CN118505261A - Intelligent retail method based on Internet of things - Google Patents

Intelligent retail method based on Internet of things Download PDF

Info

Publication number
CN118505261A
CN118505261A CN202410423279.8A CN202410423279A CN118505261A CN 118505261 A CN118505261 A CN 118505261A CN 202410423279 A CN202410423279 A CN 202410423279A CN 118505261 A CN118505261 A CN 118505261A
Authority
CN
China
Prior art keywords
people
people flow
flow
data
supermarket
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202410423279.8A
Other languages
Chinese (zh)
Inventor
李涛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xuzhou Xinmingzhi Bao Technology Co ltd
Original Assignee
Xuzhou Xinmingzhi Bao Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xuzhou Xinmingzhi Bao Technology Co ltd filed Critical Xuzhou Xinmingzhi Bao Technology Co ltd
Priority to CN202410423279.8A priority Critical patent/CN118505261A/en
Publication of CN118505261A publication Critical patent/CN118505261A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/24323—Tree-organised classifiers
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0202—Market predictions or forecasting for commercial activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
    • G06Q30/0224—Discounts or incentives, e.g. coupons or rebates based on user history
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
    • G06V10/449—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters
    • G06V10/451—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters with interaction between the filter responses, e.g. cortical complex cells
    • G06V10/454—Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/62—Extraction of image or video features relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Evolutionary Computation (AREA)
  • Accounting & Taxation (AREA)
  • Strategic Management (AREA)
  • Finance (AREA)
  • Development Economics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computing Systems (AREA)
  • Multimedia (AREA)
  • Biomedical Technology (AREA)
  • Molecular Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Physics (AREA)
  • Biophysics (AREA)
  • Marketing (AREA)
  • Game Theory and Decision Science (AREA)
  • Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses an intelligent retail method based on the Internet of things, which comprises the following steps: step one: dividing subareas and counting the flow of people; step two: realizing people flow threshold alarming by establishing a real-time monitoring module; step three: the merchant combines the predicted future flow of people to carry out sales promotion; step four: the commodity hot sale region setting is realized through product statistical analysis, wherein the people flow analysis and real-time monitoring module is used for monitoring and managing the people flow in the supermarket in real time so as to provide visual data and an alarm system; the people flow prediction and promotion activity module is used for analyzing historical data and predicting future people flow by using a random forest model so as to help merchants to more accurately formulate promotion activities; the image processing and personnel detecting module is used for processing the image captured by the camera and detecting the existence and movement of the person; the invention has the characteristics of accurate people flow prediction and effectively improving sales performance.

