A dynamic intelligent recommendation method and system based on changes in characteristics of both supply and demand sides

Abstract

The present invention discloses a dynamic intelligent recommendation method and system based on the change of characteristics of both supply and demand sides. The management server sends the registered customer information to each camera. The store smart terminal and the camera carry their own geographical location information when joining the multicast group. The store smart terminal obtains the customer queuing information of the store and sends it to its corresponding smart terminal gateway. The smart terminal gateway inserts the smart terminal geographical location into the customer queuing information and sends it to all camera gateways in the network. The camera gateway calculates the ranking index of the store smart terminal relative to the camera, adds it to the customer queuing information of the store smart terminal and sends it to the camera. After the camera identifies the target customer, it finds the application information corresponding to the target customer and sends the recommended customer queuing information. The present invention provides convenience for customers to choose the most suitable store to go to at the moment, and improves the customer's consumption experience.

Description

Technical Field

[0001] The present application belongs to the field of information recommendation technology, and in particular, relates to a dynamic intelligent recommendation method and system based on changes in characteristics of both supply and demand sides. Background Art

[0002] Large shopping districts are popular among people because they integrate catering, shopping, entertainment and activities. In order to provide convenient services to customers entering the shopping districts, many shopping districts provide information recommendation systems to recommend information to customers' smartphones with applications (APPs) installed.

[0003] However, current information recommendations are generally made to customers indiscriminately, and the actual effect of the recommendations and the customer experience are not ideal. Summary of the invention

[0004] The purpose of this application is to provide a dynamic intelligent recommendation method and system based on changes in the characteristics of both supply and demand sides, which is used to sort information when recommending information to customers, thereby improving recommendation effects and customer experience.

[0005] In order to achieve the above purpose, the technical solution of this application is as follows: A dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties, comprising: The management server sends the registered customer information to each camera and sends a unified multicast group address to the camera and the store smart terminal; When the store's smart terminals and cameras join the multicast group, they carry their own geographic location information, which is recorded by the smart terminal gateway and camera gateway respectively; The store smart terminal obtains the customer queue information of the store and sends it to its corresponding smart terminal gateway. The smart terminal gateway inserts the smart terminal's geographical location into the customer queue information and sends it to all camera gateways in the network; The camera gateway calculates the ranking index of the store smart terminal relative to the camera, adds it to the customer queue information of the store smart terminal and sends it to the camera; After the camera identifies the target customer, it finds the application information corresponding to the target customer and sends the customer queue information of all the store smart terminals recorded by the camera to the application corresponding to the target customer.

[0006] Further, the calculating the ranking index of the intelligent terminal relative to the camera includes: Set the initial value of the ranking index. When one group of customers is added to the smart terminal customer queue information, the ranking index decreases by 1, and when one group is reduced, the ranking index increases by 1. Before the first time, when the walking path between the smart terminal and the camera increases by a preset distance, the ranking index decreases by 1; After the first time, the ranking index increases by 1 for each additional preset distance between the walking path between the smart terminal and the camera.

[0007] Furthermore, the dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties also includes: The first camera records the timestamp of sending the customer queue information to the target customer's application, and sends the timestamp of the target customer to other cameras; When the second camera recognizes the same target customer, it calculates the time difference between the current time and the timestamp of the target customer. If the time difference is less than the preset time threshold, the customer queue information is not sent to the target customer's application. Otherwise, the customer queue information is sent to the target customer's application.

[0008] Furthermore, the dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties also includes: After the first camera sends the customer queue information to the target customer's application, it sends the binding relationship between the target customer and the geographic location information of the first camera to other cameras; When the second camera recognizes the same target customer, it calculates whether the walking path between the second camera and the first camera is less than a preset distance threshold. If so, the customer queuing information is not sent to the target customer's application. Otherwise, the customer queuing information is sent to the target customer's application.

[0009] This application also proposes a dynamic intelligent recommendation system based on changes in the characteristics of both supply and demand sides, including a management server, a store intelligent terminal and a camera, wherein: The management server sends the registered customer information to each camera and sends a unified multicast group address to the camera and the store smart terminal; When the store's smart terminals and cameras join the multicast group, they carry their own geographic location information, which is recorded by the smart terminal gateway and camera gateway respectively; The store smart terminal obtains the customer queue information of the store and sends it to its corresponding smart terminal gateway. The smart terminal gateway inserts the smart terminal's geographical location into the customer queue information and sends it to all camera gateways in the network; The camera gateway calculates the ranking index of the store smart terminal relative to the camera, adds it to the customer queue information of the store smart terminal and sends it to the camera; After the camera identifies the target customer, it finds the application information corresponding to the target customer and sends the customer queue information of all the store smart terminals recorded by the camera to the application corresponding to the target customer.

