A settlement method and system for a vending machine

Abstract

The application discloses a kind of settlement method and system of vending machine, communication payment technical field is related, including image acquisition processing to user, obtain vending machine state information;According to vending machine state information, analysis and judgment processing is carried out, determine user behavior information, user behavior information includes placing goods information and taking goods information.The application monitors the behavior of user taking goods and placing goods, when the user determines goods, dynamic shooting equipment carries out image acquisition to the goods taken, judges the quantity and type of the goods taken, and calculates the settlement amount, when the user closes the vending machine, it represents that the purchase is finished, the user's payment account is debited according to the settlement amount, when the user completes payment, the vending machine is self-checked to ensure that the settlement amount will not be wrong, in addition, it also avoids the user to wait for the vending machine to verify the coin value when coin is put, and shortens the waiting time of the user.

Description

Technical Field

[0001] This invention relates to the field of communication payment technology, specifically to a settlement method and system for vending machines. Background Technology

[0002] Common types of vending machines include beverage vending machines, food vending machines, multi-purpose vending machines, and cosmetics vending machines. These machines are mainly placed in high-traffic areas such as train stations, bus stations, airports, schools, and shopping malls. The machines contain goods with corresponding prices, allowing users to choose and purchase items according to their preferences.

[0003] Existing vending machines are coin-operated, requiring consumers to insert coins and wait for the machine to process the coins before dispensing the product. If the coin's value is insufficient to pay for the product, another coin must be inserted, increasing the consumer's waiting time. Summary of the Invention

[0004] To solve the above-mentioned technical problems, a settlement method and system for vending machines is provided. This technical solution addresses the problem mentioned in the background art that existing vending machines are coin-operated, requiring consumers to insert coins and wait for the vending machine to judge the coins before dispensing the goods. If the coin's value is insufficient to pay the price of the goods, more coins need to be inserted, increasing the consumer's waiting time.

[0005] To achieve the above objectives, the present invention adopts the technical solution as described in the claims:

[0006] A payment method for an automated vending machine, comprising:

[0007] Image acquisition and processing are performed on the user to obtain vending machine status information, wherein the vending machine status information is either the door is open or the door is closed;

[0008] The system analyzes and processes the vending machine status information, analyzes and monitors user behavior, and determines user behavior information, including information on placing and taking products.

[0009] User behavior information is processed by feature extraction to determine product status information;

[0010] The settlement amount is obtained by analyzing and processing the product status information and vending machine status information.

[0011] Perform a self-inspection operation on the products inside the vending machine to obtain the remaining quantity of products;

[0012] Calculate the settlement amount based on the remaining quantity of goods and obtain the verification results;

[0013] The vending machine was inspected and repaired based on the verification results to determine its real-time status.

[0014] Preferably, the step of performing image acquisition and processing on the user to obtain the vending machine status information specifically includes the following steps:

[0015] The image of the user is captured by the camera to obtain the user's image information;

[0016] The user's payment account is verified based on the user's image information to obtain account login information, which can be either successful login or failed login.

[0017] Analyze and process account login information to obtain vending machine status information;

[0018] If the account login information is displayed as successful on the display device, the vending machine door will open.

[0019] If the account login information is displayed as login failed on the display device, the vending machine door closes, and the camera re-captures the user's image for a second login attempt.

[0020] Preferably, the step of analyzing and judging user behavior based on the vending machine status information to determine user behavior information specifically includes the following steps:

[0021] If the vending machine door is open, the dynamic shooting equipment monitors the user's behavior to determine the information on placing and taking products.

[0022] If the vending machine door is closed, the dynamic shooting equipment will not monitor user behavior.

[0023] Preferably, the step of performing feature extraction processing on user behavior information to determine product status information specifically includes the following steps:

[0024] Based on the information of the goods taken, feature extraction processing is performed to obtain the quantity of goods taken and the characteristics of the goods taken. The characteristics of the goods taken include the image of the outer packaging of the goods taken, the name of the goods taken, and the volume of the goods taken.

[0025] Obtain all product information inside the vending machine, wherein the product information includes product name, product packaging image, product volume, and product price;

[0026] Establish a feature matching model based on product name, product packaging image, product volume, and product price;

[0027] The product name is used as the first matching feature, the product packaging image is used as the second matching feature, and the product volume is used as the third matching feature.