Description

Intelligent retail method based on Internet of things
Technical Field
The invention relates to the technical field of the Internet of things, in particular to an intelligent retail method based on the Internet of things.
Background
Traditional promotional campaigns are often limited in terms of selection time and manner to a particular date, such as holidays or special events. This approach has some limitations because it cannot predict future traffic conditions or changes in shopping behavior. For example, even on days without special activity, certain days may still attract a large number of customers. In addition, even on a day with a high volume of people, sales may not be increased if the merchant does not have a corresponding promotional program. In this case, the effect of the conventional promotional program is limited, and the sales potential of the store cannot be exerted to the maximum extent.
In addition, conventional merchandise display and promotion strategies are generally fixed, lacking the ability to predict traffic from actual demand. This may lead to problems such as the goods being placed in low traffic areas or the promotional strategy being underutilized during high traffic periods. Thus, there is a need for a more intelligent, data-driven method to manage supermarket operations in order to better meet customer needs and to improve overall performance. Therefore, an intelligent retail method based on the internet of things, which has accurate designer flow prediction and can effectively improve sales performance, is necessary.
Disclosure of Invention
The invention aims to provide an intelligent retail method based on the Internet of things, so as to solve the problems in the background technology.
In order to solve the technical problems, the invention provides the following technical scheme: the operation method of the intelligent retail system based on the Internet of things comprises the following steps:
step one: dividing subareas and counting the flow of people;
Step two: realizing people flow threshold alarming by establishing a real-time monitoring module;
step three: the merchant combines the predicted future flow of people to carry out sales promotion;
step four: and setting a commodity hot sale area through product statistical analysis.
According to the above technical solution, the steps of sub-region segmentation and people flow statistics include:
A deep learning neural network comprising CNN and RNN models is used for processing supermarket area images captured by a camera to realize personnel detection and motion tracking;
image preprocessing, including image correction and denoising, to improve image quality;
Extracting features by using a CNN model, including body contour detection and body key feature point recognition;
Dividing the monitoring area into different subareas, and establishing an in-out counter to count the people flow.
According to the above technical solution, the step of extracting features by using the CNN model, including body contour detection and body key feature point identification, includes:
the method comprises the steps of acquiring images from cameras arranged at different positions, preprocessing the images, firstly correcting and denoising the images to ensure the image quality, then extracting features in the images, such as edges, shapes and outlines by a convolution layer, a pooling layer and a full connection layer by means of a CNN model, detecting the body outline of a person, further using the CNN to identify key feature points of the body, such as heads, hands and feet when the existence of the person is detected, and finally using the RNN model to analyze the images between continuous frames to track the movement direction and speed of the person, and meanwhile, using a light flow analysis technology to help determine the feature point difference in adjacent frames so as to know the movement track of the person.
According to the above technical scheme, the steps of dividing the monitoring area into different subareas, establishing an in-out counter for people flow statistics include:
Dividing a monitoring area in the supermarket into different subareas based on the position and the view angle of a camera, wherein the subareas correspond to different corridors, shelves or specific shopping areas, continuously detecting and tracking personnel in each subarea, analyzing images captured by the camera by a system to detect the existence of the personnel and track the movement of the personnel, once the personnel is detected in a certain subarea, recording and tracking the position and movement information of the personnel, establishing an in-out counter for each subarea, realizing counting by detecting the position change of the personnel, and adding one to the corresponding counter when the personnel enters the certain subarea; when people leave, the counter is decremented, the people flow data in different areas are updated in real time by continuously monitoring the entering and exiting conditions in each subarea, the area people flow = the people entering number-the people leaving number, and the whole monitoring area The regional people flow i, n represent different subareas, and the characteristics of the different subareas, such as crowding degree, flow trend and residence time, can be analyzed by combining regional segmentation and people flow data, so that shopping behavior and flow conditions of different areas in a supermarket can be better understood.
According to the above technical scheme, the step of realizing people flow threshold alarming by establishing the real-time monitoring module comprises the following steps:
Setting regional people flow thresholds, acquiring historical people flow data from a monitoring module, grouping the data according to different time periods, such as hours, days or weeks, meanwhile, grouping the data according to different monitoring regions, including entrances, shelves and exit positions, representing the people flow of different regions in different time periods by using a data visualization tool, identifying peak time periods and valley time periods and additional flow possibly caused by specific events (such as promotion activities) by observing the charts, finally, setting people flow thresholds based on analysis results, and when the people flow of a certain region exceeds the preset threshold, automatically triggering an alarm system by a supermarket management team, wherein the supermarket management team can take automatic response and scheduling measures according to the received notification, including reallocating staff resources to solve queuing problems, adjusting layout or taking other measures to improve the conditions of busy regions, the automatic response of the alarm system is helpful to more rapidly respond to crowding and abnormal conditions, shopping experience and management efficiency are improved, and the supermarket operation is improved.