[0010] Furthermore, the camera gateway calculates the ranking index of the smart terminal relative to the camera and performs the following operations: Set the initial value of the ranking index. When one group of customers is added to the smart terminal customer queue information, the ranking index decreases by 1, and when one group is reduced, the ranking index increases by 1. Before the first time, when the walking path between the smart terminal and the camera increases by a preset distance, the ranking index decreases by 1; After the first time, the ranking index increases by 1 for each additional preset distance between the walking path between the smart terminal and the camera.

[0011] Furthermore, the dynamic intelligent recommendation system based on changes in characteristics of both supply and demand parties also includes: The first camera records the timestamp of sending the customer queue information to the target customer's application, and sends the timestamp of the target customer to other cameras; When the second camera recognizes the same target customer, it calculates the time difference between the current time and the timestamp of the target customer. If the time difference is less than the preset time threshold, the customer queue information is not sent to the target customer's application. Otherwise, the customer queue information is sent to the target customer's application.

[0012] Furthermore, the dynamic intelligent recommendation system based on changes in characteristics of both supply and demand parties also includes: After the first camera sends the customer queue information to the target customer's application, it sends the binding relationship between the target customer and the geographic location information of the first camera to other cameras; When the second camera recognizes the same target customer, it calculates whether the walking path between the second camera and the first camera is less than a preset distance threshold. If so, the customer queuing information is not sent to the target customer's application. Otherwise, the customer queuing information is sent to the target customer's application.

[0013] This application proposes a dynamic intelligent recommendation method and system based on the changes in the characteristics of both supply and demand. The camera gateway calculates the ranking index of the intelligent terminal relative to the camera, taking into account the number of customers queuing in the store and the walking path of the customers to the store, so as to facilitate customers to choose the most suitable store to go to at the moment and improve the customer's consumption experience. At the same time, it also improves the effect of store recommendations, making it easier for stores to provide customers with satisfactory services. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of the dynamic intelligent recommendation method based on changes in the characteristics of both supply and demand sides in this application.

[0015] Figure 2This is a schematic diagram of the structure of the dynamic intelligent recommendation system based on changes in the characteristics of both supply and demand sides in this application. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0017] Example 1: A dynamic intelligent recommendation method based on the changes in the characteristics of both supply and demand parties is proposed. Figure 1 As shown, including: Step S1: The management server sends the registered customer information to each camera and sends a unified multicast group address to the camera and the store smart terminal.

[0018] In this embodiment, for a business district, there is an information system, which includes a management server, smart terminals of each store, and cameras distributed in the business district. Figure 2 As shown, the smart terminal and the camera are connected to the information system network through their respective gateways, and the management server is also connected to the network. The information system network is networked through multicast routers, and each device can exchange information through multicast. For example, cameras 1, 2, and 3 are connected to the network through gateway B, and smart terminals 1 and 2 are connected to the network through gateway A.

[0019] In this embodiment, each camera has its own identification ID and built-in geographic location information. Smart terminals also have their own identification ID and built-in geographic location information. Cameras and smart terminals need to register with the management server, which records the identification and geographic location information of each camera and smart terminal. At the same time, the management server also records the information of the store to which the smart terminal belongs. Each store has a smart terminal, which will collect the customer queue information of the corresponding store.

[0020] In this embodiment, customers in the business district carry smartphones by default, and the smartphones have corresponding application programs installed in them to receive recommendation information. The management server sends the facial feature extraction model to all cameras. The management server has the corresponding relationship between all members' IDs, facial feature vectors, and APP-IDs. Members here refer to customers whose smartphones have installed the system's specific application APP and registered, so as to provide services for these members.

[0021] By default, customers in the business district carry smartphones with corresponding applications installed in them to receive recommendation information. Customers can register with the management server through the APP to become members, so that the management server records customer information, which may include customer number, APP number, etc. In this embodiment, the management server sends the registered customer information to each camera, and the camera can send recommendation information to the customer based on the customer information.

[0022] Step S2: When the store smart terminal and camera join the multicast group, they carry their own geographical location information, and the smart terminal gateway and camera gateway record them respectively.

