are taken according to the abnormal type of abnormal orders, which can effectively improve the security and user experience of the vending cabinet system. If the abnormal type is uncivilized shopping, the system immediately controls the vending cabinet to lock the cabinet to prevent the user from continuing to operate. After the cabinet is locked, the user cannot open or close the cabinet door or take away the goods, and the user is penalized for credit, which reduces the user's credit score.

[0131] If the abnormality type is a transaction abnormality, the system will label the user with an abnormal behavior label. Labeling methods may include: marking the user as an "abnormal transaction user" and recording their transaction history, restricting the user's payment method (such as prohibiting the use of certain payment methods), and sending notifications to users to remind them that their payment behavior is abnormal. If the payment amount is abnormal (such as the payment amount is 0 yuan or does not match the total price of the goods), the system rejects the payment request and prompts the user to pay again. If the payment method is abnormal (such as frequent changes in payment methods), the system restricts its payment method or requires the user to select a specific payment method. Record detailed information on abnormal payments (such as payment time, payment method, payment amount, etc.) for subsequent analysis.

[0132] If the abnormal type is a product abnormality, the system sends an abnormal notification to the operator of the vending cabinet. If the quantity of the product is abnormal (for example, the user did not take the product but the payment was successful), the system locks the relevant products to prevent further operations, and performs inventory calibration on the products involved in the abnormal order to ensure the accuracy of the inventory data and record detailed information on the change in the quantity of the product for subsequent analysis. Targeted control and processing measures can be taken according to the abnormal type of the abnormal order, thereby improving the security and user experience of the vending cabinet system.

[0133] This embodiment can efficiently obtain the target data of users in the process of using the vending machine, and accurately identify abnormal orders and perform management and control through abnormal order recognition models or comparison with normal behavior data. It can improve the accuracy of abnormal order recognition and reduce operating costs, thereby providing all-round support for the operation of the vending machine.

[0134] Reference Figure 4, showing a model training process diagram of a method for processing abnormal behavior orders of a vending machine provided by an embodiment of the present invention; the abnormal order recognition model can be trained in the following manner: obtaining target data of multiple users during the use of the vending machine to create sample data; the sample data includes user behavior data, order payment data, and commodity change data; annotating the sample data to obtain annotated sample data; inputting the sample data and the annotations corresponding to the sample data into a deep learning model, training the deep learning model, and adjusting the model parameters according to the training results; after the deep learning model training is completed, the abnormal order recognition model is obtained.

[0135] In this embodiment, user behavior is obtained to create a sample database, including purchase behavior data: purchased commodity categories, quantities, prices, times, etc., from which user consumption and user historical behavior information can be analyzed, whether the purchase behavior of new users is abnormal, and whether there are major changes in the purchase habits, bills, and payment information of old users. Abnormal account payment information: multiple scans to try to open the door and take the goods in a failed or unpaid state, and multiple payment abnormalities in a short period of time. Abnormal interaction sample data: users frequently operate the door in a short period of time (such as within 1 minute), cover the camera, and other malicious damage to the equipment. Abnormal data on taking goods: users frequently take out and put back goods in a short period of time (such as within 1 minute), or detect abnormal weight changes at the corresponding position through the gravity sensor, thereby recording abnormal sample data on taking goods.

[0136] Analyze and match the sample data. Obtain order video resources through surveillance cameras. Use image recognition detection algorithm technology to detect the task target in each frame of the video or image, and classify and identify the abnormal images. Use REID algorithm visual technology to determine the human motion feature extraction in the image or video. Use gravity sensors and position sensors to determine the range movement data of the product location.

[0137] Mark the behavior label based on the training model. Use deep learning algorithms (such as neural networks) to identify the user's behavior trajectory inside and outside the cabinet. For example, monitor uncivilized behaviors and actions such as violent actions on the cabinet lock, putting in foreign objects, deliberately blocking the camera, and frequently opening and closing the door. Obtain the model through data, training, and classification; use a large number of action video images simulating abnormal behaviors or through the character target model parameters in each frame or image of the video, so that the model can capture the rules in the data. By classifying images, for example, marking different images (such as blocking the camera, door opening and closing speed and duration), training the features of such images, such as blocking the camera screen will suddenly go black, the door opening and closing time is less than 1 second, and the door lock feedback fails within 10 seconds, etc., and finally achieve accurate judgment. Thereby identifying whether the pending order is an abnormal order.