According to the technical scheme, the step of carrying out the promotion activity by combining the predicted future people flow by the merchant comprises the following steps:
configuring random forest model parameters such as the number and depth of trees aiming at a complex people stream mode in a super city;
Future traffic is predicted using the model, and inventory, employee scheduling, special promotional campaigns, and shelf layout are adjusted.
According to the above technical solution, the step of configuring random forest model parameters, such as the number and depth of trees, for the complex people stream mode in the supermarket includes:
The method comprises the steps of collecting historical people flow data, including statistics of people flow per hour and marks of specific events, such as sales promotion activity dates, and preparing other characteristics, such as dates, time, seasonality and sales promotion activity types, wherein the data form the basis of a training model, then dividing the data into a training set and a testing set, wherein the training set comprises historical data, the testing set comprises data of future time periods so as to verify the accuracy of the model, then selecting a random forest model comprising a plurality of decision trees, wherein each tree can consider different characteristics and data subsets, parameters of the model, such as the number of trees and the depth of the trees, are set, in supermarket people flow analysis, the number of the trees is 100 to 200 because of a plurality of areas and complex people flow modes exist in the supermarket, meanwhile, the depth of the trees is required to be deep enough to be a potential detail and a complex mode, the depth range of the trees is 15 to 25, cross validation is used for determining the optimal depth, and then, the model can have good performance in the prediction of the flow of people and finally accurate performance of the model, and the performance of each forest event can be well-influenced by the random forest model or the random forest model, the performance parameters can be well-validated and the performance parameters can be better, the parameters can be well-assessed and the performance parameters of the model can be better are better and the specific parameters are better, the performance is better and the performance can be well validated and better and the models are better captured.
According to the above technical solution, the step of predicting the future people flow using the model, adjusting the inventory, the staff scheduling, the special sales promotion and the shelf layout includes:
According to the predicted future traffic, the business adjusts the stock according to the traffic prediction, ensures that enough products are provided in high demand, and reduces the cost, in addition, the business plans special sales promotion such as discount, gift or special preference according to the predicted peak traffic time to attract more customers and promote sales, at this time, the business can adjust staff scheduling according to the traffic prediction, ensures that better customer service is provided in the peak time, the shelf layout is also adjusted according to the prediction, high-sales volume commodities are placed in the high-traffic area, sales and traffic data are continuously monitored after sales promotion is implemented, and compared with the predicted data to evaluate the effectiveness of the sales promotion strategy, and by the method, the benefits obtained by the sales promotion of the business can be maximized.
According to the above technical scheme, the step of realizing commodity hot-sale region setting through product statistical analysis includes:
Identifying hot-sell products including smooth-sell products, high-gross-profit products and seasonal products, and carrying out product statistical analysis;
combining the people flow prediction data to determine a period and a date of high people flow;
dividing a hot sale area, intensively placing products with more purchasing people, and optimizing product display;
sales data and traffic are continuously monitored, and the layout and promotion strategy of the hot-sell areas are adjusted according to the feedback data to maximize the promotion effect.
According to the above technical solution, the system comprises:
The people flow analysis and real-time monitoring module is used for monitoring and managing the people flow in the supermarket in real time so as to provide visual data and an alarm system to cope with crowding and abnormal conditions;
the people flow prediction and promotion activity module is used for analyzing historical data and predicting future people flow by using a random forest model so as to help merchants to more accurately formulate promotion activities and resource allocation strategies;
the commodity hot-sale region setting and feedback optimizing module is used for predicting and identifying hot-sale products according to sales data and people flow, setting the hot-sale region, and continuously optimizing supermarket layout and sales promotion strategies so as to improve sales benefits and customer experience.
Compared with the prior art, the invention has the following beneficial effects: according to the invention, firstly, through an image processing technology, the system can detect the existence of personnel and track the movement of the personnel, the monitoring area is divided into subareas, the personnel detection and tracking are continuously carried out, so that the personnel flow data of different areas are updated in real time, secondly, a real-time monitoring module is established, an alarm system is configured, the alarm can be automatically triggered to cope with crowding and abnormal conditions, then, future personnel flow prediction is realized through a random forest model, a merchant is helped to better carry out sales promotion activities and adjust operation strategies, finally, the product statistics analysis is utilized to determine the commodity hot-selling area, the mass sales commodity is placed in a concentrated mode, sales data and personnel flow are continuously monitored, so that the layout and sales promotion strategies are continuously improved.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