[0023] When networking, the management server sends a unified multicast group address, such as 226.1.1.1, to the cameras and smart terminals to establish a multicast tree for message forwarding. The cameras in the business district and the smart terminals in each store have built-in geographic location information, such as longitude: 116.4074°E and latitude: 39.9042°N. When they send IGMP join messages to the gateway (A, B), they carry their own geographic location information. The gateway router (A, B) records the IP address of each smart terminal or camera connected to the gateway and its corresponding geographic location information. The gateway router (A, B) sends a PIM (*, G) join message to the PIM upstream router, so the (*, G) multicast forwarding tree is established in the entire network. Any multicast data message with a destination address of G (such as 226.1.1.1) can enter through one of the (*, G) table entries and be forwarded out by other interfaces.

[0024] Step S3: The shop smart terminal obtains the customer queuing information of the shop and sends it to its corresponding smart terminal gateway. The smart terminal gateway inserts the customer queuing information with the smart terminal's geographic location and sends it to all camera gateways in the network.

[0025] Specifically, smart terminal 1 belongs to shop 1. When smart terminal 1 finds that there are N groups of customers waiting in shop 1, it sends this information to its gateway A in a multicast manner. This message is called "customer queue information", which contains the shop name and the number of customer waiting groups. Smart terminal 1 will send the update of the number of waiting groups to gateway A in a timely manner. After receiving the update, gateway A extracts the geographical location information of smart terminal 1 from the local area and adds it to the "customer queue information" message. Or smart terminal 1 carries its own geographical location information when sending customer queue information.

[0026] Customer queue information reaches the gateways of all cameras, such as gateway B of camera 1, through the established (*, G), such as (*, 226.1.1.1) multicast forwarding tree. It is easy to understand that Figure 2In the figure, cameras 1, 2 and 3 are connected to gateway B. There may be other cameras connected to the network through other gateways, which are not described here one by one.

[0027] Step S4: The camera gateway calculates the ranking index of the shop smart terminal relative to the camera, adds it to the customer queue information of the shop smart terminal and sends it to the camera.

[0028] The camera gateways in this embodiment have built-in electronic map services, such as Amap or other brands of maps, which can be used to calculate the walking path and walking time between two locations. For any camera, this embodiment calculates the ranking index of all store smart terminals relative to the camera, taking into account the number of customers queuing in the store and the walking path of customers to the store.

[0029] Calculate the ranking index of smart terminals relative to cameras as follows: Set the initial value of the ranking index. When one group of customers is added to the smart terminal customer queue information, the ranking index decreases by 1, and when one group is reduced, the ranking index increases by 1. Before the first time, when the walking path between the smart terminal and the camera increases by a preset distance, the ranking index decreases by 1; After the first time, the ranking index increases by 1 for each additional preset distance between the walking path between the smart terminal and the camera.

[0030] Specifically, taking camera 1 as an example, the ranking index of all smart terminals is calculated. The initial value of smart terminal 1 relative to camera 1 is 100, and the ranking index decreases by 1 every time a group of customers waits for smart terminal 1, and increases by 1 every time a group of customers waits for smart terminal 1.

[0031] Before 12 noon, if the walking distance between smart terminal 1 and camera 1 is within 50 meters, the ranking index will be reduced by 1; if it exceeds 50 meters but does not exceed 100 meters, the ranking index will be reduced by 2, that is, the ranking index will be reduced by 1 for every additional 50 meters; After 12 noon, if the distance between the smart terminal 1 and the camera 1 is within 50 meters, the ranking index increases by 1; if it exceeds 50 meters but does not exceed 100 meters, the ranking index increases by 2, that is, the ranking index increases by 1 for every additional 50 meters.

[0032] It should be noted that the calculation method of the above ranking index takes the shop as a restaurant as an example. Considering that each shop has the same popularity and 12 o'clock is the peak dining time, before 12 o'clock, customers should be given priority to recommend shops that are close, because the closer the time to arrive at the shop is to 12 o'clock, the more serious the queue will be; and after 12 o'clock, customers should be given priority to recommend restaurants that are far away, because the later the arrival, the less waiting time. In addition, the distance from the customer to the shop is judged by the walking path between the camera and the smart terminal, rather than the straight-line distance, which can more objectively reflect the distance that the customer actually needs to walk to the shop. The method of calculating the ranking index in this embodiment reflects the dynamics of the changes in the characteristics of both the supply and demand sides, can accurately reflect the suitability of the shop for the customer's choice, and facilitates customers to find the most suitable shop to go to at the moment.

[0033] Therefore, gateway B calculates the ranking index of each smart terminal for camera 1, for example, smart terminal 1, ranking index 95; smart terminal 2, ranking index 93.

[0034] Then, gateway B adds the calculated ranking index to the customer queuing information of the shop smart terminal 1 and sends it to camera 1, so that camera 1 obtains the customer queuing information of smart terminal 1, that is, camera 1 records the shop name, ranking index and number of customer waiting groups corresponding to smart terminal 1.