[0028] Input the first matching feature into the feature matching model to determine the type of product to be retrieved and obtain the first matching result;

[0029] Based on the first matching result, the second matching feature is input into the feature matching model to determine the specific information of the product to be retrieved and obtain the second matching result.

[0030] Based on the second matching result, the third matching feature is input into the feature matching model to obtain the third matching result;

[0031] The prices of the goods to be taken are determined by analyzing and processing the first, second, and third matching results.

[0032] Preferably, the step of performing feature extraction processing based on the information of the goods taken to obtain the quantity and features of the goods taken specifically includes the following steps:

[0033] The quantity of goods taken is obtained by extracting quantity features from the information of the goods taken using a spot detection algorithm.

[0034] The text feature extraction process for the product information is performed using a convolutional neural network algorithm to obtain the product name and volume.

[0035] Image processing is performed on the product information using an image edge detection algorithm to obtain an image of the product's outer packaging.

[0036] Preferably, the step of analyzing and judging based on the product status information and the vending machine status information to obtain the settlement amount specifically includes the following steps:

[0037] The status information of the vending machine is judged and processed;

[0038] If the vending machine door is open, the dynamic shooting device continues to monitor the user's behavior and determine the real-time status information of the goods. The real-time status information of the goods includes either placing the goods a second time or taking the goods a second time.

[0039] The product status information is analyzed and processed again to obtain the quantity and price of the product taken a second time.

[0040] If the vending machine door is closed, the settlement amount is calculated based on the quantity and price of the items taken, and the quantity and price of the second item taken.

[0041] Preferably, the self-inspection operation of the goods inside the vending machine to obtain the remaining quantity of goods specifically includes the following steps:

[0042] Images of the products inside the vending machine are captured using dynamic imaging equipment.

[0043] The total remaining quantity of goods is obtained by processing the product image using a speckle detection algorithm.

[0044] Feature extraction processing is performed on product images to determine product type;

[0045] The remaining quantity of goods is categorized based on the total remaining quantity and the type of goods to obtain the remaining quantity.

[0046] Preferably, the step of calculating the settlement amount based on the remaining quantity of goods and obtaining the verification result specifically includes the following steps:

[0047] Obtain historical inventory data from vending machines;

[0048] Based on historical inventory data and remaining product quantity, calculations are performed to obtain product reduction information, which includes the amount of product reduction and the product type.

[0049] The decision-making process should be based on the amount of goods reduced and the quantity of goods taken.

[0050] If the quantity of goods reduced is the same as the quantity of goods taken, the first accounting amount is obtained by calculating based on the goods reduction information and the goods price;

[0051] The first verification result is obtained by calculating and processing the first accounting amount and the settlement amount.

[0052] If the amount of goods reduced is inconsistent with the number of goods taken, the vending machine is photographed again by a dynamic shooting device to obtain the information on the reduction of goods to be verified. The information on the reduction of goods to be verified includes the amount of goods reduced and the type of goods.

[0053] A second assessment is made based on the reduction in the quantity of goods to be verified and the quantity of goods taken. The second accounting amount is determined by calculating the quantity of goods taken and the price of the goods, and the second verification result is obtained.

[0054] Preferably, the step of inspecting the vending machine based on the verification results and determining the real-time status of the vending machine specifically includes the following steps:

[0055] The first verification result is then assessed and processed.

[0056] If the first verification result is greater than the set first threshold, the first calculated amount is greater than the settlement amount. The vending machine is then inspected, and the quantity and characteristics of the items taken are verified to determine whether the quantity and characteristics of the items taken by the user are consistent with the information on the reduction of items.

[0057] If the quantity and characteristics of the goods taken by the user are consistent with the information on the reduction of goods, output the correction information for the goods settlement algorithm;

[0058] If the quantity and characteristics of the goods taken by the user are inconsistent with the information on the reduction of goods, output the maintenance information of the dynamic camera equipment;

[0059] If the first verification result is equal to the set first threshold, output that the first calculated amount is equal to the settlement amount, and output the vending machine normal information.