Fig. 1 is a flowchart of an intelligent retail method based on internet of things according to an embodiment of the present invention;
Fig. 2 is a schematic diagram of module composition of an intelligent retail system based on internet of things according to a second embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiment one: fig. 1 is a flowchart of an intelligent retail method based on internet of things, which is provided in an embodiment of the present invention, and the present embodiment may be applied to a scenario where a supermarket intelligently plans a retail operation of goods, and the method may be performed by the intelligent retail method based on internet of things, provided in the present embodiment, as shown in fig. 1, and the method specifically includes the following steps:
step one: dividing subareas and counting the flow of people;
in the embodiment of the invention, a deep learning neural network is adopted to process images captured by cameras deployed at different positions, and the cameras continuously shoot the area in the supermarket to capture the movement behaviors of people;
For example, images are acquired from cameras deployed at different positions, then the images are preprocessed, firstly, image correction and denoising are carried out to ensure the image quality, then, the characteristics in the images, such as edges, shapes and outlines, are extracted through a convolution layer, a pooling layer and a full connection layer by means of a CNN model, body contour detection of a person is carried out, when existence of the person is detected, CNN is further used for identifying body key characteristic points, such as heads, hands, feet and the like, finally, the images between continuous frames are analyzed by using the RNN model to track the movement direction and speed of the person, and meanwhile, the position difference of the characteristic points in adjacent frames can be determined by using a light flow analysis technology so as to know the movement track of the person;
Illustratively, based on the position and view angle of the camera, the monitoring area in the supermarket is divided into different subareas, the subareas correspond to different corridors, shelves or specific shopping areas, the personnel detection and tracking are continuously carried out on each subarea, the system analyzes the images captured by the camera to detect the existence of the personnel and track the movement of the personnel, once the personnel is detected in a subarea, the position and movement information of the personnel are recorded and tracked, an in-out counter is established for each subarea, the counting is realized by detecting the position change of the personnel, and when the personnel enters the subarea, the corresponding counter is increased by one; when people leave, the counter is decremented, the people flow data in different areas are updated in real time by continuously monitoring the entering and exiting conditions in each subarea, the area people flow = the people entering number-the people leaving number, and the whole monitoring area The regional people flow i, n represent different subareas, and the characteristics of the different subareas, such as crowding degree, flow trend and residence time, can be analyzed by combining regional segmentation and people flow data, so that shopping behavior and flow conditions of different areas in a supermarket can be better understood.
Step two: realizing people flow threshold alarming by establishing a real-time monitoring module;
in the embodiment of the invention, a real-time monitoring module is established, and the people flow conditions of different areas are presented in a visual mode;
By way of example, a regional people flow threshold is set, historical people flow data is obtained from a monitoring module, then the data is grouped according to different time periods, such as hours, days or weeks, meanwhile, the data is also grouped according to different monitoring regions, including key positions of an entrance, a goods shelf, an exit and the like, the people flow of different regions in different time periods is represented by using a data visualization tool, peak time periods and valley time periods can be identified by observing the charts, and additional flows possibly caused by specific events (such as sales promotion activities) can be identified, finally, the people flow threshold is set based on analysis results, when the people flow of a certain region exceeds a preset threshold, an alarm system is automatically triggered, a supermarket management team can take automatic response and scheduling measures according to the received notification, including reallocating staff resources to solve queuing problems, adjusting the goods shelf layout, or taking other measures to improve the condition of the busy region, the alarm system is facilitated to respond to congestion and abnormal conditions more quickly, shopping experience and management efficiency are improved, and supermarket operation efficiency is improved.
Step three: the merchant combines the predicted future flow of people to carry out sales promotion;
in the embodiment of the invention, the random forest model is used for predicting the future people flow so that a merchant can better perform sales promotion;