[0138] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0139] Reference Figure 5 , shows a structural block diagram of a device for processing abnormal behavior orders of a vending cabinet provided by an embodiment of the present invention, which may specifically include the following modules:

[0140] The data acquisition module 301 is used to acquire target data of the user during the use of the vending cabinet; the target data includes at least one of user behavior data, order payment data, and commodity change data;

[0141] A first order determination module 302, configured to determine whether the order is an abnormal order according to the target data by using an abnormal order identification model;

[0142] The order processing module 303 is used to perform control processing if the order is an abnormal order.

[0143] Optionally, the data acquisition module 301 includes:

[0144] A video acquisition submodule, used to acquire video data acquired by an image acquisition module provided at the vending cabinet;

[0145] A behavior data determination submodule, which determines user behavior data based on the video data;

[0146] A payment data acquisition submodule is used to acquire order information and to acquire order payment data according to the order information;

[0147] The commodity data acquisition submodule is used to acquire commodity change data detected by the sensor installed in the vending cabinet.

[0148] Optionally, the first order determination module 302 includes:

[0149] An abnormality type determination submodule is used to determine whether the order is an abnormal order according to the target data through an abnormal order recognition model, and if the order is determined to be an abnormal order, determine the abnormality type;

[0150] The order processing module 303 includes:

[0151] The control and processing submodule is used to perform control and processing according to the exception type if the order is an abnormal order.

[0152] Optionally, the abnormality types include: uncivilized shopping, abnormal transactions, and abnormal commodities; the control and processing submodule includes:

[0153] The abnormal order processing unit is used to control the vending cabinet to lock the cabinet and impose a credit penalty on the user if the abnormal type of the abnormal order is the uncivilized shopping; if the abnormal type of the abnormal order is the transaction abnormality, mark the user with an abnormal behavior label; if the abnormal type of the abnormal order is the commodity abnormality, send an abnormal notification to the operator of the vending cabinet.

[0154] Optionally, the first order determining module 302 further includes:

[0155] The order abnormality confirmation submodule is used to send the target data corresponding to the order to the operator of the sales cabinet if the abnormal order identification model determines that the order is an abnormal order, so that the operation and maintenance personnel can reconfirm the abnormal condition of the order.

[0156] Optionally, the abnormal order recognition model can be trained by the following modules:

[0157] A sample data acquisition module, used to acquire target data of multiple users during the use of the vending cabinet to create sample data; the sample data includes user behavior data, order payment data, and commodity change data;

[0158] A sample data labeling module, used to label the sample data to obtain labeled sample data;

[0159] A model training module, used to input the sample data and the annotations corresponding to the sample data into a deep learning model, train the deep learning model, and adjust the model parameters according to the training results;

[0160] The model determination module is used to obtain the abnormal order recognition model after the deep learning model training is completed.

[0161] Optionally, the device further comprises:

[0162] The second order confirmation module obtains normal behavior data collected by multiple image acquisition modules installed in the vending machine; the normal behavior data includes the number of goods picked up, the speed of goods picked up, and the length of time the goods are kept for picking up; the normal behavior data is compared with the user behavior data of the current user during the use of the vending machine, and whether the order is an abnormal order is determined based on the comparison result.

[0163] This embodiment can efficiently obtain the target data of the user in the process of using the vending cabinet through the modular design of the data acquisition module, the first order determination module and the order processing module, and determine whether the order is an abnormal order through the abnormal order identification model, and perform management and control when the order is an abnormal order. This modular design can improve the accuracy of abnormal order identification, enhance the security of the vending cabinet, and optimize the user experience.

[0164] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0165] An embodiment of the present invention further provides an electronic device, including:

[0166] It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, the various processes of the above-mentioned embodiment of processing abnormal behavior orders of a vending machine are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0167] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes of the above-mentioned embodiment of processing abnormal behavior orders of a vending machine are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0168] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0169] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of 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 embodiments of 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 codes.