[0060] If the first verification result is less than the set first threshold, output that the first calculated amount is less than the settlement amount. Verify the quantity of goods taken by the user, the characteristics of the goods taken, and the price of the goods. Determine the settlement amount to be verified. Calculate the settlement amount to be verified and the first calculated amount. If the calculation result is less than or equal to the set first threshold, output the goods settlement algorithm correction information.

[0061] The second verification result is then assessed and processed.

[0062] The system counts the number of goods inside the vending machine to determine the total remaining stock in real time, and then processes and judges the total remaining stock, the number of goods taken, and historical inventory data.

[0063] If the sum of the real-time total remaining quantity of goods and the quantity of goods taken is consistent with the historical inventory data, and the second accounting amount is consistent with the settlement amount, output the dynamic camera equipment maintenance information;

[0064] If the sum of the real-time total remaining quantity of goods and the quantity of goods taken is less than the historical inventory data, and the second accounting amount is consistent with the settlement amount, output the historical inventory data abnormality information;

[0065] If the sum of the real-time total remaining quantity of goods and the quantity of goods taken is consistent with the historical inventory data, but the second accounting amount is inconsistent with the settlement amount, output the goods settlement algorithm correction information;

[0066] If the sum of the real-time total remaining inventory and the number of goods taken is less than the historical inventory data, the second accounting amount is inconsistent with the settlement amount, and the historical inventory data abnormality information and the product settlement algorithm correction information are output.

[0067] Furthermore, a settlement system for an automated vending machine is proposed to implement the settlement method for an automated vending machine as described above, including:

[0068] A camera device, used to capture user image information;

[0069] An account login module is used to verify user image information and log in to the user's payment account.

[0070] A dynamic shooting device, used to monitor user behavior and acquire user behavior information;

[0071] The information analysis module is used to analyze user behavior information to determine the quantity of goods taken, the characteristics of the goods taken, and the placement of the goods.

[0072] The model building module is used to build a feature matching model based on the product name, product packaging image, product volume, and product price, and to confirm the price of the product to be taken.

[0073] The status analysis module is used to determine the status of the vending machine and whether to generate a settlement amount.

[0074] The data storage module is used to store historical inventory data, total remaining quantity of goods, retrieved goods data, retrieved goods characteristics, and settlement amount;

[0075] The judgment module analyzes and judges historical inventory data, total remaining quantity of goods, data on goods taken, characteristics of goods taken, and settlement amount, and outputs dynamic camera equipment maintenance information, goods settlement algorithm correction information, and historical inventory data anomaly information.

[0076] A control module is used to control the information interaction between various modules.

[0077] Compared with the prior art, the present invention provides a payment method and system for vending machines, which has the following beneficial effects:

[0078] This invention monitors users' actions of taking and placing goods. Once a user selects a product, a dynamic imaging device captures images of the product being taken, determines the quantity and type of the product, and calculates the settlement amount. When the user closes the vending machine, the purchase is complete, and the payment is deducted from the user's payment account based on the settlement amount. After payment, the vending machine performs a self-check to ensure the settlement amount is accurate. Furthermore, it avoids the need for users to wait for the vending machine to verify the coin denomination when inserting coins, thus shortening the user's waiting time. Attached Figure Description

[0079] Figure 1 This is a flowchart illustrating steps S100-S700 in the settlement method for an automatic vending machine proposed in this invention.

[0080] Figure 2 This is a flowchart illustrating steps S101-S105 of the settlement method for an automatic vending machine proposed in this invention.

[0081] Figure 3 This is a flowchart illustrating steps S201-S202 in the settlement method for an automatic vending machine proposed in this invention.

[0082] Figure 4 This is a flowchart illustrating steps S301-S308 of the settlement method for an automatic vending machine proposed in this invention.

[0083] Figure 5 This is a flowchart illustrating steps S3011-S3013 in the settlement method for an automatic vending machine proposed in this invention.

[0084] Figure 6 This is a flowchart illustrating steps S401-S402 in the settlement method for an automatic vending machine proposed in this invention.

[0085] Figure 7 This is a flowchart illustrating steps S501-S504 in the settlement method for an automatic vending machine proposed in this invention.

[0086] Figure 8 This is a flowchart illustrating steps S601-S607 of the settlement method for an automatic vending machine proposed in this invention.