The method comprises the steps of collecting historical people flow data, including statistics of people flow per hour and marks of specific events, such as sales promotion activity dates, preparing other characteristics, such as dates, time, seasonality, sales promotion activity types and the like, wherein the data form the basis of a training model, then dividing the data into a training set and a testing set, wherein the training set comprises historical data, the testing set comprises data of future time periods so as to verify the accuracy of the model, then selecting a random forest model comprising a plurality of decision trees, wherein each tree can consider different characteristics and data subsets, parameters of the model, such as the number of trees and the depth of the trees, are set, in supermarket people flow analysis, the number of the trees is 100 to 200 because of the existence of a plurality of areas and complex people flow modes in the supermarket, simultaneously, the depth of the trees is required to be deep enough to capture potential details and complex modes, the depth range of the trees is between 15 and 25, and cross verification is used for determining the optimal depth, so that the models have the characteristics of accurately and the performance of the models, such as the random forest models have the accuracy and the performance of the models are well-predicted, and the performance coefficients are well-influenced by the specific events, such as the random forest models are well, the specific events are well-predicted, the parameters are well-being well-influenced, and the performance parameters of the performance of the models are well is well-being well estimated, such as the specific events are well-predicted, and the performance parameters are well-predicted events are well-estimated to be well-estimated for the forest-estimated events and the performance-statistics is well-estimated events;
Illustratively, the sales promotion is performed according to the predicted future traffic, the inventory is adjusted by the merchant according to the traffic prediction, the sufficient product is ensured to be provided in high demand, and the cost is reduced, in addition, the merchant plans special sales promotion such as discount, gift or special preference according to the predicted peak traffic period to attract more customers and promote sales, at this time, the merchant can adjust staff scheduling according to the traffic prediction, the method ensures that better customer service is provided in the peak period, the shelf layout is also adjusted according to the prediction, the high-sales-volume commodities are placed in the high-flow area, sales and people flow data are continuously monitored after sales promotion is implemented, and compared with the prediction data to evaluate the effectiveness of the sales promotion strategy, so that the benefits obtained by the sales promotion of merchants can be maximized.
Step four: and setting a commodity hot sale area through product statistical analysis.
In the embodiment of the invention, firstly, through product statistics analysis, a system can identify which products are favored by customers in the past, including free selling goods, high-priced goods or popular seasonal goods, then, people flow prediction data are utilized to determine which time periods or dates have larger people flow, such as weekends, holidays or specific promotion days, then, merchants divide the hot-sell areas, after the hot-sell areas are determined, the merchants concentrate the products with more purchase numbers in the areas, so that the display positions of the products are highlighted and the customers are convenient to browse and purchase;
Illustratively, with continuous monitoring of sales data and traffic, merchants can know the effectiveness of the hot areas and the effectiveness of the promotion strategies in real time through the visual map, by analyzing the sales data, merchants can determine which products are excellent in the hot areas, and whether the promotion strategies are successful in attracting more customers, such feedback is key to continuous improvement and optimization, and based on the feedback of the data, merchants can timely adjust the layout of the hot areas and the promotion strategies, for example, if a certain hot product is not sold as expected, merchants can reconsider the display mode or promotion strategy, otherwise, if a certain promotion strategy is very successful, merchants can further popularize and expand the strategy; by combining product statistical analysis with people flow prediction, merchants can more intelligently manage supermarket layout and sales strategies to achieve higher sales benefits, and the data-driven method enables merchants to make decisions according to actual performances, continuously improve and adapt to market changes, so that competitiveness is maintained and better shopping experience is provided.
Embodiment two: the second embodiment of the present invention provides an intelligent retail system based on internet of things, and fig. 2 is a schematic diagram of module composition of the intelligent retail system based on internet of things provided in the second embodiment of the present invention, as shown in fig. 2, the system includes:
The people flow analysis and real-time monitoring module is used for monitoring and managing the people flow in the supermarket in real time so as to provide visual data and an alarm system to cope with crowding and abnormal conditions;
the people flow prediction and promotion activity module is used for analyzing historical data and predicting future people flow by using a random forest model so as to help merchants to more accurately formulate promotion activities and resource allocation strategies;
The commodity hot-sale region setting and feedback optimizing module is used for predicting and identifying hot-sale products according to sales data and the traffic flow, setting the hot-sale region, and continuously optimizing supermarket layout and sales promotion strategies so as to improve sales benefits and customer experience;
in some embodiments of the present invention, the people flow analysis and real-time monitoring module includes:
the image processing and personnel detecting module is used for processing the image captured by the camera and detecting the existence and movement of the person;
the regional segmentation and people flow calculation module is used for segmenting the monitoring region into different subareas and calculating the people flow in each subarea;
the real-time monitoring and alarming system module is used for displaying the traffic conditions of different areas in real time and configuring an alarming system to deal with congestion and abnormal conditions;
In some embodiments of the invention, the people flow prediction and promotion module includes:
The historical data analysis and specific event identification module is used for analyzing historical data, including people flow and specific events, and identifying shopping modes and specific events;
the random forest model and future people flow prediction module is used for predicting the future people flow by using the random forest model so as to support sales promotion and resource scheduling;
the promotion activity execution and effect evaluation module is used for executing promotion activities based on people flow prediction, evaluating the effectiveness of promotion strategies and adjusting according to data feedback;
in some embodiments of the present invention, the commodity hot zone setup and feedback optimization module comprises:
the product statistical analysis and hot sale product identification module is used for identifying which products are popular in the past through product statistical analysis;
The regional setting and hot-pin product placement module is used for setting a hot-pin region according to people flow prediction and intensively placing hot-pin products;
And the continuous monitoring and strategy adjusting module is used for continuously monitoring sales data and people flow, and adjusting the distribution of the hot sale area and the promotion strategy according to data feedback.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Finally, it should be noted that: the foregoing description is only a preferred embodiment of the present invention, and the present invention is not limited thereto, but it is to be understood that modifications and equivalents of some of the technical features described in the foregoing embodiments may be made by those skilled in the art, although the present invention has been described in detail with reference to the foregoing embodiments. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. The intelligent retail method based on the Internet of things comprises the following steps of:
step one: dividing subareas and counting the flow of people;
Step two: realizing people flow threshold alarming by establishing a real-time monitoring module;
step three: the merchant combines the predicted future flow of people to carry out sales promotion;
step four: and setting a commodity hot sale area through product statistical analysis.
2. The internet of things-based intelligent retail method of claim 1, wherein: the sub-region segmentation and people flow statistics steps comprise:
A deep learning neural network comprising CNN and RNN models is used for processing supermarket area images captured by a camera to realize personnel detection and motion tracking;
image preprocessing, including image correction and denoising, to improve image quality;
Extracting features by using a CNN model, including body contour detection and body key feature point recognition;
Dividing the monitoring area into different subareas, and establishing an in-out counter to count the people flow.
3. The internet of things-based intelligent retail method of claim 2, wherein: the step of extracting the characteristics by using the CNN model, including body contour detection and body key characteristic point identification, comprises the following steps:
the method comprises the steps of acquiring images from cameras arranged at different positions, preprocessing the images, firstly correcting and denoising the images to ensure the image quality, then extracting features in the images, such as edges, shapes and outlines by a convolution layer, a pooling layer and a full connection layer by means of a CNN model, detecting the body outline of a person, further using the CNN to identify key feature points of the body, such as heads, hands and feet when the existence of the person is detected, and finally using the RNN model to analyze the images between continuous frames to track the movement direction and speed of the person, and meanwhile, using a light flow analysis technology to help determine the feature point difference in adjacent frames so as to know the movement track of the person.
4. The internet of things-based intelligent retail method of claim 2, wherein: the step of dividing the monitoring areas into different subareas and establishing an in-out counter to count the flow of people comprises the following steps:
Dividing a monitoring area in the supermarket into different subareas based on the position and the view angle of a camera, wherein the subareas correspond to different corridors, shelves or specific shopping areas, continuously detecting and tracking personnel in each subarea, analyzing images captured by the camera by a system to detect the existence of the personnel and track the movement of the personnel, once the personnel is detected in a certain subarea, recording and tracking the position and movement information of the personnel, establishing an in-out counter for each subarea, realizing counting by detecting the position change of the personnel, and adding one to the corresponding counter when the personnel enters the certain subarea; when people leave, the counter is decremented, the people flow data in different areas are updated in real time by continuously monitoring the entering and exiting conditions in each subarea, and the area people flow = people entering number-people leaving number is monitored in the whole N represents different subareas, and features of the different subareas, such as crowding degree, flow trend and residence time, can be analyzed by combining the subarea segmentation and the people flow data, so that shopping behavior and flow conditions of different areas in a supermarket can be better understood.
5. The internet of things-based intelligent retail method of claim 1, wherein: the step of realizing people flow threshold alarming by establishing the real-time monitoring module comprises the following steps:
