WO2022142899A1

WO2022142899A1 - Store sales data analysis method, apparatus, electronic device, and storage medium

Info

Publication number: WO2022142899A1
Application number: PCT/CN2021/133107
Authority: WO
Inventors: 黄焯真
Original assignee: 深圳云天励飞技术股份有限公司
Priority date: 2020-12-31
Filing date: 2021-11-25
Publication date: 2022-07-07
Also published as: CN112883775A

Abstract

Provided are a store sales data analysis method, apparatus, electronic device, and storage medium, said method comprising: obtaining a first designated area image and performing face detection on said first designated area image to obtain first face statistical data (101); obtaining a second designated area image and performing face detection on said second designated area image to obtain second face statistical data (102); obtaining a third designated area image and performing face detection on said third designated area image to obtain third face statistical data (103); obtaining a fourth designated area image and performing face detection on said fourth designated area image to obtain fourth face statistical data (104); said fourth designated area being located in the checkout area of said store; according to said first face statistical data, second face statistical data, third face statistical data, and fourth face statistical data, performing sales analysis of the store. The described method can help a store to reduce the trial-and-error costs during the stocking process.

Description

Store sales data analysis method, device, electronic device and storage medium

This application claims the priority of the Chinese patent application filed on December 31, 2020 with the application number 202011640061.6 and the invention name is "store sales data analysis method, device, electronic device and storage medium", the entire content of which is approved by Reference is incorporated in this application.

technical field

The invention relates to the field of artificial intelligence processing, in particular to a method, device, electronic device and storage medium for analyzing store sales data.

Background technique

With the rapid development of information technology, artificial intelligence applications have gradually been integrated into the lives of the public. In particular, the application of face recognition is not only widely used in the field of payment, but also widely used in other fields, such as smart shops, smart supermarkets, etc. With the rapid development of the current mobile Internet, online shopping has become the mainstream of the current world. Offline brick-and-mortar stores are struggling and facing severe survival challenges. It seems very cautious, and they dare not make decisions easily. Moreover, in the start-up stage of the store, in the absence of a stable customer source and word of mouth, due to the diversity of new customers and the subjectivity of customers and store employees, it is difficult Most offline store employees are struggling in terms of traffic drainage and promotion. There are few sales data for reference, and it is difficult for purchases to meet customers' purchase expectations. The trial and error costs of multiple purchases are high, resulting in the store after a period of operation. Announced to close. Therefore, in the start-up stage of existing stores, due to imperfect data in various aspects, few sales data for reference, and high trial and error costs, the sales effect of the stores is poor.

SUMMARY OF THE INVENTION

The embodiment of the present invention provides a method for analyzing store sales data, which can analyze the implicit relationship among customers, employees, and commodities that affect store sales through face statistical data in various areas of the store, so as to improve the performance of the store in each area. Dimensional sales data can help stores locate customers' purchasing expectations, thereby reducing trial and error costs during the purchase process.

In a first aspect, an embodiment of the present invention provides a method for analyzing store sales data, the method comprising:

Obtain the first designated area image, perform face detection on the first designated area image, and obtain first face statistical data, the first designated area is located in the area outside the store, and the first face The statistical data includes face statistics of pedestrians passing through the shop;

Obtain the second designated area image, perform face detection on the second designated area image, and obtain second face statistical data, the second designated area is located in the door area of the store, and the second face statistical data The data includes facial statistics of customers entering the store;

Obtain a third designated area image, perform face detection on the third designated area image, and obtain third face statistical data, the third designated area is located in the shelf area of the store, and the third face statistical data Data includes browsing product data, employee and browsing customer facial statistics;

Obtain a fourth designated area image, perform face detection on the fourth designated area image, and obtain fourth face statistical data, the fourth designated area is located in the cashier area of the store, and the fourth face statistical data The data includes the facial statistics of the checkout customer and the checkout item data;

The sales analysis of the store is performed according to the first face statistics data, the second face statistics data, the third face statistics data, and the fourth face statistics data.

Optionally, the sales analysis includes sales data analysis and sales forecast analysis. The store conducts sales analysis, including:

Form the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data into a graphic report, and analyze the sales data according to the graphic report;

The correlation calculation is performed according to at least two of the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data, and the correlation calculation is carried out according to the result of the correlation calculation. Sales forecast analysis.

Optionally, the correlation calculation is performed according to at least two of the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data, and the correlation is calculated according to the correlation. The results of the calculation are carried out for the sales forecast analysis, including:

According to the standard deviation corresponding to any two items in the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data and the covariance between the any two items, Calculate the correlation between the any two items, and carry out the correlation according to any two items in the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data. The sales forecast analysis.

According to the standard deviation of the first face statistical data, the standard deviation of the fourth face statistical data, and the covariance between the first face statistical data and the fourth face statistical data, calculate the the first correlation between the first face statistics and the fourth face statistics; and/or

According to the standard deviation of the second human face statistical data, the standard deviation of the fourth human face statistical data, and the covariance between the second human face statistical data and the fourth human face statistical data, calculate the a second correlation between the second face statistics and the fourth face statistics; and/or

According to the standard deviation of the third human face statistical data, the standard deviation of the fourth human face statistical data, and the covariance between the third human face statistical data and the fourth human face statistical data, calculate the The third correlation of the fourth face statistical data of the third face statistical data is performed, and the sales forecast analysis is performed according to the first correlation and/or the second correlation and/or the third correlation.

Optionally, according to the standard deviation of the first human face statistical data, the standard deviation of the fourth human face statistical data, the difference between the first human face statistical data and the fourth human face statistical data Calculate the first correlation between the first face statistics and the fourth face statistics, including:

According to the statistical data of the faces of the pedestrians passing through the shop, the traffic flow of the pedestrians in the shop is obtained according to statistics;

calculating the standard deviation of the pedestrian flow of said pedestrian;

According to the checkout product data, the number of checkout products is obtained by statistics;

calculating the standard deviation of the checkout item quantity;

calculating the covariance between the flow of the pedestrian and the quantity of the checkout item;

According to the standard deviation of the pedestrian flow, the standard deviation of the checkout item quantity, and the covariance between the pedestrian flow and the checkout item quantity, calculate the difference between the pedestrian flow rate and the checkout item quantity correlation as the first correlation.

Optionally, according to the standard deviation of the second human face statistical data, the standard deviation of the fourth human face statistical data, the difference between the second human face statistical data and the fourth human face statistical data The covariance of , calculating the second correlation between the second face statistics and the fourth face statistics, including:

According to the face statistics of the customers entering the store, the number of customers entering the store is obtained by statistics;

calculating the standard deviation of the number of customers entering the store;

Calculate the number of in-store customers and the check-out commodity according to the standard deviation of the number of in-store customers, the standard deviation of the check-out commodity number, and the covariance between the in-store customer number and the check-out commodity number The correlation of the quantity serves as the second correlation.

Optionally, according to the standard deviation of the third human face statistical data, the standard deviation of the fourth human face statistical data, the difference between the third human face statistical data and the fourth human face statistical data , calculate the third correlation of the fourth face statistical data of the third face statistical data, including:

According to the browsing customer's face statistics data and browsing commodity data, the number of times or time of commodity wandering is obtained according to statistics;

Calculate the standard deviation corresponding to the number of times or time the commodity lingers;

According to the standard deviation of the number of times of commodity wandering, the standard deviation of the number of checkout commodities, and the covariance between the number of times of commodity wandering and the number of checkout commodities, calculate the difference between the number of times of commodity wandering and the number of checkout commodities, respectively correlation as a third correlation; or

According to the standard deviation of the commodity lingering time, the standard deviation of the checkout commodity quantity, and the covariance between the commodity lingering time and the checkout commodity quantity, the difference between the commodity lingering time and the checkout commodity quantity is calculated respectively. correlation as a third correlation.

In a second aspect, an embodiment of the present invention further provides a store sales data analysis device, the device comprising:

The first acquisition module is used to acquire a first designated area image, and perform face detection on the first designated area image to obtain first face statistical data, and the first designated area is located in the area outside the store. , the first face statistical data includes the face statistical data of pedestrians passing through the store;

The second acquisition module is configured to acquire an image of a second designated area, perform face detection on the image of the second designated area, and obtain second face statistical data, and the second designated area is located in the door area of the store, The second face statistical data includes face statistical data of customers entering the store;

a third acquisition module, configured to acquire an image of a third designated area, perform face detection on the image of the third designated area, and obtain third face statistical data, and the third designated area is located in the shelf area of the store, The third face statistical data includes browsing product data, employee and browsing customer face statistical data;

a fourth acquisition module, configured to acquire a fourth designated area image, perform face detection on the fourth designated area image, and obtain fourth face statistical data, where the fourth designated area is located in the cashier area of the store, The fourth face statistical data includes face statistical data of checkout customers and checkout commodity data;

The analysis module is used to analyze the sales of the shop according to the first face statistics data, the second face statistics data, the third face statistics data and the fourth face statistics data.

In a third aspect, an embodiment of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, when the processor executes the computer program The steps in the store sales data analysis method provided by the embodiment of the present invention are implemented.

In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the method for analyzing store sales data provided by the embodiment of the present invention is implemented steps in .

In this embodiment of the present invention, an image of a first designated area is acquired, and face detection is performed on the image of the first designated area to obtain first statistical data of faces, and the first designated area is located in an area outside the shop, The first face statistical data includes face statistical data of pedestrians passing through the store; a second designated area image is obtained, and face detection is performed on the second designated area image to obtain second face statistical data, The second designated area is located at the entrance area of the store, and the second face statistical data includes the face statistical data of customers entering the store; the image of the third designated area is acquired, and the image of the third designated area is processed. face detection to obtain third face statistical data, the third designated area is located in the shelf area of the store, and the third face statistical data includes browsing product data, employee and browsing customer face statistical data; Four designated area images, and face detection is performed on the fourth designated area image to obtain fourth face statistical data, the fourth designated area is located in the cashier area of the store, and the fourth face statistical data includes The facial statistics data of the checkout customer and the checkout commodity data; the sales analysis of the store is performed according to the first facial statistics data, the second facial statistics data, the third facial statistics data, and the fourth facial statistics data. The present invention can analyze the implicit relationship among customers, employees and commodities that affect the sales of the store through the acquired face statistics data in various areas of the store, thereby improving the sales data of the store in various dimensions and helping the store to locate. Customers' purchase expectations, thereby reducing the trial and error cost in the purchase process, and ensuring the continuous operation of the store.