[0087] Figure 9 , Figure 10 This is a flowchart illustrating steps S701-S7012 in the settlement method for an automatic vending machine proposed in this invention.

[0088] Figure 11 This is a structural block diagram of a payment system for an automatic vending machine proposed in this invention. Detailed Implementation

[0089] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0090] Reference Figure 1 As shown, a payment method for an automatic vending machine includes:

[0091] S100. Perform image acquisition and processing on the user to obtain vending machine status information, wherein the vending machine status information is either the door is open or the door is closed.

[0092] S200: Analyze and process the vending machine status information, analyze and monitor user behavior, and determine user behavior information, including information on placing and taking products.

[0093] S300: Perform feature extraction processing on user behavior information to determine product status information;

[0094] S400: Analyze and process the product status information and vending machine status information to obtain the settlement amount;

[0095] S500 performs a self-inspection of the products inside the vending machine to obtain the remaining quantity of products;

[0096] S600: Calculate the settlement amount based on the remaining quantity of goods and obtain the verification result;

[0097] S700. Based on the verification results, inspect the vending machine and determine its real-time status.

[0098] Those skilled in the art will understand that the process involves capturing the user's image, logging into the user's virtual payment account, and opening the vending machine door upon successful login. If login fails, the vending machine door cannot be opened. The system analyzes the user's product selection behavior using dynamic camera equipment, determines the price of the product based on the analysis, and deducts the corresponding amount from the user's virtual payment account after the user closes the vending machine door. The deducted amount is verified against the products inside the vending machine using a judgment module and a control module. If the deducted amount does not match the calculated amount, staff will go to the site to inspect and repair the vending machine.

[0099] Reference Figure 2 As shown, the process of acquiring and processing images of users to obtain vending machine status information includes the following steps:

[0100] S101. Capture images of the user using a camera to obtain user image information;

[0101] S102. Verify the user's payment account based on the user's image information and obtain account login information, wherein the account login information is either successful login or failed login.

[0102] S103. Analyze and process the account login information to obtain the vending machine status information;

[0103] S104. If the account login information is displayed as successful login on the display device, the vending machine cabinet door opens;

[0104] S105. If the account login information is displayed as login failure on the display device, the vending machine door is closed, and the camera device re-captures the user's image for a second login.

[0105] In this embodiment, if the user's payment account login fails, the vending machine door cannot be opened, ensuring the safety of the goods inside the vending machine and reducing property loss. Only when the user's payment account is successfully logged in can the vending machine door be opened to select goods.

[0106] Reference Figure 3 As shown, the process of analyzing and processing vending machine status information, analyzing and monitoring user behavior, and determining user behavior information specifically includes the following steps:

[0107] S201. If the vending machine door is open, the dynamic shooting device monitors the user's behavior to determine the information on placing and taking goods.

[0108] S202. If the vending machine door is closed, the dynamic shooting equipment will not monitor user behavior.

[0109] Reference Figure 4 As shown, the specific steps for extracting features from user behavior information to determine product status information include the following:

[0110] S301. Perform feature extraction processing based on the information of the goods taken to obtain the quantity of goods taken and the features of the goods taken, wherein the features of the goods taken include the image of the outer packaging of the goods taken, the name of the goods taken, and the volume of the goods taken.

[0111] S302. Obtain all product information inside the vending machine, wherein the product information includes product name, product packaging image, product volume, and product price;

[0112] S303. Establish a feature matching model based on the product name, product packaging image, product volume, and product price;

[0113] S304. Take the product name as the first matching feature, take the product outer packaging image as the second matching feature, and take the product volume as the third matching feature.

[0114] S305. Input the first matching feature into the feature matching model, determine the type of product to be retrieved, and obtain the first matching result.

[0115] S306. Based on the first matching result, input the second matching feature into the feature matching model to determine the specific information of the product to be retrieved and obtain the second matching result;

[0116] S307. Based on the second matching result, input the third matching feature into the feature matching model to obtain the third matching result;

[0117] S308. Analyze and process the first matching result, the second matching result and the third matching result to determine the price of the goods to be taken;

[0118] Those skilled in the art will understand that products from the same merchant may have the same name, but their volume and images may differ. Furthermore, products from different merchants may have similar packaging images but different names, potentially leading to price differences. Therefore, a feature matching model is established based on the product name, packaging image, volume, and price. Images of the product in the user's hand are then captured using dynamic imaging equipment. Feature acquisition determines the product's packaging image, name, and volume. These three elements are input into the feature matching model. Only when all three matching features are identical can the product price be determined. This avoids errors during subsequent checkout, preventing under- or over-deductions and ensuring the accuracy of vending machine payments.

[0119] Reference Figure 5 As shown, the process of extracting features based on the information of the goods taken to obtain the quantity and features of the goods taken includes the following steps:

[0120] S3011. The quantity of goods taken is obtained by performing quantity feature extraction processing on the information of the goods taken through the spot detection algorithm;

[0121] S3012. Use a convolutional neural network algorithm to extract text features from the information of the retrieved goods to obtain the name and volume of the retrieved goods.

[0122] S3013. The image of the product being picked up is processed using an image edge detection algorithm to obtain an image of the outer packaging of the product being picked up.

[0123] Reference Figure 6 As shown, the process of analyzing and processing product status information and vending machine status information to obtain the settlement amount includes the following steps:

[0124] S401. Perform judgment and processing on the status information of the vending machine;

[0125] If the vending machine door is open, the dynamic shooting device continues to monitor the user's behavior and determine the real-time status information of the goods. The real-time status information of the goods includes either placing the goods a second time or taking the goods a second time.

[0126] S402. Analyze and process the real-time status information of the goods again to obtain the quantity and price of the goods taken a second time;

[0127] If the vending machine door is closed, the settlement amount is calculated based on the quantity and price of the items taken, and the quantity and price of the second item taken.

[0128] Reference Figure 7 As shown, the self-checking operation of the vending machine to obtain the remaining quantity of goods includes the following steps:

[0129] S501. Acquire product images by capturing images of the products inside the vending machine using dynamic shooting equipment;

[0130] S502. Perform image processing on the product image using a speckle detection algorithm to obtain the total remaining quantity of the product;

[0131] S503. Perform feature extraction processing on the product image to determine the product type;

[0132] S504. Classify and process the remaining goods according to the total remaining quantity and the type of goods to obtain the remaining quantity of goods;

[0133] In this embodiment, after the user finishes selecting goods and closes the vending machine door, the vending machine's contents are inventoried using a dynamic imaging device to determine the quantity of each item. When the inventory of a certain item is insufficient, a replenishment message is sent to the remote control center via the control module, ensuring that the vending machine has an ample supply of goods. In addition, since the same merchant may produce many types of goods, but the packaging may be the same, it is necessary to extract keyword features from the product images, classify different products from the same merchant, and inventory their quantities. The remaining quantity of goods is a data set corresponding to the quantity of different goods, with each item corresponding to a specific quantity.

[0134] Reference Figure 8 As shown, the process of calculating the settlement amount based on the remaining quantity of goods and obtaining the verification result includes the following steps:

[0135] S601. Obtain historical inventory data of vending machines;

[0136] S602. Based on historical inventory data and remaining product quantity, calculate and process to obtain product reduction information, which includes the product reduction amount and product type;

[0137] S603. Make a judgment and process based on the amount of goods reduced and the number of goods taken;

[0138] S604. If the quantity of goods reduced is the same as the quantity of goods taken, calculate and process the amount of the first accounting amount by using the information on the reduction of goods and the price of the goods.

[0139] S605. Calculate and process the first accounting amount and settlement amount to obtain the first verification result;

[0140] S606. If the amount of goods reduced is inconsistent with the number of goods taken, the vending machine is photographed again by a dynamic shooting device to obtain the information on the reduction of goods to be verified. The information on the reduction of goods to be verified includes the amount of goods reduced and the type of goods.

[0141] S607. Based on the reduction in the quantity of goods to be verified and the quantity of goods taken, a second judgment is made. The second accounting amount is determined by calculating the quantity of goods taken and the price of the goods, and the second verification result is obtained.

[0142] Those skilled in the art will understand that by judging the remaining quantity of goods and the quantity of goods taken, and combining this with historical inventory data, it can be determined whether the sum of the remaining quantity of goods and the quantity of goods taken after the user purchases goods equals the historical inventory data. If they do not equal, it may be due to an error in registering the historical inventory data, or it may be due to an error in the dynamic camera equipment collecting the quantity of goods taken by the user. The cause of the error needs to be eliminated. If the sum of the remaining quantity of goods and the quantity of goods taken after the user purchases goods equals the historical inventory data, then the quantity of goods taken, the price of goods, and the settlement amount are compared to determine whether the actual amount of goods taken is consistent with the settlement amount. If they are inconsistent, the goods settlement algorithm is corrected.

[0143] Reference Figure 9 , 10 As shown, the process of inspecting and determining the real-time status of a vending machine based on the verification results includes the following steps:

[0144] S701. The first verification result is judged and processed;

[0145] S702. If the first verification result is greater than the set first threshold, output that the first calculated amount is greater than the settlement amount, inspect the vending machine, verify the quantity and characteristics of the goods taken, and determine whether the quantity and characteristics of the goods taken by the user are consistent with the information on the reduction of goods.

[0146] S703. If the quantity and characteristics of the goods taken by the user are consistent with the information on the reduction of goods, output the correction information for the goods settlement algorithm.

[0147] S704. If the quantity of goods taken by the user, the characteristics of the goods taken, and the information on the reduction of goods are inconsistent with the information on the reduction of goods, output the maintenance information of the dynamic camera equipment.

[0148] S705. If the first verification result is equal to the set first threshold, output that the first calculated amount is equal to the settlement amount, and output the vending machine normal information.

[0149] S706. If the first verification result is less than the set first threshold, output that the first calculated amount is less than the settlement amount, verify the quantity of goods taken by the user, the characteristics of the goods taken, and the price of the goods, determine the settlement amount to be verified, calculate the settlement amount to be verified and the first calculated amount, and if the calculation result is less than or equal to the set first threshold, output the goods settlement algorithm correction information.

[0150] Those skilled in the art will understand that if the first threshold is set to 0, and the first verification result is greater than 0, it means that the settlement amount is less than the actual deduction amount. This situation may be due to a problem with the product settlement algorithm or a problem with the image captured by the dynamic camera equipment, which requires the dynamic camera equipment to be repaired. If the first verification result is less than the set first threshold, it means that the settlement amount is greater than the actual deduction amount, which requires the product settlement algorithm to be corrected. In addition, the judgment and processing of the first verification result involves staff going to the vending machine to manually count the goods and determine the specific goods to be discussed and the quantity of goods to be taken.

[0151] S707. The second verification result is judged and processed.

[0152] S708. Inventory the quantity of goods inside the vending machine, determine the real-time total remaining quantity of goods, and make judgments and processes on the real-time total remaining quantity of goods, the quantity of goods taken, and historical inventory data.

[0153] S709. If the sum of the real-time total remaining quantity of goods and the quantity of goods taken is consistent with the historical inventory data, and the second accounting amount is consistent with the settlement amount, output the dynamic camera equipment maintenance information.

[0154] When the amount of goods reduced is inconsistent with the number of goods taken, staff go to the vending machine's location to inventory the goods inside. If the total remaining goods and the sum of the number of goods taken are consistent with the historical inventory data, and the second accounting amount is consistent with the settlement amount, it indicates that there is an abnormality in the dynamic camera equipment's inventory of the remaining goods inside the vending machine, and maintenance is required.

[0155] S7010. If the sum of the real-time total remaining quantity of goods and the quantity of goods taken is less than the historical inventory data, and the second accounting amount is consistent with the settlement amount, output the historical inventory data abnormality information.

[0156] Staff analyzed the images captured by the dynamic camera equipment on site, determined the quantity and price of the goods taken, and verified that the second accounting amount was consistent with the settlement amount. This indicated that there was an error in the data registration when placing goods into the vending machine, and the historical inventory data was modified.

[0157] S7011. If the sum of the real-time total remaining quantity of goods and the quantity of goods taken is consistent with the historical inventory data, but the second accounting amount is inconsistent with the settlement amount, output the goods settlement algorithm correction information.

[0158] S7012. If the sum of the real-time total remaining quantity of goods and the quantity of goods taken is less than the historical inventory data, the second accounting amount is inconsistent with the settlement amount. Output the historical inventory data abnormality information and the goods settlement algorithm correction information.

[0159] Reference Figure 11 As shown, a settlement system for an automatic vending machine is used to implement the settlement method for an automatic vending machine as described above, including:

[0160] A camera device, used to capture user image information;

[0161] An account login module is used to verify user image information and log in to the user's payment account.

[0162] A dynamic shooting device, used to monitor user behavior and acquire user behavior information;

[0163] The information analysis module is used to analyze user behavior information to determine the quantity of goods taken, the characteristics of the goods taken, and the placement of the goods.

[0164] The model building module is used to build a feature matching model based on the product name, product packaging image, product volume, and product price, and to confirm the price of the product to be taken.

[0165] The status analysis module is used to determine the status of the vending machine and whether to generate a settlement amount.

[0166] The data storage module is used to store historical inventory data, total remaining quantity of goods, retrieved goods data, retrieved goods characteristics, and settlement amount;

[0167] The judgment module analyzes and judges historical inventory data, total remaining quantity of goods, data on goods taken, characteristics of goods taken, and settlement amount, and outputs dynamic camera equipment maintenance information, goods settlement algorithm correction information, and historical inventory data anomaly information.

[0168] A control module, which is used to control the information interaction between the various modules;

[0169] In this embodiment, the vending machine cabinet is an intelligent door opener. Users can only open the vending machine cabinet door and select products by logging into their payment account.

[0170] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A settlement method of a vending machine, characterized by, The application comprises the following steps: image acquisition and processing of the user to obtain vending machine state information, which is any one of the cabinet door opening and the cabinet door closing; analysis and judgment processing according to the vending machine state information to analyze and monitor the user behavior and determine the user behavior information, which includes the placed goods information and the taken goods information; feature extraction processing of the user behavior information to determine the goods state information; analysis and judgment processing according to the goods state information and the vending machine state information to obtain the settlement amount; self-checking operation of the goods inside the vending machine to obtain the goods remaining amount; accounting operation of the settlement amount according to the goods remaining amount to obtain the verification result, which specifically comprises the following steps: obtaining the historical inventory data of the vending machine; calculating and processing according to the historical inventory data and the goods remaining amount to obtain the goods reduction information, which includes the goods reduction amount and the goods type; judgment processing according to the goods reduction amount and the taken goods quantity; if the goods reduction amount and the taken goods quantity are consistent, calculating and processing through the goods reduction information and the goods price to obtain the first accounting amount; calculating and processing according to the first accounting amount and the settlement amount to obtain the first verification result; if the goods reduction amount and the taken goods quantity are inconsistent, image acquisition of the vending machine again through the dynamic shooting device to obtain the to-be-verified goods reduction information, which includes the to-be-verified goods reduction amount and the goods type; second judgment according to the to-be-verified goods reduction amount and the taken goods quantity, calculating and processing through the taken goods quantity and the goods price to determine the second accounting amount and obtain the second verification result; and, according to the verification result, repairing the vending machine to determine the real-time state of the vending machine, wherein when the verification result indicates an abnormality, comprehensively analyzing and judging the historical inventory data, the real-time total goods remaining amount, the taken goods information and the settlement amount to determine the root cause of the abnormality and outputting specific repair information selected from the group consisting of dynamic camera repair information, goods settlement algorithm correction information and historical inventory data abnormal information.

2. The settlement method of an automatic vending machine according to claim 1, characterized by: The image acquisition and processing of the user to obtain the vending machine state information specifically comprises the following steps: image acquisition of the user through the shooting device to obtain user image information; verification processing of the user payment account according to the user image information to obtain account login information, which is any one of login success and login failure; analysis and processing according to the account login information to obtain the vending machine state information; if the account login information is displayed as login success on the display device, the vending machine cabinet door is opened; if the account login information is displayed as login failure on the display device, the vending machine cabinet door is closed and the shooting device re-acquires the image of the user for second login.

3. The settlement method of an automatic vending machine according to claim 2, characterized by, The analysis and judgment processing according to the vending machine state information to analyze and monitor the user behavior and determine the user behavior information specifically comprises the following steps: If the vending machine cabinet door is in an open state, the dynamic shooting device monitors the user behavior to determine the placed commodity information and the taken commodity information. If the vending machine cabinet door is in a closed state, the dynamic shooting device does not monitor the user behavior.

4. The settlement method of an automatic vending machine according to claim 3, characterized by, The feature extraction processing of the user behavior information to determine the commodity state information specifically includes the following steps: According to the taken commodity information, the feature extraction processing is performed to obtain the taken commodity quantity and the taken commodity features, and the taken commodity features include the taken commodity outer packaging image, the taken commodity name and the taken commodity volume. All commodity information in the vending machine is obtained, wherein the all commodity information includes the commodity name, the commodity outer packaging image, the commodity volume and the commodity price. A feature matching model is established according to the commodity name, the commodity outer packaging image, the commodity volume and the commodity price. The taken commodity name is taken as the first matching feature, the taken commodity outer packaging image is taken as the second matching feature, and the taken commodity volume is taken as the third matching feature. The first matching feature is input into the feature matching model to determine the taken commodity type and obtain the first matching result. According to the first matching result, the second matching feature is input into the feature matching model to determine the taken commodity specific information and obtain the second matching result. According to the second matching result, the third matching feature is input into the feature matching model to obtain the third matching result. According to the first matching result, the second matching result and the third matching result, the taken commodity price is determined.

5. The settlement method of an automatic vending machine according to claim 4, characterized in that, The feature extraction processing of the taken commodity information to obtain the taken commodity quantity and the taken commodity features specifically includes the following steps: The taken commodity information is subjected to quantity feature extraction processing by a spot detection algorithm to obtain the taken commodity quantity. The taken commodity information is subjected to text feature extraction processing by a convolutional neural network algorithm to obtain the taken commodity name and the taken commodity volume. The taken commodity information is subjected to image processing by an image edge detection algorithm to obtain the taken commodity outer packaging image.

6. The settlement method of an automatic vending machine according to claim 5, wherein The settlement amount is obtained by analyzing and judging the commodity state information and the vending machine state information, specifically including the following steps: The vending machine state information is subjected to judgment processing. If the vending machine cabinet door is in an open state, the dynamic shooting device continues to monitor the user behavior to determine the commodity real-time state information, and the commodity real-time state information includes any one of the second placed commodity and the second taken commodity. The second taken commodity quantity and the second taken commodity price are obtained by re-analyzing the commodity real-time state information. If the vending machine cabinet door is in a closed state, the settlement amount is obtained by calculating the taken commodity quantity, the taken commodity price, the second taken commodity quantity and the second taken commodity price.

7. The settlement method of an automatic vending machine according to claim 6, wherein The self-checking operation of the commodities in the vending machine to obtain the commodity remaining quantity specifically includes the following steps: The commodity image is obtained by image acquisition of the commodities in the vending machine through the dynamic shooting device. The total commodity remaining quantity is obtained by image processing of the commodity image through the spot detection algorithm. The commodity type is determined by feature extraction processing of the commodity image. According to the total amount of goods and commodity type classification processing, obtain commodity balance.

8. A settlement system of a vending machine for implementing the vending machine settlement method according to any one of claims 1 to 7, characterized by Comprise: Information acquisition module, the information acquisition module carries out image acquisition processing to the user, obtains the vending machine state information; Behavior monitoring module, the behavior monitoring module carries out analysis and judgment processing to the vending machine state information, analyzes and monitors the user behavior, determines the user behavior information; Feature extraction module, the feature extraction module carries out feature extraction processing to the user behavior information, determines the commodity state information; Settlement module, the settlement module carries out analysis and judgment processing according to the commodity state information and the vending machine state information, obtains the settlement amount; Self-checking module, the self-checking module carries out self-checking operation to the internal goods of vending machine, obtains the commodity balance; Accounting module, the accounting module carries out accounting operation to the settlement amount according to the commodity balance, obtains the verification result; And, the judgment module carries out analysis and judgment according to the historical inventory data, the total amount of goods, the data of taking goods, the characteristics of taking goods, the settlement amount, outputs any one or more of dynamic camera equipment maintenance information, commodity settlement algorithm correction information, historical inventory data exception information.

Images