Setting regional people flow thresholds, acquiring historical people flow data from a monitoring module, grouping the data according to different time periods, such as hours, days or weeks, meanwhile, grouping the data according to different monitoring regions, including entrances, shelves and exit positions, representing the people flow of different regions in different time periods by using a data visualization tool, identifying peak time periods and valley time periods and additional flow possibly caused by specific events (such as promotion activities) by observing the charts, finally, setting people flow thresholds based on analysis results, and when the people flow of a certain region exceeds the preset threshold, automatically triggering an alarm system by a supermarket management team, wherein the supermarket management team can take automatic response and scheduling measures according to the received notification, including reallocating staff resources to solve queuing problems, adjusting layout or taking other measures to improve the conditions of busy regions, the automatic response of the alarm system is helpful to more rapidly respond to crowding and abnormal conditions, shopping experience and management efficiency are improved, and the supermarket operation is improved.
6. The internet of things-based intelligent retail method of claim 1, wherein: the step of the merchant carrying out a promotional program in combination with the predicted future traffic of people comprises the steps of:
configuring random forest model parameters such as the number and depth of trees aiming at a complex people stream mode in a super city;
Future traffic is predicted using the model, and inventory, employee scheduling, special promotional campaigns, and shelf layout are adjusted.
7. The internet of things-based intelligent retail method of claim 6, wherein: the step of configuring random forest model parameters, such as the number and depth of trees, for the complicated people stream mode in the supermarket comprises the following steps:
The method comprises the steps of collecting historical people flow data, including statistics of people flow per hour and marks of specific events, such as sales promotion activity dates, and preparing other characteristics, such as dates, time, seasonality and sales promotion activity types, wherein the data form the basis of a training model, then dividing the data into a training set and a testing set, wherein the training set comprises historical data, the testing set comprises data of future time periods so as to verify the accuracy of the model, then selecting a random forest model comprising a plurality of decision trees, wherein each tree can consider different characteristics and data subsets, parameters of the model, such as the number of trees and the depth of the trees, are set, in supermarket people flow analysis, the number of the trees is 100 to 200 because of a plurality of areas and complex people flow modes exist in the supermarket, meanwhile, the depth of the trees is required to be deep enough to be a potential detail and a complex mode, the depth range of the trees is 15 to 25, cross validation is used for determining the optimal depth, and then, the model can have good performance in the prediction of the flow of people and finally accurate performance of the model, and the performance of each forest event can be well-influenced by the random forest model or the random forest model, the performance parameters can be well-validated and the performance parameters can be better, the parameters can be well-assessed and the performance parameters of the model can be better are better and the specific parameters are better, the performance is better and the performance can be well validated and better and the models are better captured.
8. The internet of things-based intelligent retail method of claim 6, wherein: the step of predicting future traffic using the model, adjusting inventory, employee scheduling, special promotional campaigns, and shelf layout, comprises:
According to the predicted future traffic, the business adjusts the stock according to the traffic prediction, ensures that enough products are provided in high demand, and reduces the cost, in addition, the business plans special sales promotion such as discount, gift or special preference according to the predicted peak traffic time to attract more customers and promote sales, at this time, the business can adjust staff scheduling according to the traffic prediction, ensures that better customer service is provided in the peak time, the shelf layout is also adjusted according to the prediction, high-sales volume commodities are placed in the high-traffic area, sales and traffic data are continuously monitored after sales promotion is implemented, and compared with the predicted data to evaluate the effectiveness of the sales promotion strategy, and by the method, the benefits obtained by the sales promotion of the business can be maximized.
9. The internet of things-based intelligent retail method of claim 1, wherein: the step of realizing commodity hot sale region setting through product statistical analysis comprises the following steps:
Identifying hot-sell products including smooth-sell products, high-gross-profit products and seasonal products, and carrying out product statistical analysis;
combining the people flow prediction data to determine a period and a date of high people flow;
dividing a hot sale area, intensively placing products with more purchasing people, and optimizing product display;
sales data and traffic are continuously monitored, and the layout and promotion strategy of the hot-sell areas are adjusted according to the feedback data to maximize the promotion effect.
10. A smart retail system that performs the internet of things-based smart retail method as recited in claim 1, characterized by: the system comprises:
The people flow analysis and real-time monitoring module is used for monitoring and managing the people flow in the supermarket in real time so as to provide visual data and an alarm system to cope with crowding and abnormal conditions;
the people flow prediction and promotion activity module is used for analyzing historical data and predicting future people flow by using a random forest model so as to help merchants to more accurately formulate promotion activities and resource allocation strategies;
the commodity hot-sale region setting and feedback optimizing module is used for predicting and identifying hot-sale products according to sales data and people flow, setting the hot-sale region, and continuously optimizing supermarket layout and sales promotion strategies so as to improve sales benefits and customer experience.
CN202410423279.8A 2024-04-09 2024-04-09 Intelligent retail method based on Internet of things Pending CN118505261A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410423279.8A CN118505261A (en) 2024-04-09 2024-04-09 Intelligent retail method based on Internet of things