[0170] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0171] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0172] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0173] 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 have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0174] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0175] The above is a detailed introduction to a method for processing abnormal behavior orders of a sales cabinet and a device for processing abnormal behavior orders of a sales cabinet provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for processing abnormal behavior orders of a vending cabinet, characterized in that: The method comprises: Acquire target data of the user during use of the vending machine; the target data includes at least one of user behavior data, order payment data, and commodity change data; According to the target data, determining whether the order is an abnormal order through an abnormal order identification model; If the order is an abnormal order, control and management will be carried out.

2. The method for processing abnormal behavior orders of a vending machine according to claim 1, characterized in that: The step of obtaining target data of the user during use of the vending machine includes: Acquire video data collected by an image acquisition module provided at the vending cabinet; Determining user behavior data based on the video data; Obtaining order information, and extracting order payment data according to the order information; Obtain commodity change data detected by a sensor disposed in the vending cabinet.

3. The method for processing abnormal behavior orders of a vending machine according to claim 1, characterized in that: The determining whether the order is an abnormal order by using an abnormal order identification model according to the target data includes: According to the target data, determining whether the order is an abnormal order through an abnormal order identification model, and if it is determined that the order is an abnormal order, determining the abnormality type; If the order is an abnormal order, control and management are performed, including: If the order is an abnormal order, control and management will be performed according to the abnormal type.

4. The method for processing abnormal behavior orders of a vending machine according to claim 3, characterized in that: The abnormal types include: uncivilized shopping, abnormal transactions, and abnormal commodities; If the order is an abnormal order, control and management are performed according to the abnormal type, including: If the abnormal type of the abnormal order is uncivilized shopping, the vending cabinet is controlled to be locked and a credit penalty is imposed on the user; If the abnormal type of the abnormal order is the transaction abnormality, an abnormal behavior label is marked on the user; If the abnormal type of the abnormal order is that the product is abnormal, an abnormal notification is sent to the operator of the vending cabinet.

5. The method for processing abnormal behavior orders of a vending machine according to claim 3, characterized in that: The determining whether the order is an abnormal order by using an abnormal order identification model according to the target data further includes: If the abnormal order identification model determines that the order is an abnormal order, the target data corresponding to the order is sent to the operator of the vending machine so that the operation and maintenance personnel can reconfirm the abnormal condition of the order.

6. According to the method for processing abnormal behavior orders of a vending machine according to claim 1, the abnormal order recognition model can be trained in the following way: Acquire target data of multiple users in the process of using the vending machine to create sample data; the sample data includes user behavior data, order payment data, and commodity change data; Annotating the sample data to obtain annotated sample data; Inputting the sample data and the annotations corresponding to the sample data into a deep learning model, training the deep learning model, and adjusting model parameters according to the training results; After the deep learning model training is completed, the abnormal order recognition model is obtained.

7. The method for processing abnormal behavior orders of a vending machine according to claim 1, characterized in that: The method further comprises: Acquire normal behavior data collected by multiple image acquisition modules provided in the vending cabinet; the normal behavior data includes the number of goods taken, the speed of taking goods, and the length of time the goods are taken; The normal behavior data is compared with the user behavior data of the current user during the use of the vending machine, and whether the order is an abnormal order is determined based on the comparison result.

8. A device for processing abnormal behavior orders of a vending cabinet, characterized in that: The device comprises: A data acquisition module, used to acquire target data of the user during the use of the vending cabinet; the target data includes at least one of user behavior data, order payment data, and commodity change data; A first order determination module, configured to determine whether the order is an abnormal order according to the target data by using an abnormal order identification model; The order processing module is used to perform control and processing if the order is an abnormal order.

9. A vending device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein when the computer program is executed by the processor, the steps of a method for processing abnormal behavior orders of a vending machine as described in any one of claims 1-7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of a method for processing abnormal behavior orders of a vending machine according to any one of claims 1 to 7 are implemented.

Images