Description of drawings

1 is a flowchart of a method for analyzing store sales data provided by an embodiment of the present invention;

2 is a schematic structural diagram of a store sales data analysis device provided by an embodiment of the present invention;

3 is a schematic structural diagram of an analysis module provided by an embodiment of the present invention;

4 is a schematic structural diagram of a second processing submodule provided by an embodiment of the present invention;

5 is a schematic structural diagram of a first computing unit provided by an embodiment of the present invention;

6 is a schematic structural diagram of a second computing unit provided by an embodiment of the present invention;

7 is a schematic structural diagram of a third computing unit provided by an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention.

Detailed ways

Please refer to FIG. 1. FIG. 1 is a flowchart of a method for analyzing store sales data provided by an embodiment of the present invention. As shown in FIG. 1, the method includes the following steps:

101. Acquire a first designated area image, and perform face detection on the first designated area image to obtain first face statistical data.

In the embodiment of the present invention, the above-mentioned first designated area is located in an area outside the shop door, and the first face statistics data include face statistics data of pedestrians passing by the shop.

Further, the image of the first designated area may be collected by a first camera set at the door of the shop and shooting toward the outside of the shop. When a pedestrian passes by in the first designated area outside the store (which may be an area in front of the store), the first camera will capture images regardless of whether the pedestrian enters the store or not.

The above-mentioned face detection for the first designated area may be performed by a first face detection engine, and the image of the first designated area is a large image, and the large image includes at least one pedestrian. Through the first face detection engine, you can Detect how many pedestrians are included in the large image, and output the pedestrian's face map, face feature values, etc. Further, the above-mentioned first face statistical data may include camera ID, camera position, pedestrian's face map, pedestrian's face feature value, capture time, number of shots, etc., and the above-mentioned pedestrian's face feature value may include the age of the pedestrian. , gender, etc.

In a possible embodiment, the above-mentioned first face detection engine may be integrated in the first camera, and perform face detection on the captured image of the first designated area in real time.

102. Acquire an image of a second designated area, and perform face detection on the image of the second designated area to obtain second face statistical data.

In the embodiment of the present invention, the above-mentioned second designated area is located in the above-mentioned store entrance area, and the above-mentioned second face statistical data may include face statistical data of customers entering the store.

Further, the above-mentioned second designated area image may be collected by a second camera disposed in the store and photographing at the entrance of the store. When there is a customer entering the store in the second designated area of the store entrance area (which may be an area directly below the store entrance), the second camera will capture images regardless of whether the entering customer will buy a product.

The above-mentioned face detection on the image of the second designated area may be performed by a second face detection engine, and the image of the second designated area is a large image, and the large image includes at least one customer entering the store. The engine can detect how many customers enter the store in the big picture, and output the face map and face feature values of the customers corresponding to the store. Further, the above-mentioned second face statistical data may include the camera ID, the position of the camera, the face map of the customer entering the store, the facial feature value of the customer entering the store, the capture time, etc. The above-mentioned facial feature value of the customer entering the store may include: Age, gender, glasses, hats, accessories, etc. of customers entering the store. In addition, it is also possible to perform human body detection on the image of the second designated area to obtain the wearing feature values of the customers entering the store, such as the feature values corresponding to tops and bottoms, and bind the wearing feature values to the second face statistical data.

In a possible embodiment, the above-mentioned second face detection engine may be integrated in the second camera to perform face detection on the captured image of the second designated area in real time.

103. Acquire a third designated area image, and perform face detection on the third designated area image to obtain third face statistical data.

The above-mentioned third designated area is located in the shelf area of the store, and the above-mentioned third face statistical data may include browsing product data, employee and browsing customer face statistical data.

Further, the image of the above-mentioned third designated area may be collected by a third camera that is arranged in the store and shoots at the shelf. When there is a browsing customer in the third designated area of the shelf area (which may be the area directly in front of the shelf), the third camera will capture images regardless of whether the browsing customer will purchase a product.

The above-mentioned face detection on the image of the third designated area may be carried out by a third face detection engine, and the image of the third designated area is a large image, and the large image includes at least one browsing customer. , you can detect how many browsing customers and whether there are employees in the big picture, and output the products browsed by the browsing customers, the face maps of employees and browsing customers, and the facial feature values, etc. correspondingly. Further, the above-mentioned third face statistical data may include camera ID, camera position, browsing customer's face map, browsing customer's facial feature value, capture time, browsing products, employee's face map, etc. The facial feature values may include the browsing customer's age, gender, glasses, hats, accessories, and the like. In addition, it is also possible to perform human body detection on the image of the third designated area to obtain the wearing characteristic values of the browsing customers, such as the characteristic values corresponding to tops and bottoms, and bind the wearing characteristic values to the third face statistical data.

In a possible embodiment, the above-mentioned third face detection engine may be integrated in a third camera to perform face detection on the captured image of the third designated area in real time.

104. Acquire an image of the fourth designated area, and perform face detection on the image of the fourth designated area to obtain fourth face statistical data.

In the embodiment of the present invention, the fourth designated area is located in the cash register area of the store, and the fourth face statistical data may include face statistical data of checkout customers and checkout commodity data.

Further, the image of the above-mentioned fourth designated area may be collected by a fourth camera that is set in the shop and shoots in front of the cashier. When there is a checkout customer in the fourth designated area of the cash register area (which may be the area directly in front of the cash register), the fourth camera will perform image capture.

The above-mentioned face detection on the image of the fourth designated area may be performed by a fourth face detection engine, and the image of the fourth designated area is a large image, and the large image includes at least one checkout customer. , you can detect how many checkout customers and corresponding checkout products are included in the big picture, and output the checkout products that the checkout customer needs to checkout, the face map of the checkout customer, and the facial feature values, etc. Further, the above-mentioned fourth face statistical data may include the camera ID, the position of the camera, the face map of the checkout customer, the face feature value of the checkout customer, the capture time, the checkout product, etc. The face feature value of the above-mentioned checkout customer may include: Checkout customer's age, gender, glasses, hats, accessories, etc. In addition, human body detection can also be performed on the image of the fourth designated area to obtain the characteristic values of the checkout customer's wearing, such as the characteristic values corresponding to tops and bottoms, and the wearing characteristic values are bound to the fourth face statistical data.

In a possible embodiment, the above-mentioned fourth face detection engine may be integrated in the third camera to perform face detection on the captured image of the fourth designated area in real time.

105. Perform a sales analysis on the store according to the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data.

In the embodiment of the present invention, the sales analysis of the store may be performed according to a single item of the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data, for example, by The traffic flow of pedestrians in the first face statistics data can be used to analyze the sales of shops. For example, if there are a lot of people after get off work at noon and in the afternoon, you can sell food and beverage products. The sales analysis of the store is carried out through the age and gender of the customers entering the store in the second face statistics data. Further, for example, if most of the customers entering the store are young women, jewelry-type products can be sold. The sales analysis of the store is carried out through the browsing customer and browsing product data in the third-party face statistical data. Further, for example, if the browsing time of the browsing customer or the number of browsing times exceeds a certain number of times, the similar products of the browsing product can be sold. . Through the checkout customers and checkout products in the fourth face statistical data, we can perform a shadowless analysis on the store. For example, if the checkout customer buys a shirt and pays the bill, the checkout product is a shirt, and the shirt can continue to be sold.

Optionally, the above-mentioned sales analysis includes sales data analysis and sales forecast analysis, and the above-mentioned first face statistical data, second face statistical data, third face statistical data, and fourth face statistical data can be formed into graphic reports. , and carry out the above sales data analysis according to the above-mentioned graphic report; carry out correlation calculation according to at least two of the above-mentioned first face statistical data, second face statistical data, third face statistical data, and fourth face statistical data , and conduct sales forecast analysis based on the results of the above correlation calculation.

Further, the above sales data analysis can be called static analysis, and the sales forecast analysis can be called dynamic analysis.

For the above-mentioned first face statistical data, the above-mentioned graphic report may be pedestrian traffic statistics, specifically, time-sharing traffic and facial feature value statistics. For example, from 8:00 to 9:00 on September 11, 2020, The total flow of people is 235, the flow of people over 20 is 200, the flow of men is 111, the flow of women is 124, etc.

For the above-mentioned second face statistical data, the above-mentioned graphic report can be the traffic flow of customers entering the store, and specifically can be time-sharing traffic flow and facial feature value statistics, for example, from 8:00 to 9:00 on September 11, 2020 , the total number of people entering the store is 21, the number of people over the age of 20 is 18, the number of people who enter the store is 10, the number of people who wear a hat is 11, the number of people wearing hats is 3, the number of people wearing glasses is 4, and the number of people wearing jeans The traffic flow is 21, the traffic flow of people wearing black tops is 13, etc.

For the above-mentioned third face statistical data, the above-mentioned graphic and text report can be the wandering time or the number of wandering times of customers browsing in front of the shelves, and specifically can be the time-sharing face statistics on each shelf, the number of people wandering on each shelf, and the total wandering time statistics on each shelf. , Statistics on the number of shelf wanderings, and statistics on employee turnover. For example, from 8:00 to 9:00 on September 11, 2020, there were 5 people wandering on shelf A, and the total wandering time on shelf B was 4 hours (the wandering time in the case of multiple people is superimposed ), the number of wanderings on the C shelf is 24 times, and the staff on the D shelf has visited once, etc.

For the above-mentioned fourth face statistical data, the above-mentioned graphic and text report can be the statistics of the number of customers who check out before the checkout and the type of goods to be checked out. For example, from 8:00 to 9:00 on September 11, 2020, 4 customers purchased goods And checkout, the goods purchased by checkout customer A are a, b, and the goods purchased by checkout customer B are a, c, and so on.

The above-mentioned graphic and text report can also be a time-sharing graph, a time-sharing trend graph, etc. of the first face statistical data and the second face statistical data, or can also be the distribution map of each shelf hot spot of the third face statistical data, The histogram, the proportion of employees and browsing customers in the time period, etc., can also be the distribution and placement map of the proportion of checkout customers in the third face statistical data in the fourth face statistical data, and the placement map of checkout products, etc. .

The above-mentioned sales forecast analysis may be correlation analysis based on historical first face statistical data, second face statistical data, third face statistical data, and fourth face statistical data. For example, through the historical first face statistics and second face statistics, it is possible to analyze whether the flow of pedestrians is positively correlated with the flow of customers entering the store. Through the historical first face statistical data and the fourth face statistical data, it can be analyzed whether there is a positive correlation between the flow of pedestrians and the number of checkout customers (or the number of checkout items). Through the historical second face statistics and fourth face statistics, it can be analyzed whether there is a positive correlation between the flow of people entering the store and the number of checkout customers (or the number of checkout items). Through the third face statistical data and the fourth face statistical data, it can be analyzed whether the wandering time or the number of wandering times of browsing customers is positively correlated with the number of checkout customers (or the number of checkout items), and so on.

Optionally, the standard deviation corresponding to any two items in the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data and the covariance between any two items may be used. , calculate the correlation between any two items, and carry out the sales according to the correlation of any two items in the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data Predictive analytics.

Further, the correlation between any two items can be calculated as follows:

Among them, the above ρ _xy is the correlation between the two variables of X and Y, the above Cov(X, Y) is the covariance between the two variables of X and Y, the above

is the standard deviation of the X variable, the above

is the standard deviation of the Y variable. In the above formula, X can be any one of the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data, and Y can be any of the remaining three items. one. For example, if the first face statistical data is X, then Y may be any one of the second face statistical data, the third face statistical data, and the fourth face statistical data. The above X is the mean value of the X variable, and Y is the mean value of the Y variable. When the above correlation is 0, the two variables X and Y have no relationship. When the value of X increases (decreases) and the value of Y increases (decreases), the two variables are positively correlated, and the correlation is between 0.00 and 1.00. When the value of X increases (decreases) and the value of Y decreases (increases), the two variables are negatively correlated, and the correlation is between -1.00 and 0.00.

Optionally, the first face can be calculated according to the standard deviation of the first face statistical data, the standard deviation of the fourth face statistical data, and the covariance between the first face statistical data and the fourth face statistical data. the first correlation of the statistical data with the fourth human face statistical data; and/or according to the standard deviation of the second human face statistical data, the standard deviation of the fourth human face statistical data, the second human face statistical data and the fourth human face covariance between statistical data, calculating the second correlation between the second statistical data and the fourth statistical data; and/or according to the standard deviation of the third statistical data, the fourth statistical data The standard deviation, the covariance between the third face statistical data and the fourth face statistical data, calculate the third correlation of the third face statistical data and the fourth face statistical data, and based on the first correlation and/ Or second correlation and/or third correlation for sales forecast analysis.

Further, the first face statistics are calculated according to the standard deviation of the first face statistics, the standard deviation of the fourth face statistics, and the covariance between the first face statistics and the fourth face statistics. The first correlation between the data and the fourth face statistical data may specifically be that, according to the face statistical data of the pedestrians passing through the shop, the pedestrian flow of the shop is obtained by statistics; the standard deviation of the pedestrian flow is calculated; according to the checkout commodity data, Calculate the number of checkout items; calculate the standard deviation of the checkout items; calculate the covariance between the pedestrian flow and the checkout item; The covariance between the quantities is calculated, and the correlation between the flow of people and the number of checkout items is calculated as the first correlation. The above-mentioned first correlation can determine whether the number of goods sold by the store is proportional to the flow of pedestrians, and it can also be understood as whether the more pedestrians, the better the goods are sold.

Further, the second face statistics are calculated according to the standard deviation of the second face statistics, the standard deviation of the fourth face statistics, and the covariance between the second face statistics and the fourth face statistics. The second correlation between the data and the fourth face statistical data, specifically, the number of customers entering the store can be obtained according to the statistical data of the faces of the customers entering the store; ); according to the standard deviation of the number of customers entering the store, the standard deviation of the number of checkout items, and the covariance between the number of customers entering the store and the number of items checked out, the correlation between the number of customers entering the store and the number of items checked out is calculated as the second Correlation. The above-mentioned second correlation can determine whether the number of goods sold by the store is proportional to the number of customers entering the store, and it can also be understood whether the more people who enter the store, the better the goods are sold.

Optionally, calculate the third face statistics according to the standard deviation of the third face statistical data, the standard deviation of the fourth face statistical data, and the covariance between the third face statistical data and the fourth face statistical data The third correlation of the data fourth face statistics, you can count the number or time of product wandering according to the browsing customer’s face statistics and browsing product data; calculate the standard deviation corresponding to the number of product wandering or time; The standard deviation of , the standard deviation of the number of items at checkout, and the covariance between the number of items lingering and the number of items at checkout are calculated, and the correlation between the number of items lingering and the number of items at checkout is calculated as the third correlation; or according to the standard of lingering time of items The difference, the standard deviation of the checkout item quantity, and the covariance between the item lingering time and the checkout item quantity, respectively calculate the correlation between the item lingering time and the checkout item quantity as the third correlation. The above-mentioned third correlation can determine whether the number of products sold by the store is proportional to the wandering time or the number of wandering times of browsing customers, and it can also be understood whether the more and longer the viewing, the easier it is to generate purchase behavior.

Through the above method, it is also possible to predict whether the successfully sold product is proportional to the employee's guidance, and then determine the direction of employee training through the number of successful product sales and the number of employees' wandering, and then determine the direction of employee training. The correlation between the product portfolios can be found, and the shelves can be optimized.

In the embodiment of the present invention, an image of a first designated area is acquired, and face detection is performed on the image of the first designated area to obtain first statistical data of faces, and the first designated area is located in the area outside the store. , the first face statistical data includes the face statistical data of pedestrians passing through the store; obtain a second designated area image, and perform face detection on the second designated area image to obtain second face statistical data , the second designated area is located in the door area of the store, and the second face statistical data includes the face statistical data of customers entering the store; the third designated area image is acquired, and the third designated area image is processed. face detection to obtain third face statistical data, the third designated area is located in the shelf area of the store, and the third face statistical data includes browsing product data, employee and browsing customer face statistical data; obtaining a fourth designated area image, and performing face detection on the fourth designated area image to obtain fourth face statistical data, the fourth designated area is located in the cashier area of the store, and the fourth face statistical data Including the face statistics data of checkout customers and checkout commodity data; according to the first face statistics data, the second face statistics data, the third face statistics data, the fourth face statistics data, the sales analysis of the store is carried out . The embodiment of the present invention can analyze the implicit relationship among customers, employees, and commodities that affect the sales of the store through the acquired face statistical data in various areas of the store, thereby improving the sales data of the store in various dimensions, helping to The store locates the customer's purchase expectations, thereby reducing the trial and error cost in the purchase process and ensuring the continuous operation of the store. Through the store sales data analysis method provided by the embodiment of the present invention, it is possible to successfully predict whether the products of the store conform to the population in the area, and which type of people the products of the store are biased towards, so as to assist the store in purchasing better types of goods and optimize the shelf life. Staff arrangement, optimization of shelf arrangement, etc., can effectively predict the future business direction of the store, and guide different groups of people to consume different products, and ultimately increase the overall turnover of the store.

It should be noted that the method for analyzing store sales data provided by the embodiments of the present invention can be applied to devices such as mobile phones, monitors, computers, and servers that can analyze store sales data.

Please refer to Fig. 2, Fig. 2 is the structural representation of a kind of shop sales data analysis device that the embodiment of the present invention provides, as shown in Fig. 2, described device comprises:

The first acquisition module 201 is used to acquire an image of a first designated area, perform face detection on the image of the first designated area, and obtain first face statistical data, and the first designated area is located outside the door of the store area, the first face statistical data includes the face statistical data of pedestrians passing through the store;

The second obtaining module 202 is configured to obtain an image of a second designated area, perform face detection on the image of the second designated area, and obtain second face statistical data, and the second designated area is located in the door area of the store , the second face statistical data includes face statistical data of customers entering the store;

The third acquisition module 203 is configured to acquire a third designated area image, perform face detection on the third designated area image, and obtain third face statistical data, and the third designated area is located in the shelf area of the store , the third face statistical data includes browsing product data, employee and browsing customer face statistical data;

The fourth acquisition module 204 is configured to acquire a fourth designated area image, perform face detection on the fourth designated area image, and obtain fourth face statistical data, and the fourth designated area is located in the cashier area of the store , the fourth face statistical data includes face statistical data of checkout customers and checkout commodity data;

The analysis module 205 is configured to perform a sales analysis on the store according to the first face statistics data, the second face statistics data, the third face statistics data and the fourth face statistics data.

Optionally, as shown in FIG. 3 , the sales analysis includes sales data analysis and sales forecast analysis, and the analysis module 205 includes:

The first processing submodule 2051 is used to form the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data into a graphic report, and according to the graphic Reports to analyze the sales data;

The second processing submodule 2052 is configured to perform correlation calculation according to at least two of the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data, and according to the The sales forecast analysis is performed according to the result of the correlation calculation.

Optionally, the second processing sub-module 2052 is further configured to correspond to any two of the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data. The standard deviation and the covariance between the any two items, calculate the correlation between the any two items, and according to the first face statistical data, the second face statistical data, the third face statistical data and the correlation between any two items in the fourth face statistical data to perform the sales forecast analysis.

Optionally, as shown in FIG. 4 , the second processing sub-module 2052 includes:

The first calculation unit 20521 is configured to calculate according to the standard deviation of the first face statistical data, the standard deviation of the fourth face statistical data, the first face statistical data and the fourth face statistical data covariance between, calculating the first correlation between the first face statistics and the fourth face statistics; and/or

The second calculation unit 20522 is configured to calculate according to the standard deviation of the second human face statistical data, the standard deviation of the fourth human face statistical data, the second human face statistical data and the fourth human face statistical data covariance between, calculating the second correlation between the second face statistics and the fourth face statistics; and/or

The third calculation unit 20523 is configured to calculate according to the standard deviation of the third human face statistical data, the standard deviation of the fourth human face statistical data, the third human face statistical data and the fourth human face statistical data covariance between, calculate the third correlation of the fourth face statistical data of the third face statistical data, and according to the first correlation and/or the second correlation and/or the third correlation to conduct the sales forecast analysis.

Optionally, as shown in FIG. 5 , the first computing unit 20521 includes:

The first statistical subunit 205211 is used to obtain statistics on the flow of pedestrians in the store according to the face statistics data of the pedestrians passing through the store;

a first calculation subunit 205212, configured to calculate the standard deviation of the pedestrian flow;

The second statistics subunit 205213 is used to obtain the number of checkout commodities according to the checkout commodity data;

The second calculation subunit 205214 is used to calculate the standard deviation of the checkout commodity quantity;

The third calculation subunit 205215 is configured to calculate the said pedestrian flow according to the standard deviation of the pedestrian flow, the standard deviation of the checkout item quantity, and the covariance between the pedestrian flow rate and the checkout item quantity The correlation between the flow of pedestrians and the quantity of the checkout commodities is taken as the first correlation.

Optionally, as shown in Figure 6, the second computing unit 20522 includes:

The third statistical subunit 205221 is used to count the number of customers entering the store according to the face statistical data of the customers entering the store;

The fourth calculation subunit 205222 is used to calculate the standard deviation of the number of customers entering the store;

The fifth calculation subunit 205223 is configured to calculate the The correlation between the number of customers entering the store and the number of the checkout commodities is used as the second correlation.

Optionally, as shown in Figure 7, the third computing unit 20523 includes:

The fourth statistical subunit 205231 is used to obtain statistics on the wandering times or time of commodities according to the browsing customer's face statistics data and browsing commodity data;

The sixth calculation subunit 205232 is used to calculate the standard deviation corresponding to the number of times or time the commodity lingers;

The seventh calculation subunit 205233 is configured to calculate the commodity wandering according to the standard deviation of the commodity wandering times, the standard deviation of the checkout commodity quantity, and the covariance between the commodity wandering frequency and the checkout commodity quantity The correlation between the number of times and the number of the checkout commodities is used as the third correlation; or the seventh calculation subunit 205233 is further used to calculate the standard deviation of the lingering time of the commodity, the standard deviation of the checkout commodity quantity, and the lingering time of the commodity and the covariance between the checkout item quantity, respectively calculate the correlation between the item lingering time and the checkout item quantity as a third correlation.

It should be noted that the apparatus for analyzing store sales data provided by the embodiments of the present invention can be applied to devices such as mobile phones, monitors, computers, and servers that can analyze store sales data.

The store sales data analysis device provided by the embodiment of the present invention can realize the various processes implemented by the store sales data analysis method in the above method embodiments, and can achieve the same beneficial effects. In order to avoid repetition, details are not repeated here.

Referring to FIG. 8, FIG. 8 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. As shown in FIG. 8, it includes: a memory 802, a processor 801, and a memory 802 and a processor 801 stored in the memory 802 and available on the processor A computer program running on 801, which:

The processor 801 is used for calling the computer program stored in the memory 802, and performs the following steps:

Optionally, the sales analysis includes sales data analysis and sales forecast analysis. The sales analysis of the store is carried out on the face statistics, including:

Optionally, the correlation calculation performed by the processor 801 is performed according to at least two of the first human face statistical data, the second human face statistical data, the third human face statistical data, and the fourth human face statistical data, The sales forecast analysis is performed according to the result of the correlation calculation, including:

Optionally, according to the standard deviation of the first face statistical data, the standard deviation of the fourth face statistical data, the first face statistical data and the fourth person executed by the processor 801 Covariance between face statistics, calculating the first correlation between the first face statistics and the fourth face statistics, including:

calculating the standard deviation of the pedestrian flow of said pedestrian;

calculating the standard deviation of the checkout item quantity;

Optionally, according to the standard deviation of the second face statistical data, the standard deviation of the fourth face statistical data, the second face statistical data and the fourth person executed by the processor 801 Covariance between face statistics, calculating the second correlation between the second face statistics and the fourth face statistics, including:

Optionally, according to the standard deviation of the third face statistical data, the standard deviation of the fourth face statistical data, the third face statistical data and the fourth person executed by the processor 801 The covariance between the face statistical data, and the third correlation of the fourth face statistical data of the third face statistical data is calculated, including:

It should be noted that the above-mentioned electronic device may be a mobile phone, a monitor, a computer, a server and other devices that can be applied to analyze the sales data of a store.

The electronic device provided by the embodiment of the present invention can realize the various processes realized by the method for analyzing the store sales data in the above method embodiment, and can achieve the same beneficial effect. In order to avoid repetition, details are not repeated here.

Embodiments of the present invention further provide a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, each process of the method for analyzing store sales data provided by the embodiments of the present invention is implemented, and The same technical effect can be achieved, and in order to avoid repetition, details are not repeated here.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. During execution, the processes of the embodiments of the above-mentioned methods may be included. The storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random Access Memory, RAM for short), and the like.

The above disclosures are only preferred embodiments of the present invention, and of course, the scope of the rights of the present invention cannot be limited by this. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

A method for analyzing store sales data, comprising the following steps:

Obtain the first designated area image, perform face detection on the first designated area image, and obtain first face statistical data, the first designated area is located in the area outside the store, and the first face The statistical data includes face statistics of pedestrians passing through the shop;

Obtain the second designated area image, perform face detection on the second designated area image, and obtain second face statistical data, the second designated area is located in the door area of the store, and the second face statistical data The data includes facial statistics of customers entering the store;

Obtain a third designated area image, perform face detection on the third designated area image, and obtain third face statistical data, the third designated area is located in the shelf area of the store, and the third face statistical data Data includes browsing product data, employee and browsing customer facial statistics;

Obtain a fourth designated area image, perform face detection on the fourth designated area image, and obtain fourth face statistical data, the fourth designated area is located in the cashier area of the store, and the fourth face statistical data The data includes the facial statistics of the checkout customer and the checkout item data;

The sales analysis of the store is performed according to the first face statistics data, the second face statistics data, the third face statistics data, and the fourth face statistics data.
The method of claim 1, wherein the sales analysis includes sales data analysis and sales forecast analysis, and the first face statistical data, the second face statistical data, and the third face statistical data The data and the fourth face statistical data are used to analyze the sales of the store, including:

Form the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data into a graphic report, and analyze the sales data according to the graphic report;

The correlation calculation is performed according to at least two of the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data, and the correlation calculation is carried out according to the result of the correlation calculation. Sales forecast analysis.
The method according to claim 2, characterized in that, performing the method according to at least two of the first human face statistical data, the second human face statistical data, the third human face statistical data, and the fourth human face statistical data Relevance calculation, the sales forecast analysis is performed according to the result of the correlation calculation, including:

According to the standard deviation corresponding to any two items in the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data and the covariance between the any two items, Calculate the correlation between the any two items, and carry out the correlation according to any two items in the first face statistical data, the second face statistical data, the third face statistical data, and the fourth face statistical data. The sales forecast analysis.
The method according to claim 2, characterized in that, performing the method according to at least two of the first human face statistical data, the second human face statistical data, the third human face statistical data, and the fourth human face statistical data Relevance calculation, the sales forecast analysis is performed according to the result of the correlation calculation, including:

According to the standard deviation of the first face statistical data, the standard deviation of the fourth face statistical data, and the covariance between the first face statistical data and the fourth face statistical data, calculate the the first correlation between the first face statistics and the fourth face statistics; and/or

According to the standard deviation of the second human face statistical data, the standard deviation of the fourth human face statistical data, and the covariance between the second human face statistical data and the fourth human face statistical data, calculate the a second correlation between the second face statistics and the fourth face statistics; and/or

According to the standard deviation of the third human face statistical data, the standard deviation of the fourth human face statistical data, and the covariance between the third human face statistical data and the fourth human face statistical data, calculate the The third correlation of the fourth face statistical data of the third face statistical data is performed, and the sales forecast analysis is performed according to the first correlation and/or the second correlation and/or the third correlation.
The method according to claim 4, wherein the standard deviation of the first human face statistical data, the standard deviation of the fourth human face statistical data, the first human face statistical data and the Calculate the covariance between the fourth human face statistical data, and calculate the first correlation between the first human face statistical data and the fourth human face statistical data, including:

According to the statistical data of the faces of the pedestrians passing through the shop, the traffic flow of the pedestrians in the shop is obtained according to statistics;

calculating the standard deviation of the pedestrian flow of said pedestrian;

According to the checkout product data, the number of checkout products is obtained by statistics;

calculating the standard deviation of the checkout item quantity;

calculating the covariance between the flow of the pedestrian and the quantity of the checkout item;

According to the standard deviation of the pedestrian flow, the standard deviation of the checkout item quantity, and the covariance between the pedestrian flow rate and the checkout item quantity, calculate the pedestrian flow and the checkout item The correlation of the quantity is taken as the first correlation.
The method according to claim 5, wherein the standard deviation of the second human face statistical data, the standard deviation of the fourth human face statistical data, the second human face statistical data and the and calculating the covariance between the fourth human face statistical data, and calculating the second correlation between the second human face statistical data and the fourth human face statistical data, including:

According to the face statistics of the customers entering the store, the number of customers entering the store is obtained by statistics;

calculating the standard deviation of the number of customers entering the store;

Calculate the number of in-store customers and the check-out commodity according to the standard deviation of the number of in-store customers, the standard deviation of the check-out commodity number, and the covariance between the in-store customer number and the check-out commodity number The correlation of the quantity serves as the second correlation.
The method according to claim 6, wherein the standard deviation of the third human face statistical data, the standard deviation of the fourth human face statistical data, the third human face statistical data and the The covariance between the fourth human face statistical data, and the third correlation of the fourth human face statistical data of the third human face statistical data is calculated, including:

According to the browsing customer's face statistics data and browsing commodity data, the number of times or time of commodity wandering is obtained according to statistics;

Calculate the standard deviation corresponding to the number of times or time the commodity lingers;

According to the standard deviation of the number of times of commodity wandering, the standard deviation of the number of checkout commodities, and the covariance between the number of times of commodity wandering and the number of checkout commodities, the correlation between the number of times of commodity wandering and the number of checkout commodities is calculated sex as a third relevance; or

According to the standard deviation of the commodity lingering time, the standard deviation of the checkout commodity quantity, and the covariance between the commodity lingering time and the checkout commodity quantity, the difference between the commodity lingering time and the checkout commodity quantity is calculated respectively. correlation as a third correlation.
A store sales data analysis device, characterized in that the device comprises:

The first acquisition module is used to acquire a first designated area image, and perform face detection on the first designated area image to obtain first face statistical data, and the first designated area is located in the area outside the store. , the first face statistical data includes the face statistical data of pedestrians passing through the store;

The second acquisition module is configured to acquire an image of a second designated area, perform face detection on the image of the second designated area, and obtain second face statistical data, and the second designated area is located in the door area of the store, The second face statistical data includes face statistical data of customers entering the store;

a third acquisition module, configured to acquire an image of a third designated area, perform face detection on the image of the third designated area, and obtain third face statistical data, and the third designated area is located in the shelf area of the store, The third face statistical data includes browsing product data, employee and browsing customer face statistical data;

a fourth acquisition module, configured to acquire a fourth designated area image, perform face detection on the fourth designated area image, and obtain fourth face statistical data, where the fourth designated area is located in the cashier area of the store, The fourth face statistical data includes face statistical data of checkout customers and checkout commodity data;

The analysis module is configured to analyze the sales of the store according to the first face statistics data, the second face statistics data, the third face statistics data and the fourth face statistics data.
An electronic device, characterized by comprising: a memory, a processor, and a computer program stored on the memory and running on the processor, the processor implementing the computer program as claimed in claim 1 when the processor executes the computer program Steps in the store sales data analysis method described in any one of to 7.
A computer-readable storage medium, characterized in that, a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the store sale according to any one of claims 1 to 7 is realized Steps in a data analysis method.