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410423279.8A CN118505261A (en) 2024-04-09 2024-04-09 Intelligent retail method based on Internet of things

Publications (1)

Publication Number Publication Date
CN118505261A true CN118505261A (en) 2024-08-16

Family

ID=92245656

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410423279.8A Pending CN118505261A (en) 2024-04-09 2024-04-09 Intelligent retail method based on Internet of things

Country Status (1)

Country Link
CN (1) CN118505261A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100036706A1 (en) * 2005-06-24 2010-02-11 Beonic Corporation Pty Ltd Queue early warning system
CN109871804A (en) * 2019-02-19 2019-06-11 上海宝尊电子商务有限公司 A kind of method and system of shop stream of people discriminance analysis
CN112330359A (en) * 2020-11-04 2021-02-05 上饶市中科院云计算中心大数据研究院 Smart tourist attraction saturation evaluation method and device
CN112950254A (en) * 2021-01-27 2021-06-11 北京市商汤科技开发有限公司 Information processing method and device, electronic equipment and storage medium
CN113473592A (en) * 2021-06-21 2021-10-01 上海邮电设计咨询研究院有限公司 Passenger flow monitoring system and method
CN116029395A (en) * 2023-03-24 2023-04-28 深圳市明源云科技有限公司 Pedestrian flow early warning method and device for business area, electronic equipment and storage medium

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100036706A1 (en) * 2005-06-24 2010-02-11 Beonic Corporation Pty Ltd Queue early warning system
CN109871804A (en) * 2019-02-19 2019-06-11 上海宝尊电子商务有限公司 A kind of method and system of shop stream of people discriminance analysis
CN112330359A (en) * 2020-11-04 2021-02-05 上饶市中科院云计算中心大数据研究院 Smart tourist attraction saturation evaluation method and device
CN112950254A (en) * 2021-01-27 2021-06-11 北京市商汤科技开发有限公司 Information processing method and device, electronic equipment and storage medium
CN113473592A (en) * 2021-06-21 2021-10-01 上海邮电设计咨询研究院有限公司 Passenger flow monitoring system and method
CN116029395A (en) * 2023-03-24 2023-04-28 深圳市明源云科技有限公司 Pedestrian flow early warning method and device for business area, electronic equipment and storage medium

Similar Documents

Publication Publication Date Title
US8812344B1 (en) Method and system for determining the impact of crowding on retail performance
US8577705B1 (en) Method and system for rating the role of a product category in the performance of a store area
US10915131B2 (en) System and method for managing energy
US10713670B1 (en) Method and system for finding correspondence between point-of-sale data and customer behavior data
US8189926B2 (en) Method and system for automatically analyzing categories in a physical space based on the visual characterization of people
US9747497B1 (en) Method and system for rating in-store media elements
US8873794B2 (en) Still image shopping event monitoring and analysis system and method
JP4972491B2 (en) Customer movement judgment system
US8254625B2 (en) Automated service measurement, monitoring and management
US20070253595A1 (en) Still Image Queue Analysis System and Method
US9846811B2 (en) System and method for video-based determination of queue configuration parameters
JP6517325B2 (en) System and method for obtaining demographic information
KR20170028815A (en) Analytics system and method
Jung et al. Grocery customer behavior analysis using RFID-based shopping paths data
CN118505261A (en) Intelligent retail method based on Internet of things
CN117236996B (en) User behavior prediction method and system based on big data analysis
Tarokh et al. A new model to speculate CLV based on Markov chain model
US11151584B1 (en) Method and system for collecting shopper response data tied to marketing and merchandising elements
Hossam et al. Revolutionizing Retail Analytics: Advancing Inventory and Customer Insight with AI
US12143756B2 (en) Multi-camera person re-identification
US20240330823A1 (en) Risk assessment with prescriptive recommendations
US20230112584A1 (en) Multi-camera person re-identification
CN115131070B (en) Online commodity predetermined quantity virtual mark identification and processing method and device
Keh et al. Video-Based Gender Profiling on Challenging Camera Viewpoint for Restaurant Data Analytics
CN117670419A (en) ERP system-based passenger flow analysis method, device, equipment and medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination