A control method and system for unmanned retail terminal based on intelligent vending machine

Abstract

The present invention provides an unmanned retail terminal control method and system based on a smart vending machine, wherein the method includes: step 1: obtaining the shipping record of the smart vending machine; step 2: determining the aisle suggestion information of the first commodity based on the shipping record; step 3: when the replenisher replenishes the goods, based on the aisle suggestion information, controlling the smart vending machine to display a replenishment reminder message; step 4: obtaining video information within a preset range in front of the smart vending machine in real time, determining the target person based on the video information, and recommending goods to the target person. The unmanned retail terminal control method and system based on the smart vending machine of the present invention determines the aisle suggestion information based on the shipping record stored locally in the smart vending machine, thereby improving the suitability of the commodity arrangement of the first commodity; introducing the replenishment reminder message, without the need for manual statistics to determine the replenishment quantity, is more intelligent and convenient. Recommending goods to the target person based on video information improves the recommendation efficiency of the smart vending machine.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to an unmanned retail terminal control method and system based on an intelligent vending machine. Background Art

[0002] At present, when smart vending machines are placed in different locations, the quantity of different goods sold is also different. For example, at the entrance of a shopping mall, more masks are sold, and in a movie theater, more popcorn and drinking water are sold. Since it is not clear what kind of facilities are lacking or what kind of goods are in short supply around the location where the smart vending machine is placed, blindly arranging the same types of goods for sale in the vending machine can easily lead to inappropriate types of goods being sold. At the same time, manual inventory is required for replenishment, which is less convenient and not smart enough. In addition, when smart vending machines are equipped with display devices for product recommendations, they often make indiscriminate recommendations and have low recommendation efficiency.

[0003] Therefore, a solution is urgently needed. Summary of the invention

[0004] One of the purposes of the present invention is to provide an unmanned retail terminal control method based on a smart vending machine. Based on the shipping records stored locally in the smart vending machine, the aisle recommendation information is determined, thereby improving the suitability of the commodity arrangement of the first commodity; a replenishment reminder message is introduced to remind the replenishment personnel to replenish the corresponding amount of commodities, without the need for manual statistics to determine the replenishment quantity, which is more intelligent and convenient; commodity recommendations are made to target personnel based on video information, thereby improving the recommendation efficiency of the smart vending machine.

[0005] An embodiment of the present invention provides an unmanned retail terminal control method based on an intelligent vending machine, comprising:

[0006] Step 1: Obtain multiple shipment records stored locally in the smart vending machine;

[0007] Step 2: Based on the shipping record, determine the aisle recommendation information for the first product, wherein the aisle recommendation information includes: a recommended aisle position and a recommended number;

[0008] Step 3: When the replenisher replenishes the goods, based on the aisle suggestion information, the smart vending machine is controlled to display a replenishment reminder message at a preset display location;

[0009] Step 4: Obtain video information within a preset range in front of the smart vending machine in real time, determine the target person based on the video information, and recommend corresponding products to the target person.

[0010] Preferably, the step 2: determining the channel suggestion information of the first commodity based on the shipping record includes:

[0011] Parse the shipping records to obtain the shipping quantity of the first products corresponding to each shipping record;

[0012] Calculate the first occurrence number of the shipment number corresponding to the same first commodity and the same shipment number, and at the same time, calculate the second occurrence number of the shipment record corresponding to the same first commodity;

[0013] determining an adjustment factor based on the first number of occurrences and the corresponding second number of occurrences;

[0014] Multiply the adjustment coefficient by the corresponding shipment quantity to obtain the recommended value;

[0015] Accumulate and calculate each recommendation value to obtain the recommendation degree;

[0016] Sort the first product by recommendation degree from high to low to obtain a product recommendation sequence;

[0017] Determine a recommended aisle position for the first product based on the product recommendation sequence;

[0018] Based on the recommendation degree, the recommended number of channels is determined.

[0019] Preferably, determining the recommended position of the aisle of the first commodity based on the commodity recommendation sequence includes:

[0020] Obtain the historical residence time of the target person in the preset range in front of the smart vending machine;

[0021] If the dwell time is greater than or equal to a preset first threshold, obtaining the face position and face orientation of the historical target person, and constructing a viewing vector based on the face position and face orientation;

[0022] Obtaining a display orientation of a display area of ​​the smart vending machine, and constructing a display vector based on the display position and the display orientation;

[0023] During the dwell time, the vector angle between the viewing vector and the display vector is continuously calculated at a preset period;

[0024] Accumulate and calculate the angle of each vector to obtain the sum of the vector angles;

[0025] If the sum of the vector angles is greater than or equal to a preset second threshold, the first height of the corresponding historical target person is obtained;

[0026] Based on the first height, a recommended position of an aisle for the first product is determined.

[0027] Preferably, determining the recommended position of the aisle of the first commodity based on the first height includes:

[0028] Calculate an average height of the first heights, and determine that a second height at a preset distance below the average height corresponds to a horizontal position on the smart vending machine;

[0029] Obtaining a vertical distance between the position of each cargo lane and the horizontal position, obtaining a first reference value corresponding to the vertical distance, and associating it with the corresponding cargo lane;

[0030] Obtain the horizontal distance between each cargo lane and the middle cargo lane of the cargo lane row, determine a second reference value corresponding to the horizontal distance, and associate it with the corresponding cargo lane;

[0031] Accumulate and calculate the first reference value and the second reference value associated with each cargo channel to obtain a reference degree;

[0032] Traverse the first product in the product recommendation sequence from the beginning to the end, and use the first product being traversed as the second product;

[0033] When traversing to the second product, the target position of the currently unassigned aisle with the greatest reference degree is obtained and used as the recommended position corresponding to the second product.

[0034] Preferably, the step 3: when the replenisher replenishes the goods, based on the aisle suggestion information, controlling the smart vending machine to display a replenishment reminder message at a preset display position includes:

[0035] Continuously obtain machine status signals of smart vending machines;

[0036] Analyze the machine status signal, and when at least one replenishment signal is obtained, obtain the remaining quantity information in the intelligent vending machine, wherein the remaining quantity information includes: remaining goods and the number of remaining goods;

[0037] Generate replenishment reminder information based on the preset replenishment reminder information template and according to the remaining quantity information and channel recommendation information.

[0038] Preferably, target personnel are identified, including:

[0039] Based on the frame processing technology, the video information is processed in frames to obtain multiple frames of images;

[0040] Determine the target person based on multiple frames of images.

[0041] Preferably, the recommending corresponding products to the target person includes:

[0042] Based on the preset feature extraction template, the first character feature of the target person is obtained;

[0043] Querying a preset character type library to determine a first character type corresponding to the first character feature;

[0044] The target customer of the goods sold in the smart vending machine is obtained, and the second character type of the target customer is determined; if the first character type is consistent with any second character type, the goods sold corresponding to the second character type are used as recommended goods;

[0045] Generate recommended product information based on a preset recommended product information generation template and based on the recommended products;

[0046] Based on the preset recommended product information display rules, according to the recommended product information, the corresponding recommended product information is displayed at the display position and the corresponding target personnel are reminded to view it.

[0047] Preferably, the step of obtaining target customers of goods sold in the smart vending machine includes:

[0048] Obtain purchase records of multiple products for sale from the big data platform;

[0049] Parse the purchase records to obtain multiple one-to-one correspondences between sold goods and purchasing users;

[0050] Obtain the preset rejection condition feature library;

[0051] Extract purchase features of purchase records;

[0052] Match the purchase features with the elimination features in the elimination condition feature library, determine the matching purchase records and eliminate them;

[0053] Split the remaining purchase records to obtain the products sold and the target customers corresponding to the products sold.

[0054] An embodiment of the present invention provides an unmanned retail terminal control system based on an intelligent vending machine, which is characterized by comprising:

[0055] An acquisition module is used to acquire multiple shipment records stored locally in the smart vending machine;

[0056] A determination module, configured to determine, based on the shipping record, aisle recommendation information for the first commodity, wherein the aisle recommendation information includes: a recommended location and a recommended number of aisles;

[0057] A control module, used to control the smart vending machine to display a replenishment reminder message at a preset display position based on the aisle suggestion information when the replenisher replenishes the goods;

[0058] The recommendation module is used to obtain video information within a preset range in front of the smart vending machine in real time, determine the target person based on the video information, and recommend corresponding products to the target person.

[0059] Preferably, the determining module determines the channel suggestion information of the first commodity based on the shipping record, including:

[0060] Parse the shipping records to obtain the shipping quantity of the first products corresponding to each shipping record;

[0061] Calculate the first occurrence number corresponding to the shipment number, and at the same time, calculate the second occurrence number corresponding to each shipment number;

[0062] determining an adjustment factor based on the second number of occurrences and the first number of occurrences;

[0063] Multiply the adjustment coefficient by the corresponding second occurrence number to obtain the recommended value;

[0064] Accumulate and calculate each recommendation value to obtain the recommendation degree;

[0065] Sort the first product by recommendation degree from high to low to obtain a product recommendation sequence;

[0066] Determine a recommended aisle position for the first product based on the product recommendation sequence;

[0067] Based on the recommendation degree, the recommended number of cargo lanes is determined. Other features and advantages of the present invention will be described in the subsequent description, and in part will become apparent from the description, or will be understood by practicing the present invention. The objects and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0068] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0070] Figure 1 Schematic diagram of an unmanned retail terminal control method based on a smart vending machine in an embodiment of the present invention;

[0071] Figure 2 The figure is a schematic diagram of an unmanned retail terminal control system based on a smart vending machine in an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0073] The embodiment of the present invention provides an unmanned retail terminal control method based on an intelligent vending machine, such as Figure 1 As shown, including:

[0074] Step 1: Obtain multiple shipment records stored locally in the smart vending machine;

[0075] Step 2: Based on the shipping record, determine the aisle recommendation information for the first product, wherein the aisle recommendation information includes: a recommended aisle position and a recommended number;

[0076] Step 3: When the replenisher replenishes the goods, based on the aisle suggestion information, the smart vending machine is controlled to display a replenishment reminder message at a preset display location;

[0077] Step 4: Obtain video information within a preset range in front of the smart vending machine in real time, determine the target person based on the video information, and recommend corresponding products to the target person.

[0078] The working principle and beneficial effects of the above technical solution are:

[0079] The shipping record is: what kind of goods and what quantity were sold at what time, for example: two bottles of drinking water were sold at x. Based on the shipping record, determine the aisle recommendation information of the first product; the aisle recommendation information includes: the recommended position and recommended number of the aisle of the first product. The recommended position is: the aisle in which the first product is recommended to be placed, and the recommended number is: how many aisles are recommended for the first product, for example: two aisles of potato chips are placed in aisles 4 rows and 4 columns and 4 rows and 5 columns. Generally, the better the sales of the first product, the more conspicuous the recommended position, and the more recommended number. When the replenisher replenishes the goods, based on the aisle recommendation information, control the smart vending machine to display a replenishment reminder message at a preset display position (for example: at the preset display device of the smart vending machine); the replenishment reminder message is: the replenishment reminder corresponding to the aisle recommendation.

[0080] Usually, smart vending machines are generally equipped with display screens, and vending machines are generally placed in places with dense traffic such as business districts and communities. When there is no demand for vending machines, the display screen can be used for advertising, but all-weather advertising will cause a waste of power resources. Therefore, video information within a preset range in front of the smart vending machine is obtained in real time, and based on the video information, the target person is determined to recommend goods; the preset range is set manually; the video information is: image messages of activities or movements captured by a shooting device (e.g., a camera) pre-set on the smart vending machine.

[0081] This application determines the aisle recommendation information based on the shipping records stored locally in the smart vending machine, thereby improving the suitability of the product arrangement of the first product; introduces a replenishment reminder message to remind the replenishment personnel to replenish the corresponding number of products, without the need for manual statistics to determine the replenishment quantity, which is more intelligent and convenient; recommends products to target personnel based on video information, thereby improving the recommendation efficiency of the smart vending machine.

[0082] In one embodiment, the step 2: determining the channel recommendation information of the first commodity based on the shipping record includes:

[0083] Parse the shipping records to obtain the shipping quantity of the first products corresponding to each shipping record;

[0084] Calculate the first occurrence number of the shipment number corresponding to the same first commodity and the same shipment number, and at the same time, calculate the second occurrence number of the shipment record corresponding to the same first commodity;

[0085] determining an adjustment factor based on the first number of occurrences and the corresponding second number of occurrences;

[0086] Multiply the adjustment coefficient by the corresponding shipment quantity to obtain the recommended value;

[0087] Accumulate and calculate each recommendation value to obtain the recommendation degree;

[0088] Sort the first product by recommendation degree from high to low to obtain a product recommendation sequence;

[0089] Determine a recommended aisle position for the first product based on the product recommendation sequence;

[0090] Based on the recommendation degree, the recommended number of channels is determined.

[0091] The working principle and beneficial effects of the above technical solution are:

[0092] Parse the shipment records, obtain multiple first commodities and shipment numbers corresponding to each other, and associate them with the corresponding shipment records. In particular, when a certain vending machine has a large shipment volume of a certain commodity, it is unreasonable to blindly believe that the commodity is suitable for sale. Some emergencies may occur, for example: a company meeting urgently needs water, but the company logistics has not purchased it in time, so a large amount of drinking water needs to be purchased. This purchase demand is not continuous and has low reference value. Therefore, the first occurrence number of shipments corresponding to the same first commodity and the same number of shipments is calculated (for example: 6 times for buying 1 bottle of water, 3 times for buying 2 bottles of water, and 1 time for buying 10 bottles of water). At the same time, calculate the second occurrence number of shipment records corresponding to the same first commodity (for example: 10 times). Based on the first occurrence number and the corresponding second occurrence number, determine the adjustment coefficient (when determining, the first occurrence number is divided by the corresponding second occurrence number). Multiply the adjustment coefficient and the corresponding number of shipments to obtain the recommended value. Cumulatively calculate each recommended value to obtain the recommendation degree (the higher the recommendation degree, the more the number of recommended aisles corresponding to the first commodity and the closer the recommended aisle is to the middle). Sort the first product by recommendation degree from high to low, and determine the recommended position of the first product according to the obtained product recommendation sequence (the earlier the first product is in the product recommendation sequence, the more prominent the aisle position should be). Based on the recommendation degree, determine the number of recommended aisles (the higher the recommendation degree, the more recommended aisles there are).

[0093] This application introduces an adjustment coefficient to determine the recommendation degree based on the shipment quantity of the first product, thereby improving the accuracy of obtaining the recommendation degree; introduces a product recommendation sequence to determine the channel recommendation information of the first product based on the order of the first product in the product sequence, which is more appropriate.

[0094] In one embodiment, determining the recommended position of the aisle of the first commodity based on the commodity recommendation sequence includes:

[0095] Obtain the historical residence time of the target person in the preset range in front of the smart vending machine;

[0096] If the dwell time is greater than or equal to a preset first threshold, obtaining the face position and face orientation of the historical target person, and constructing a viewing vector based on the face position and face orientation;

[0097] Obtaining a display orientation of a display area of ​​the smart vending machine, and constructing a display vector based on the display position and the display orientation;

[0098] During the dwell time, the vector angle between the viewing vector and the display vector is continuously calculated at a preset period;

[0099] Accumulate and calculate the angle of each vector to obtain the sum of the vector angles;

[0100] If the sum of the vector angles is greater than or equal to a preset second threshold, the first height of the corresponding historical target person is obtained;

[0101] Based on the first height, a recommended position of an aisle for the first product is determined.

[0102] The working principle and beneficial effects of the above technical solution are:

[0103] Obtain the residence time of the historical target person in the preset range in front of the smart vending machine. The historical target person is: the person who has stopped in front of the smart vending machine in history. The preset range is: within 2m, and the residence time is: the length of time the historical target person stays in front of the smart vending machine (for example: 3 minutes). Since not all people passing in front of the smart vending machine buy the goods sold in the vending machine, if the residence time is greater than or equal to the preset first threshold (for example: 3 minutes), then obtain the face position and face orientation of the historical target person, and construct a viewing vector based on the face position and face orientation. Construct a display vector based on the display position and the display area of ​​the vending machine (for example: the LED screen for displaying information at the preset position of the vending machine) (the vector construction technology belongs to the prior art and will not be repeated). Obtain the vector angle (for example: 170 degrees) between the viewing vector and the display vector during the residence time of the historical target person at a preset period (for example: 20 seconds). The vector angle sum of each vector angle is calculated, and if the vector angle sum is greater than or equal to a preset second threshold (the second threshold is manually preset), the first height (for example: 1.70m) of the corresponding historical target person is obtained. Based on the first height, the recommended position of the first product is determined.

[0104] The present application introduces dwell time to determine the first height of a historical target person who viewed a product and had a long dwell time, and determines a suitable recommended position for the first product based on the first height, thereby improving the rationality of determining the recommended position.

[0105] In one embodiment, determining a recommended aisle position for the first commodity based on the first height includes:

[0106] Calculate an average height of the first height, and determine a horizontal position of the smart vending machine corresponding to the average height;

[0107] Obtaining a vertical distance between the position of each cargo lane and the horizontal position, obtaining a first reference value of the vertical distance, and associating it with the corresponding cargo lane;

[0108] Obtain the horizontal distance between each cargo lane and the middle cargo lane of the cargo lane row, and use it as a second reference value and associate it with the corresponding cargo lane;

[0109] Accumulate and calculate the first reference value and the second reference value associated with each cargo channel to obtain a reference degree;

[0110] Traverse the first product in the product recommendation sequence from the beginning to the end, and use the first product being traversed as the second product;

[0111] When traversing to the second product, the target position of the currently unassigned aisle with the greatest reference degree is obtained and used as the recommended position corresponding to the second product.

[0112] The working principle and beneficial effects of the above technical solution are:

[0113] Generally, people will give priority to seeing the range at the same height as their eye height, and the goods that are directly in the sight range are more likely to be noticed by the purchaser. Since the heights of people who buy goods are different, it is necessary to determine the average height, and the position of the human eye is usually 15 cm below the height. Therefore, the second height at a preset distance below the average height (for example, 15 cm) is determined to correspond to the horizontal position on the smart vending machine. The above horizontal position is the height at which the human eye directly sees the vending machine. The farther the vertical distance from the horizontal position, the less likely the corresponding aisle is to be noticed by the customer. The first reference value corresponding to the vertical distance is obtained (the smaller the vertical distance, the larger the first reference value), and the first reference value is associated with the corresponding aisle. In each row of aisles, the middle aisle is the easiest to see. Therefore, the horizontal distance of each aisle from the middle aisle of the aisle row is obtained (the larger the horizontal distance, the less likely the corresponding aisle is to be seen), and the second reference value corresponding to the horizontal distance is determined (the smaller the horizontal distance, the larger the second reference value), and the second reference value is associated with the corresponding aisle. Add the first reference value and the second reference value associated with each aisle to obtain the reference degree (the larger the reference degree, the easier it is to see the corresponding aisle). From the beginning to the end, the first product in the product recommendation sequence is taken as the second product. The earlier the second product is traversed, the earlier it should be arranged in a conspicuous aisle. When the second product is traversed, the target position of the aisle with the largest unassigned reference degree is determined as the recommended position of the second product.

[0114] The present application introduces vertical distance and horizontal distance to determine the first reference value and the second reference value respectively, and calculates the reference degree based on the first reference value and the second reference value, thereby improving the comprehensiveness and rationality of obtaining the reference degree; determining the recommended position of the second product in the product recommendation sequence based on the reference degree is more appropriate.

[0115] In one embodiment, the step 3: when the replenisher replenishes the goods, based on the channel suggestion information, controlling the smart vending machine to display a replenishment reminder message at a preset display position includes:

[0116] Continuously obtain machine status signals of smart vending machines;

[0117] Analyze the machine status signal, and when at least one replenishment signal is obtained, obtain the remaining quantity information in the intelligent vending machine, wherein the remaining quantity information includes: remaining goods and the number of remaining goods;

[0118] Generate replenishment reminder information based on the preset replenishment reminder information template and according to the remaining quantity information and channel recommendation information.

[0119] The working principle and beneficial effects of the above technical solution are:

[0120] The machine status signal is an electrical signal that characterizes the device status of the smart vending machine. The above machine status signal is parsed, and when a replenishment signal is obtained (an electrical signal that characterizes the device status of the vending machine is a replenishment status), the balance information is obtained. The balance information includes: multiple groups of one-to-one corresponding remaining goods and the number of remaining goods (for example: two bags of potato chips, three bottles of mineral water, etc.). Based on the preset replenishment reminder information generation template, replenishment reminder information is generated according to the balance information and channel suggestion information (for example: please replenish 8 bags of potato chips, 17 bottles of mineral water); the preset replenishment reminder information generation template is: what kind of replenishment reminder information is generated based on what kind of balance information.

[0121] This application introduces a template for generating replenishment reminder information, which determines the replenishment reminder information used for replenishment reminders based on the remaining quantity information and channel recommendation information, thereby improving the standardization of the acquisition of replenishment reminder information.

[0122] In one embodiment, the unmanned retail terminal control method based on the intelligent vending machine, the step of determining the target person based on the video information, includes:

[0123] Based on the frame processing technology, the video information is processed in frames to obtain multiple frames of images;

[0124] Determine the target person based on multiple frames of images.

[0125] The working principle and beneficial effects of the above technical solution are:

[0126] When determining the target person, the video information needs to be decomposed into frames of images (based on frame processing technology, which will not be described in detail). The target person is determined based on multiple frames of images. The specific determination process is as follows:

[0127] Sort multiple frames of images according to shooting time to obtain a video sequence;

[0128] Get the gray value of each image in the video sequence;

[0129] Subtract the grayscale values ​​of the images of two adjacent frames to obtain the differential grayscale value function D of the differential image n (x,y);

[0130] D n (x,y)=|f n (x,y)-f n-1 (x,y)|

[0131] Among them, f n (x, y) is the gray value function of the nth frame image, f n-1 (x, y) is the gray value function of the n-1th frame image, |…| is the absolute value of…;

[0132] Binarize the difference image and get the processed image function R ′ n (x,y);

[0133]

[0134] Among them, T is a threshold value preset manually, the point with a gray value of 255 is the corresponding point of the target person, and the point with a gray value of 0 is the background point in the processed image except the corresponding point of the target person;

[0135] Connectivity processing is performed on the processed image to obtain a recognized image. Connectivity processing of an image belongs to the prior art and will not be elaborated on. After the connectivity processing is completed, the target person is determined from the processed image after the connectivity processing is completed.

[0136] This application introduces image frame processing technology to process video information, thereby improving the efficiency of subsequent face recognition; introducing face recognition technology to determine target personnel, rather than blindly placing advertisements 24 hours a day, reduces system energy consumption.

[0137] In one embodiment, the recommending corresponding products to the target person includes:

[0138] Based on the preset feature extraction template, the first character feature of the target person is obtained;

[0139] Querying a preset character type library to determine a first character type corresponding to the first character feature;

[0140] Obtain the target customers of the products sold in the smart vending machine and determine the second person type of the target customers;

[0141] If the first character type is consistent with any second character type, the sales product corresponding to the second character type is used as a recommended product;

[0142] Generate recommended product information based on a preset recommended product information generation template and based on the recommended products;

[0143] Based on the preset recommended product information display rules, according to the recommended product information, the corresponding recommended product information is displayed at the display position and the corresponding target personnel are reminded to view it.

[0144] The working principle and beneficial effects of the above technical solution are:

[0145] Based on the preset feature extraction template, the first character feature of the target person is obtained; the preset feature extraction template is: a pre-set template for extracting character features, the first character feature is, for example, height, hair color, etc. The preset character type library includes: multiple second character features and first character types, the first character type is, for example, the elderly, children and young people, etc., and another example is: men and women. The second character type of the target customer (target advertising customer) of the goods sold in the smart vending machine is obtained (for example: the goods sold are lollipops, and the second character type of the target customer is children). The first character type is consistent with any second character type, and the corresponding goods sold are used as recommended goods. Based on the preset recommended product information generation template, recommended product information is generated according to the recommended products: the preset product information generation template is, for example: the original price of the lollipop is xx, the current price is xx, etc. The preset recommended product information display rule is, for example: display on the display screen of the smart vending machine. Based on the preset recommended product information display rules, the corresponding recommended product information is displayed at the display position and the corresponding target person is reminded to view it. When reminding, a reminder voice can be played through the speaker, for example: "The original price of the lollipop is xx, and the current price is only xx. Please check the smart vending machine details page for details."

[0146] This application introduces a feature extraction template to extract the first character feature of the target person, thereby improving the accuracy of obtaining the first character feature; introduces a character type library and a second character type of the target customer to determine recommended products, thereby improving the rationality of obtaining recommended products; introduces a recommended product information generation template and a recommended product information display rule to generate and display corresponding product information and remind people to check, which is more humane.

[0147] In one embodiment, obtaining target customers of goods sold in the smart vending machine includes:

[0148] Obtain purchase records of multiple products for sale from the big data platform;

[0149] Parse the purchase records to obtain multiple one-to-one correspondences between sold goods and purchasing users;

[0150] Obtain the preset rejection condition feature library;

[0151] Extract purchase features of purchase records;

[0152] Match the purchase features with the elimination features in the elimination condition feature library, determine the matching purchase records and eliminate them;

[0153] Split the remaining purchase records to obtain the products sold and the target customers corresponding to the products sold.

[0154] The working principle and beneficial effects of the above technical solution are:

[0155] When obtaining the purchase records corresponding to the sold goods, you can obtain them from the big data platform to determine the multiple sold goods and purchasing users in the purchase records. The exclusion condition feature library includes multiple exclusion condition features for elimination, for example: if the payment method is purchasing on behalf of others, then the corresponding purchasing user is not the target user. Extract the purchase features of the purchase records, such as payment method, purchasing user information, etc. Match the purchase features with the exclusion features in the exclusion condition feature library, determine the matching purchase records and eliminate them. Split the remaining purchase records after elimination, determine the sold goods and the purchasing users corresponding to the sold goods, and use the purchasing users as the target customers of the corresponding sold goods.

[0156] This application introduces a rejection condition feature library to first reject purchase records that meet the rejection conditions before acquiring target customers for selling goods, thereby improving the efficiency of acquiring target customers.

[0157] The embodiment of the present invention provides an unmanned retail terminal control system based on an intelligent vending machine, such as Figure 2 As shown, including:

[0158] Acquisition module 1, used to acquire multiple shipment records stored locally in the smart vending machine;

[0159] Determining module 2, used to determine the cargo lane recommendation information of the first commodity based on the shipping record, wherein the cargo lane recommendation information includes: a recommended location and a recommended number of cargo lanes;

[0160] The control module 3 is used to control the smart vending machine to display a replenishment reminder message at a preset display position based on the aisle suggestion information when the replenisher replenishes the goods;

[0161] The recommendation module 4 is used to obtain video information within a preset range in front of the smart vending machine in real time, determine the target person based on the video information, and recommend corresponding products to the target person.

[0162] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A control method for an unmanned retail terminal based on an intelligent vending machine, characterized in that: include: Step 1: Obtain multiple shipment records stored locally in the smart vending machine; Step 2: Based on the shipping record, determine the aisle recommendation information for the first product, wherein the aisle recommendation information includes: a recommended aisle position and a recommended number; Step 3: When the replenisher replenishes the goods, based on the aisle suggestion information, the smart vending machine is controlled to display a replenishment reminder message at a preset display location; Step 4: Obtain video information within a preset range in front of the smart vending machine in real time, determine the target person based on the video information, and recommend corresponding products to the target person; The step 2: determining the channel recommendation information of the first commodity based on the shipping record includes: Parse the shipping records to obtain the shipping quantity of the first products corresponding to each shipping record; Calculate the first occurrence number of the shipment number corresponding to the same first commodity and the same shipment number, and at the same time, calculate the second occurrence number of the shipment record corresponding to the same first commodity; determining an adjustment factor based on the first number of occurrences and the corresponding second number of occurrences; Multiply the adjustment coefficient by the corresponding shipment quantity to obtain the recommended value; Accumulate and calculate each recommendation value to obtain the recommendation degree; Sort the first product by recommendation degree from high to low to obtain a product recommendation sequence; Determine a recommended aisle position for the first product based on the product recommendation sequence; Based on the recommendation degree, the recommended number of channels is determined.

2. The unmanned retail terminal control method based on the intelligent vending machine according to claim 1, characterized in that: The step of determining a recommended location of a goods channel for the first goods based on the goods recommendation sequence includes: Obtain the historical residence time of the target person in the preset range in front of the smart vending machine; If the dwell time is greater than or equal to a preset first threshold, obtaining the face position and face orientation of the historical target person, and constructing a viewing vector based on the face position and face orientation; Obtaining a display orientation of a display area of ​​the smart vending machine, and constructing a display vector based on the display position and the display orientation; During the dwell time, the vector angle between the viewing vector and the display vector is continuously calculated at a preset period; Accumulate and calculate the angle of each vector to obtain the sum of the vector angles; If the sum of the vector angles is greater than or equal to a preset second threshold, the first height of the corresponding historical target person is obtained; Based on the first height, a recommended position of an aisle for the first product is determined.

3. The unmanned retail terminal control method based on the intelligent vending machine according to claim 2, characterized in that: The step of determining a recommended aisle position for the first commodity based on the first height includes: Calculate an average height of the first heights, and determine that a second height at a preset distance below the average height corresponds to a horizontal position on the smart vending machine; Obtaining a vertical distance between the position of each cargo lane and the horizontal position, obtaining a first reference value corresponding to the vertical distance, and associating it with the corresponding cargo lane; Obtain the horizontal distance between each cargo lane and the middle cargo lane of the cargo lane row, determine a second reference value corresponding to the horizontal distance, and associate it with the corresponding cargo lane; Accumulate and calculate the first reference value and the second reference value associated with each cargo channel to obtain a reference degree; Traverse the first product in the product recommendation sequence from the beginning to the end, and use the first product being traversed as the second product; When traversing to the second product, the target position of the currently unassigned aisle with the greatest reference degree is obtained and used as the recommended position corresponding to the second product.

4. The unmanned retail terminal control method based on the intelligent vending machine according to claim 1, characterized in that: The step 3: when the replenisher replenishes the goods, based on the aisle suggestion information, the smart vending machine is controlled to display a replenishment reminder message at a preset display position, including: Continuously obtain machine status signals of smart vending machines; Analyze the machine status signal, and when at least one replenishment signal is obtained, obtain the balance information in the intelligent vending machine, wherein the balance information includes: the remaining goods and the number of the remaining goods; Generate replenishment reminder information based on the preset replenishment reminder information template and according to the remaining quantity information and channel recommendation information.

5. The unmanned retail terminal control method based on the intelligent vending machine according to claim 1, characterized in that: Determining the target person based on the video information includes: Based on the frame processing technology, the video information is processed in frames to obtain multiple frames of images; Determine the target person based on multiple frames of images.

6. The unmanned retail terminal control method based on the intelligent vending machine according to claim 1, characterized in that: The recommending corresponding products to the target person includes: Based on the preset feature extraction template, the first character feature of the target person is obtained; Querying a preset character type library to determine a first character type corresponding to the first character feature; Obtain the target customers of the products sold in the smart vending machine and determine the second person type of the target customers; If the first character type is consistent with any second character type, the sales product corresponding to the second character type is used as a recommended product; Generate recommended product information based on a preset recommended product information generation template and based on the recommended products; Based on the preset recommended product information display rules, according to the recommended product information, the corresponding recommended product information is displayed at the display position and the corresponding target personnel are reminded to view it.

7. The unmanned retail terminal control method based on the intelligent vending machine according to claim 6, characterized in that: The obtaining of target customers of goods sold in the smart vending machine includes: Obtain purchase records of multiple products for sale from the big data platform; Parse the purchase records to obtain multiple one-to-one correspondences between sold goods and purchasing users; Obtain the preset rejection condition feature library; Extract purchase features of purchase records; Match the purchase features with the elimination features in the elimination condition feature library, determine the matching purchase records and eliminate them; Split the remaining purchase records to obtain the products sold and the target customers corresponding to the products sold.

8. An unmanned retail terminal control system based on an intelligent vending machine, characterized in that: include: An acquisition module is used to acquire multiple shipment records stored locally in the smart vending machine; A determination module, configured to determine, based on the shipping record, aisle recommendation information for the first commodity, wherein the aisle recommendation information includes: a recommended location and a recommended number of aisles; A control module, used to control the smart vending machine to display a replenishment reminder message at a preset display position based on the aisle suggestion information when the replenisher replenishes the goods; The recommendation module is used to obtain video information within a preset range in front of the smart vending machine in real time, determine the target person based on the video information, and recommend corresponding products to the target person; The determining module determines the channel suggestion information of the first commodity based on the shipping record, including: Parse the shipping records to obtain the shipping quantity of the first products corresponding to each shipping record; Calculate the first occurrence number corresponding to the shipment number, and at the same time, calculate the second occurrence number corresponding to each shipment number; determining an adjustment factor based on the second number of occurrences and the first number of occurrences; Multiply the adjustment coefficient by the corresponding second occurrence number to obtain the recommended value; Accumulate and calculate each recommendation value to obtain the recommendation degree; Sort the first product by recommendation degree from high to low to obtain a product recommendation sequence; Determine a recommended aisle position for the first product based on the product recommendation sequence; Based on the recommendation degree, the recommended number of channels is determined.

Images