Intelligent design method of personalized discount coupon, electronic device and storage medium

Abstract

The application relates to an intelligent design method of personalized discount coupons, an electronic device and a storage medium, the method comprising the following steps: acquiring a video stream captured by a camera, performing a target tracking task on pedestrians and goods in the video stream to obtain a tracking result, and calculating a commodity sales condition feature according to the tracking result, wherein the commodity sales condition feature at least comprises a browsing frequency of the goods, a consultation frequency of the goods, a taking frequency of the goods and a shopping cart adding frequency of the goods; training a regression model by using the commodity sales condition feature to obtain a sales amount prediction model; predicting the sales amount of the goods under a given discount by using the sales amount prediction model; and generating a discount coupon under the condition of maximum profit according to the predicted sales amount. The application can help a merchant to reasonably control marketing costs and improve profits without invading the privacy of users, and solves the problem of high marketing costs, low sales and low profits caused by the previous discount coupon issuing mode based on personal experience.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, and in particular to an intelligent design method for personalized discount coupons, an electronic device, and a storage medium. BACKGROUND

[0002] Issuing discount coupons is a common promotional behavior, and shopping malls attract users to consume through the way of discount coupons. At present, for large shopping malls or supermarkets, the merchants rely heavily on personal experience for discount coupons, and cannot intelligently generate discount coupon amounts for specific or different goods, resulting in increased marketing costs, but cannot improve the sales volume and profits of the goods.

[0003] At present, the way of issuing discount coupons through personal experience in the related art causes the problems of increased marketing costs, low sales volume, and low profits, and no effective solution has been proposed. SUMMARY

[0004] The embodiments of the present application provide an intelligent design method for personalized discount coupons, an electronic device, and a storage medium to at least solve the problem of increased marketing costs, low sales volume, and low profits caused by the way of issuing discount coupons through personal experience in the related art.

[0005] In a first aspect, the embodiments of the present application provide an intelligent design method for personalized discount coupons, which comprises the following steps:

[0006] Obtaining a video stream captured by a camera;

[0007] Performing a target tracking task on pedestrians and goods in the video stream to obtain a tracking result;

[0008] Calculating a goods sales condition feature according to the tracking result, the goods sales condition feature at least including the number of times the goods are browsed, the number of times the goods are consulted, the number of times the goods are taken, and the number of times the goods are added to a shopping cart;

[0009] Training a regression model using the goods sales condition feature to obtain a sales amount prediction model;

[0010] Predicting the sales volume of the goods under a given discount using the sales amount prediction model;

[0011] Generating a discount coupon under the condition of maximum profit according to the predicted sales volume.

[0012] In some embodiments, the goods sales condition feature further includes the number of times the goods are browsed for a duration longer than a preset threshold but not taken, and the number of times the goods are taken but not added to a shopping cart.

[0013] In some embodiments, after obtaining the video stream captured by the camera, the method further comprises:

[0014] identifying a person identity of the pedestrian through an image classification model, and determining whether the person identity is a service staff or a customer;

[0015] when the person identity of the pedestrian is identified as a customer, generating a temporary ID of the customer, performing a pedestrian attribute identification task on the customer, and adding age interval features and gender features of the customer extracted to the commodity sales condition features, and deleting the temporary ID when the customer leaves the shooting area.

[0016] In some embodiments, before the identifying a person identity of the pedestrian through an image classification model, and determining whether the person identity is a service staff or a customer, the method further comprises:

[0017] detecting a rectangular area of a position of the pedestrian appearing in the video stream picture through a target detection model, and outputting a target detection result;

[0018] cutting a pedestrian image according to the target detection result, and performing quality evaluation on the cut pedestrian image through a quality evaluation module;

[0019] if the pedestrian image does not meet a preset quality evaluation standard of the quality evaluation module, not performing identity information identification on the pedestrian.

[0020] In some embodiments, after the calculating the commodity sales condition features according to the tracking result, the method further comprises:

[0021] for a new commodity without commodity sales condition features, finding a most similar commodity with commodity sales condition features through inherent attributes of the new commodity.

[0022] In some embodiments, after the calculating the commodity sales condition features according to the tracking result, the method further comprises:

[0023] calculating a commodity attention score according to the commodity sales condition features;

[0024] screening commodities according to the commodity attention score and sales volume.

[0025] In some embodiments, the regression model is a linear regression model, a random forest model or a gradient boosting decision tree model.

[0026] In some embodiments, the generating a discount coupon under the condition of maximum profit comprises:

[0027] calculating a discount value at which the profit is maximum;

[0028] calling a preset discount coupon template, the discount coupon template including a page and appearance of the discount coupon;

[0029] writing the discount coupon attribute, the discount date and the discount value into corresponding positions of the discount coupon template, generating the discount coupon, and issuing the discount coupon to the customer through the member system.

[0030] In a second aspect, the embodiments of the present application provide an electronic device, comprising a memory and a processor, the memory storing a computer program, and the processor being configured to implement the method for intelligent design of personalized discount coupons according to the first aspect when running the computer program.

[0031] In a third aspect, the embodiments of the present application provide a storage medium, the storage medium storing a computer program, and the computer program being configured to implement the method for intelligent design of personalized discount coupons according to the first aspect when running.

[0032] Compared with the related art, the method for intelligent design of personalized discount coupons provided by the embodiments of the present application obtains a video stream captured by a camera, performs a target tracking task on pedestrians and goods in the video stream to obtain a tracking result; in this way, the video stream data is processed to analyze the behavior of customers (pedestrians), and the sales status of goods is calculated according to the tracking result, a regression model is trained using the sales status of goods to obtain a sales amount prediction model, the personal information of customers is not used, the sales status of goods is analyzed without infringing on the privacy of customers, the sales amount prediction model can accurately predict the sales volume of goods under different discount amounts, and the detection of the sales status of goods is realized in a single network (sales amount prediction model) to reduce the inference time by sharing most of the calculations, the purpose of performing inference and prediction at a video frame rate is achieved, the sales amount prediction model is used to predict the sales volume of goods under a given discount, in this way, the sales volume of a large number of different goods can be quickly predicted, the degree of intelligence is high and the practicability is strong, the discount coupon is generated under the condition of maximum profit according to the predicted sales volume, in this way, the discount coupon is intelligently designed without infringing on the privacy of users, which not only helps the merchant to reasonably control the marketing cost but also realizes the improvement of profit, and solves the problem of increasing marketing cost, low sales volume and low profit caused by the previous way of issuing discount coupons based on personal experience. BRIEF DESCRIPTION OF DRAWINGS

[0033] The accompanying drawings, which are included to provide a further understanding of the application and constitute a part of this application, illustrate certain illustrative embodiments of the application and together with the description serve to explain the application. In the drawings:

[0034] Figure 1 is a first flowchart of a method for intelligent design of personalized discount coupons according to an embodiment of the present application;

[0035] Figure 2Fig. 4 is a second flowchart illustrating a method for intelligently designing personalized discount coupons according to an embodiment of the present application;

[0036] Figure 3 Fig. 5 is a third flowchart illustrating a method for intelligently designing personalized discount coupons according to an embodiment of the present application;

[0037] Figure 4 Fig. 6 is a flowchart illustrating a method for obtaining the number of times a product is browsed according to an embodiment of the present application;

[0038] Figure 5 Fig. 7 is a flowchart illustrating a method for obtaining the number of times a product is consulted or the time a product is consulted according to an embodiment of the present application;

[0039] Figure 6 Fig. 8 is a flowchart illustrating a method for obtaining the number of times a product is taken or the number of times a product is added to a shopping cart according to an embodiment of the present application;

[0040] Figure 7 Fig. 9 is a fourth flowchart illustrating a method for intelligently designing personalized discount coupons according to an embodiment of the present application;

[0041] Figure 8 Fig. 10 is a graph illustrating the discount of a product versus the sales of the product according to an embodiment of the present application;

[0042] Figure 9 Fig. 11 is a flowchart illustrating a method for generating discount coupons under a profit maximization condition according to an embodiment of the present application;

[0043] Figure 10 Fig. 12 is a schematic diagram of the internal structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0044] In order to make the objectives, technical solutions, and superiorities of the present application clearer, the following will describe and explain the present application in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. Based on the embodiments provided by the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of the present application. In addition, it should be understood that although the effort made during the development process can be complex and lengthy, some design, manufacture, or production changes made by those of ordinary skill in the art based on the technical content disclosed in the present application are only routine technical means and should not be understood as the insufficiency of the content disclosed in the present application.

[0045] Reference to an "embodiment" in this application means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily mutually exclusive of one another. It is expressly understood that that the embodiments described in this application are open to be combined with each other in their various permutations and combinations.

[0046] Unless otherwise defined, technical terms and scientific terms used in the present application shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terms "comprise", "comprising", "include", "including", "contain", "containing", "have", "having", "carry", "carrying", "or" and the like are not intended to exclude the possibility of one or more additional elements, steps, or components. The terms "a", "an", and "the" are not intended to be limiting in that they can mean one or more than one. The terms "include", "including", "have", "having", "contain", "containing", "comprise", "comprising", and the like are not limiting and are used synonymously with "comprising" or "including". The term "connected" is not intended to be limited to direct connections, and can include indirect connections. The term "multiple" means two or more. The term "and / or" means that the associated objects can exist together or separately. The terms "first", "second", "third", and the like are not intended to denote a particular order or ranking, but are merely used to distinguish one object from another.

[0047] The present application provides an intelligent design method of personalized discount coupons.

[0048] Figure 1 is a first flowchart of an intelligent design method of personalized discount coupons according to an embodiment of the present application, referring to Figure 1 In an embodiment of the present application, the intelligent design method of personalized discount coupons is applied to a large shopping mall / supermarket with a camera installed in the commodity area. The method comprises the following steps:

[0049] In step S101, a video stream captured by the camera is obtained.

[0050] In step S102, target tracking tasks are performed on pedestrians and commodities in the video stream to obtain tracking results. For example, video stream data is processed to analyze the behavior of customers (pedestrians).

[0051] It should be noted that the target tracking task in this embodiment is a single-step multi-target tracking task, and commodity detection and Re-ID (pedestrian re-identification) feature extraction can be realized simultaneously through a single model. Commodity detection and identity embedding are completed in a single network (sales prediction model) to reduce inference time by sharing most of the calculations, and the purpose of performing inference and prediction at a video frame rate is achieved. In this embodiment, the target tracking task is implemented by using a FairMOT (Fair Multi-Object Tracking) model. Of course, in some other embodiments, other models can also be used to implement this, which is not limited here.

[0052] In step S103, the commodity sales status features are calculated according to the tracking results. The commodity sales status features at least include the number of times of browsing the commodity, the number of times of consulting the commodity, the number of times of taking the commodity, and the number of times of adding the commodity to the shopping cart. Of course, in some other embodiments, the commodity sales status features can also include commodity basic information features or other features, which are not limited here. For example, the basic information features include commodity identification, commodity name, commodity unit price, and commodity discount information.

[0053] As can be easily understood, the more the commodity sales status features, the better the prediction effect of the subsequent sales prediction model, and the higher the prediction accuracy. In this embodiment, the commodity sales status features are calculated according to the tracking results in the last time span. In other words, the commodity sales status features at least include the number of times of browsing the commodity in the last time span, the number of times of consulting the commodity in the last time span, the number of times of taking the commodity in the last time span, and the number of times of adding the commodity to the shopping cart in the last time span. The last time span refers to the time period in which the last commodity price and discount are unchanged. For example, the last promotion period of the commodity: 0:00:00 on January 6 to 0:00:00 on January 9, during which the commodity price is 200 yuan and the discount amount is 180 yuan (9 times the price). For example, the last promotion period of the commodity: 0:00:00 on January 15 to 0:00:00 on January 20, during which the commodity price is 100 yuan and the discount amount is 80 yuan (8 times the price). In addition, the commodity sales status features can also be extracted by using existing video algorithms, which are not described here.

[0054] In step S104, the regression model is trained by using the commodity sales status features to obtain the sales prediction model. In this way, the prediction of the sales is realized.

[0055] Step S105, using the sales prediction model to predict the sales of the commodity under a given discount; in this way, a large number of different commodities can be quickly predicted, and the degree of intelligence is high and the practicability is strong; wherein, according to the user demand, the given discount can be set according to the user demand, and the given discount can be a given amount or a given discount (for example, 7 times, 8 times, 9 times, etc.), which is not limited here;

[0056] Step S106, generating a discount coupon under the condition of maximizing profit according to the predicted sales; in this way, not only can it help the merchant to reasonably control the marketing cost, but also can realize the profit increase, and solve the problem of increasing marketing cost, low sales and low profit caused by the previous discount coupon issued by personal experience. It should be noted that the present application can generate different discount coupons for different commodities, and the discount coupons do not distinguish customers;

[0057] It should be noted that the above steps S101 to S106 in the embodiment are completed without infringing on the user's privacy, wherein, not infringing on the user's privacy means that the method does not use the customer's personal privacy data. It is worth noting that in the scenario of the embodiment of the present application, the above steps can be used to intelligently generate discount coupons for all commodities in the mall.

[0058] Through the above steps S101 to S106, the video stream captured by the camera is obtained, and the target tracking task is performed on the pedestrians and commodities in the video stream to obtain tracking results; in this way, the video stream data is processed to analyze the behavior of the customers (pedestrians), and the commodity sales condition characteristics are calculated according to the tracking results, the regression model is trained using the commodity sales condition characteristics, and the sales prediction model is obtained. Without using the customer's personal information, the customer's privacy is not infringed, the commodity sales condition is analyzed, and the sales prediction model can accurately predict the sales of the commodity under different discount amounts. Moreover, the detection of the commodity sales condition is realized in a single network (sales prediction model) to reduce the inference time by sharing most of the calculations, the purpose of performing inference and prediction at a video frame rate is achieved, the sales prediction model is used to predict the sales of the commodity under a given discount, in this way, a large number of different commodities can be quickly predicted, and the degree of intelligence is high and the practicability is strong, and the discount coupon is generated under the condition of maximizing profit according to the predicted sales. In this way, the discount coupon is intelligently designed without infringing on the user's privacy, which not only helps the merchant to reasonably control the marketing cost, but also can realize the profit increase, and solves the problem of increasing marketing cost, low sales and low profit caused by the previous discount coupon issued by personal experience.

[0059] Since there are still a large number of potential customers in the shopping process, many potential customers are in a state of observation, and once they receive good news, they are easy to convert into new customers. In an embodiment, the commodity sales condition feature also includes the number of times the commodity is browsed for more than a preset threshold but not taken and the number of times the commodity is taken but not added to the shopping cart. In this way, by increasing the features of the behavior of a large number of potential customers to the commodity, the subsequent training of the sales prediction model can be facilitated to better improve the prediction accuracy; wherein the preset threshold is set according to user demand, for example, the preset threshold can be 1 day, 2 days or more, which is not limited here.

[0060] Further, in some embodiments, the commodity sales condition feature is composed by splicing the number of times the commodity is browsed for more than a preset threshold but not taken, the number of times the commodity is taken but not added to the shopping cart, and the number of times the commodity is browsed, the number of times the commodity is consulted, the number of times the commodity is taken, and the number of times the commodity is added to the shopping cart, and in the splicing process, the system does not save or record the original data of the customer, and does not focus on individual data, so as not to infringe on the privacy of the user.

[0061] Figure 2 FIG. 2 is a second flow diagram of an intelligent design method of a personalized discount coupon according to an embodiment of the present application, referring to Figure 2 In some embodiments, after obtaining the video stream captured by the camera, the method further includes:

[0062] In step S201, the identity of the pedestrian is identified by the image classification model, and it is determined whether the pedestrian is a service personnel or a customer; wherein the identity of the pedestrian can be distinguished between the service personnel and the customer by the uniform clothing of the service personnel (for example, brightly colored work clothes, uniform LOGO, etc.), and in addition, the image classification model can use the MoibleNets series model to achieve this, wherein the MoibleNets series model includes MoibleNetV1, MoibleNetV2, MobileNetV3, etc. The MoibleNets series model used in the present embodiment is a lightweight model, which not only has a small number of parameters but also has a small amount of calculation, so its running speed is fast, which is conducive to improving the training speed of the model;

[0063] In step S202, if the pedestrian is identified as a customer, a temporary ID is generated for the customer and Pedestrian Attribute Recognition (PAR) is performed on the customer. The extracted age range and gender characteristics are added to the product sales status characteristics. When the customer leaves the capture area, the temporary ID is deleted. Since the above steps do not save the user's original data and do not focus on individual data, they do not infringe on user privacy. It is easy to understand that the purpose of the human attribute recognition task is to mine the attribute information of pedestrians (for example, age range features and gender features, etc.) from the input image, and the recognition and mining obtains the high-level semantic information of pedestrians; the available training methods include: based on the RAP algorithm of deep learning, by using manually designed low-level features, such as HOG (Histogram of Oriented Gradient, also known as directional gradient histogram features), SIFT (Scale-invariant feature transform, also known as scale-invariant feature transform), and then combining the classification algorithm SVM and conditional random field (CRF), it is convenient to train the human attribute recognition task. Among them, the existing technologies of RAP algorithm, HOG, SIFT, classification algorithm SVM and CRF are known to those skilled in the art, so they are not described one by one.

[0064] Figure 3 This is a third flow chart of an intelligent design method for personalized discount coupons according to an embodiment of the present application, referring to Figure 3 In actual applications, since pedestrians can move freely within the camera's field of view, their movement speed and posture may change at any time. Therefore, it is impossible to guarantee that every frame of the video is clear and usable. To improve the recognition effect of features, in some embodiments, before identifying the pedestrian's identity through an image classification model and determining whether the pedestrian is a service staff or a customer, the method further includes:

[0065] Step S301: Detecting a rectangular area of ​​pedestrians in a video stream using a target detection model, and outputting a target detection result; wherein the target detection model is an existing technology in the art and will not be described in detail here;

[0066] Step S302: segmenting the pedestrian image according to the target detection result, and performing quality evaluation on the segmented pedestrian image using a quality evaluation module; wherein the quality evaluation module is an existing technology in the art and will not be described in detail here;

[0067] Step S303: If the pedestrian image does not meet the preset quality evaluation standard of the quality evaluation module, the identity information of the pedestrian will not be recognized.

[0068] Figure 4is a flowchart of a process of obtaining the number of times a commodity is browsed according to Embodiment One of the present application, referring to Figure 4 In some embodiments, obtaining the number of times a commodity is browsed comprises the following steps:

[0069] Step S401, when it is detected that the temporary ID of the customer is the same, the time of the customer staying in the target commodity area is calculated.

[0070] Step S402, if the staying time is greater than the preset staying time and the activity range of the customer is less than the preset spatial range, then the pedestrian stays in the last time span, and the activity range of the customer is the circumscribed rectangle of the activity track of the customer within the preset staying time, which is recorded as a browsing time; wherein the preset staying time is set according to customer demand, which is not specifically limited here.

[0071] Step S403, if the activity range of the customer is greater than the preset spatial range, then the activity range of the customer is the circumscribed rectangle of the activity track of the customer within the staying time, which is recorded as a browsing time.

[0072] Figure 5 is a flowchart of a process of obtaining the number of times a commodity is consulted or the time a commodity is consulted according to Embodiment One of the present application, referring to Figure 5 In an embodiment, obtaining the number of times a commodity is consulted or the time a commodity is consulted comprises the following steps:

[0073] Step S501, the distance between the customer and the service personnel is calculated, when it is detected that the distance is less than the preset distance, the communication time T1 is started, and when it is detected that the distance is equal to the preset distance, the communication time T2 is ended; wherein the preset distance is set according to user demand, which is not specifically limited here; this embodiment can distinguish whether the customer or the service personnel through step S201, which is not described one by one here.

[0074] Step S502, the time a target commodity is consulted within the last time span is calculated as the end communication time T2 minus the start communication time T1. For example, the start communication time T1 is 2021-10-00-00, and the end communication time T2 is 2021-10-30-00, then the consultation time is 30 minutes.

[0075] Figure 6 is a flowchart of a process of obtaining the number of times a commodity is taken and the number of times a commodity is added to a shopping cart according to Embodiment One of the present application, referring to Figure 6 In an embodiment, obtaining the number of times a commodity is taken and the number of times a commodity is added to a shopping cart comprises the following steps:

[0076] Step S601, detecting the shopping cart position and the shelf position; wherein, the shelf position is relatively fixed, and can be marked manually; in addition, the number of shelves and the fine granularity of the marking can be adjusted according to the needs of the scene.

[0077] Step S602, when the distance between the pedestrian and the shelf position is less than a preset fixed threshold, start detecting the hand position and tracking the hand position in real time; wherein, the preset fixed threshold is set according to user needs, which is not limited here;

[0078] Step S603, if the hand is detected to enter the shelf area, record the number of times of picking up the goods once;

[0079] Step S604, if the hand is detected to enter the shopping cart area, record the number of times of putting the goods into the shopping cart once.

[0080] In order to solve the cold start problem of new goods and improve practicability, in an embodiment, after calculating the sales status features of the goods according to the tracking results, the method further comprises:

[0081] For new goods without sales status features, find the most similar goods with sales status features through the inherent attributes of the new goods. Wherein, the inherent attributes of the new goods are categories, prices and brands, etc. For example, the new goods are "Jiudao Bao", but Jiudao Bao has no sales status features, and the most similar goods with sales status features are found through the inherent attributes (categories, prices and brands, etc.) of "Jiudao Bao", that is, "Wanglaoji"; execute steps S104 to S106 to generate discount coupons for new goods. Since the sales status features of the goods in this embodiment are all calculated according to the tracking results of the goods in the previous time span, in some other embodiments, in order to improve the effect of the model, the sales status features of the goods in multiple time spans can also be used, for example, the time of the previous three goods promotion.

[0082] Figure 7 is a fourth flowchart of an intelligent design method of a personalized discount coupon according to an embodiment of the present application, referring to Figure 7 In order to meet the needs of going out of stock or increasing sales combined with business, in an embodiment, after calculating the sales status features of the goods according to the tracking results, the method further comprises:

[0083] Step S701, calculating the attention score of the goods according to the sales status features of the goods; wherein, the calculation formula of the attention score of the goods is as follows:

[0084]

[0085] Wherein, Score ItemItem represents the attention score of the product, and Item represents the discounted product, x represents the weight of different features, which can be manually set according to actual conditions, i xi represents the sales condition feature of the i-th product, and n is the total number of product sales condition features.

[0086] Step S702, screening the product according to the attention score and the sales volume of the product. The screening product formula is as follows:

[0087]

[0088] wherein Sale Item represents the sales volume of the product (Item), Score Item represents the attention score of the product.

[0089] In addition, considering that the purchase condition and the discount condition in the transaction will affect the sales volume of the product, in some other embodiments, the purchase condition and the discount condition of the user can also be analyzed by some software program or algorithm to realize the quantitative calculation of the discount sensitivity of the user to obtain the discount and sales curve. For different products, the discount and sales curve is different. Figure 8 is a schematic diagram of the product discount and sales curve according to the embodiment one of the present application, referring to Figure 8 It can be seen that the sales volume rapidly increases with the increase of the discount at the initial stage, but when the discount decreases to a certain extent, the change tends to stagnate, and the sales volume stops increasing due to the shortage of supply or the saturation of purchasing power. In the actual scene, the discount cannot reach 100%. Selecting the product sensitive to the change of the discount can bring more significant effect. Note that the discount and sales curve at this time only represents the general trend, and therefore cannot be used for accurate prediction of the sales volume, but only for analysis of the sensitivity of the product to the discount.

[0090] In some embodiments, the regression model is a linear regression model, a random forest model or a gradient boosting decision tree model. Of course, in some other embodiments, the regression model can also select other models with better regression effect according to the user demand, which is not limited here.

[0091] Figure 9 is a process schematic diagram of generating a discount coupon under the condition of profit maximization according to the embodiment of the present application, referring to Figure 9 In some embodiments, generating a discount coupon under the condition of profit maximization includes:

[0092] Step S901, calculating the discount value when the profit is maximum; the skilled person in the art can calculate the discount value when the profit is maximum through the formula of total sales and the formula of total profit; wherein the formula of total sales is: total sales = sales quantity * sales price = sales quantity * (original price - discount amount), in addition, the formula of total profit is: total profit = sales quantity * single piece profit = sales quantity * (original price - discount amount - cost), since the sales quantity of the commodity under different discounts can be predicted through the model, and according to the above formula of total sales or the formula of total profit, the discount value corresponding to the maximum calculation value can be calculated;

[0093] Step S902, calling a preset discount coupon template, the discount coupon template including the page and appearance of the discount coupon; wherein the page and appearance can be set according to the user's demand, which is not limited here;

[0094] Step S903, writing the discount coupon attribute, discount date and discount value calculated in step S901 into the corresponding position of the discount coupon template to generate the discount coupon, and issuing the discount coupon to the customer through the member system. Wherein the discount coupon attribute, discount date and discount value can be set according to the user's demand, which is not limited here;

[0095] Wherein the member system is an existing technology, and some basic information of the customer is recorded in the member system, and the use of member system data means that the use of these information has been authorized by the user; as shown in the following table content:

[0096] Name Source Customer Unique ID Product Membership System Gender Product Membership System Age Product Membership System User's Purchased Products Historical Transaction Data User's Transaction History Historical Transaction Data

[0097] Wherein the unique identifier refers to the ID used to identify the user's identity in the membership system, the gender refers to the gender of the customer, the age refers to the age of the customer, and the user's purchase of goods refers to the actual purchase of goods identification, type and quantity, etc.; the user's transaction refers to the payment method, payment amount, discount method and discount amount, etc.

[0098] In order to maximize the conversion rate of the discount coupon, of course, in some other embodiments, combined with expert experience and industry knowledge, the skilled person in the art can also analyze the use of the discount coupon through some software algorithms, programs or models, and obtain the relationship between the use rate of the discount coupon and the use time, the validity period of the discount coupon, and the type of the commodity, and write the setting method of the validity period of the discount coupon as a rule (i.e. if A commodity is used in B situation with C discount, the validity period is D) to realize the selection of suitable discount coupon time attribute for specific commodity, increase the conversion rate of discount coupon, and improve the profit of commodity.

[0099] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0100] This embodiment further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0101] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0102] Optionally, in this embodiment, the processor may be configured to execute the following steps through a computer program:

[0103] Step S101, obtaining a video stream captured by a camera;

[0104] Step S102, performing a target tracking task on pedestrians and goods in the video stream to obtain a tracking result;

[0105] Step S103, calculating product sales characteristics based on the tracking results, the product sales characteristics at least including the number of times the product was viewed, the number of times the product was consulted, the number of times the product was taken, and the number of times the product was added to the shopping cart;

[0106] Step S104, using the product sales characteristics to train a regression model to obtain a sales forecast model;

[0107] Step S105, using the sales forecasting model to forecast the sales volume of the product under a given discount;

[0108] Step S106: Generate discount coupons based on the predicted sales volume under the profit maximization condition.

[0109] In addition, in conjunction with the intelligent design method for personalized discount coupons in the above embodiments, embodiments of the present application may provide a storage medium for implementation. The storage medium stores a computer program; when the computer program is executed by a processor, it implements any of the intelligent design methods for personalized discount coupons in the above embodiments.

[0110] In one embodiment, a computer device is provided, which can be a terminal. The computer device comprises a processor, a memory, a network interface, a display screen and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is configured to communicate with an external terminal through a network connection. The computer program, when executed by the processor, implements an intelligent design method of personalized discount coupons. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0111] In one embodiment, Figure 10 is a schematic diagram of the internal structure of an electronic device according to an embodiment of the present application, as Figure 10 shown, an electronic device is provided, which can be a server, and a schematic diagram of the internal structure of the electronic device can be as Figure 10 shown. The electronic device comprises a processor, a network interface, an internal memory and a non-volatile memory connected through an internal bus, wherein the non-volatile memory stores an operating system, a computer program and a database. The processor is configured to provide computing and control capabilities, the network interface is configured to communicate with an external terminal through a network connection, the internal memory is configured to provide an environment for running the operating system and the computer program, the computer program, when executed by the processor, implements an intelligent design method of personalized discount coupons, and the database is configured to store data.

[0112] Those skilled in the art can understand that Figure 10 the structure shown in the above is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device can comprise more or fewer components than those shown in the diagram, or combine certain components, or have a different arrangement of components.

[0113] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiment methods. Any reference to memory, storage, database or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0114] Those skilled in the art should understand that each technical feature of the above-mentioned embodiments can be combined arbitrarily, and in order to make the description simple, not all possible combinations of each technical feature in the above-mentioned embodiments are described, however, as long as the combination of these technical features does not exist contradictory, it should be considered as the scope of the present application.

[0115] The above-mentioned embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of protection of the patent of the present application should be subject to the appended claims.

Claims

1. A method for intelligent design of personalized discount coupons, characterized in that, The method is applied to a large-scale shopping mall / supermarket with a camera installed in a commodity area, and comprises the following steps: acquiring a video stream captured by the camera; performing a single-step multi-target tracking task on pedestrians and commodities in the video stream to obtain tracking results, wherein the tracking results correspond to behavior analysis results of the pedestrians; identifying the identities of the pedestrians by an image classification model and determining whether the pedestrians are service personnel or customers; when a customer is identified, generating a temporary ID of the customer, performing a pedestrian attribute identification task on the customer, and adding age interval features and gender features of the customer to commodity sales status features, and deleting the temporary ID when the customer leaves the shooting area; calculating commodity sales status features from the tracking results, wherein the commodity sales status features at least include the number of times a commodity is browsed, the number of times a commodity is consulted, the number of times a commodity is taken, the number of times a commodity is added to a shopping cart, the number of times a commodity is browsed for more than a preset threshold but not taken, and the number of times a commodity is taken but not added to a shopping cart, and the age interval features and gender features of customers among pedestrians; wherein for a new commodity without commodity sales status features, the most similar commodity with commodity sales status features is found by inherent attributes of the new commodity, such as category, price, and brand; training a regression model using the commodity sales status features to obtain a sales prediction model; predicting the sales volume of a commodity at a given discount using the sales prediction model; generating a discount coupon under the condition of maximizing profit according to the predicted sales volume.

2. The method of claim 1, wherein, Before the step of identifying the identities of the pedestrians by the image classification model and determining whether the pedestrians are service personnel or customers, the method further comprises: detecting the rectangular area of the position of the pedestrian appearing in the video stream frame by a target detection model, and outputting a target detection result; cutting the pedestrian image according to the target detection result, and performing quality evaluation on the cut pedestrian image by a quality evaluation module; if the pedestrian image does not meet the preset quality evaluation standard of the quality evaluation module, the identity information of the pedestrian is not identified.

3. The method of claim 1, wherein, After the step of calculating commodity sales status features from the tracking results, the method further comprises: calculating a commodity attention score according to the commodity sales status features; screening commodities according to the commodity attention score and sales volume.

4. The method according to any one of claims 1 to 3, characterized in that, The regression model is a linear regression model, a random forest model, or a gradient boosting decision tree model.

5. The method of claim 1, wherein, The step of generating a discount coupon under the condition of maximizing profit comprises: calculating the discount value when the profit is maximized; calling a preset discount coupon template, which includes the page and appearance of the discount coupon; writing the discount coupon attributes, discount date, and discount value into the corresponding positions of the discount coupon template to generate the discount coupon, and issuing the discount coupon to the customer through a member system. 6.An electronic device comprising a memory and a processor, the electronic device comprising: The memory stores a computer program, and the processor is configured to run the computer program to execute the intelligent design method of the personalized discount coupon according to any one of claims 1 to 5.

7. A storage medium, characterized by The storage medium has stored therein a computer program, wherein the computer program is configured to execute the intelligent design method of the personalized discount coupon according to any one of claims 1 to 5 when running.

Images