[0035] Similarly, through the above method, the camera 1 can also obtain the customer queue information of the smart terminal 2, which will not be repeated here.

[0036] It is easy to understand that when the smart terminal 1 or 2 updates the number of customer waiting groups, the relevant information will be sent to the gateway B via multicast, and the gateway B will send the calculated updated ranking index to each camera.

[0037] Based on the same method, other cameras can also receive customer queuing information from smart terminals, and record information such as the store name, ranking index, and number of customer waiting groups corresponding to each smart terminal.

[0038] It should be noted that, taking catering as an example, one table of customers is a group, and the number of customer waiting groups is the number of tables of customers waiting. It can also be subdivided into several groups of customers waiting for a table of 2, several groups of customers waiting for a table of 4, several groups of customers waiting for a table of 8, etc.

[0039] Step S5: After identifying the target customer, the camera finds the application information corresponding to the target customer and sends the customer queue information of all the store smart terminals recorded by the camera to the application corresponding to the target customer.

[0040] In a common implementation, after capturing customer Z1 in the business district, camera 1 extracts the customer's facial feature vector, for example, [0.1, -0.2, 0.3, ..., 0.7, -0.8, 0.9], and sends the facial feature vector and the customer queue information of all smart terminals in the shops recorded by the camera to the management server. The management server finds the corresponding member's APP-ID (0x000001) based on the facial feature vector, and sends the customer queue information of all smart terminals in the shops recorded by camera 1 to the customer's smartphone based on the APP-ID.

[0041] In this way, customers can receive customer queue information from all current store smart terminals, and choose the store that best suits them to dine in based on the ranking index.

[0042] However, in order to reduce the pressure on the management server and prevent the management server from becoming a bottleneck for information push, this embodiment adopts a distributed implementation method, placing all the steps that require calculation in the gateway or camera for execution, thereby flexibly utilizing the computing power of the devices in the entire network. As a registration and management device, the management server does not require much computing power, which reduces the cost of networking.

[0043] The management server will update the member information in time and send it to all cameras, including the member ID and the corresponding facial feature vector, and the connection address information of the member APP (which can be accessed through the network). When camera 1 finds customer Z1, it extracts the facial feature vector and performs a similarity comparison. If it is found that the similarity between the feature vector of a member and customer Z1 exceeds the threshold, the store information is pushed to the APP of customer Z1 through the connection address of the customer APP. The store information is sorted according to the ranking index. Obviously, this method can avoid the management server from calculating and comparing the feature vectors of members in large quantities and pushing messages, which causes a waste of network bandwidth and camera computing resources.

[0044] Embodiment 2: This embodiment aims at the situation where there are relatively concentrated cameras in one area. A customer may be captured by multiple cameras, resulting in repeated push of information, causing discomfort to the customer.

[0045] In this embodiment, the dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties further includes: The first camera records the timestamp of sending the customer queue information to the target customer's application, and sends the timestamp of the target customer to other cameras; When the second camera recognizes the same target customer, it calculates the time difference between the current time and the timestamp of the target customer. If the time difference is less than the preset time threshold, the customer queue information is not sent to the target customer's application. Otherwise, the customer queue information is sent to the target customer's application.

[0046] In Example 1, camera 1 has captured customer Z1 and accordingly sent recommendation information (i.e., customer queue information of the store smart terminal) to customer Z1. Different from Example 1, in this example, when camera 1 sends the recommendation information, it records the timestamp of sending the customer queue information to the target customer's application, for example, 11:10 on November 19, 2024.

[0047] When camera 2 also captures customer Z1, the time difference between the current time and the timestamp of the target customer is calculated. If the time difference is less than the preset time threshold, the customer queue information is not sent to the target customer's application. This embodiment avoids repeated disturbing of customers based on the time difference.

[0048] Specifically, when camera 1 captures customer Z1, it generates a facial feature vector, performs facial matching based on the registered customer information sent by the management server, and sends recommendation information to customer Z1 after a successful match. At this time, camera 1 needs to multicast the ID of customer Z1 to the entire network through the (*, G) multicast forwarding tree, and add the current timestamp (for example, November 19, 2024 11:10), and at the same time add the current timestamp to the ID of customer Z1 saved by itself.

[0049] After receiving the message, other cameras, such as Camera 2, extract the customer ID and the above timestamp (November 19, 2024 11:10) in the message, and locally bind the above timestamp to the member corresponding to the customer ID. That is, Camera 2 records customer Z1 and the timestamp November 19, 2024 11:10. When Camera 2 receives the message containing customer Z1 and the timestamp again, it will update the timestamp bound to customer Z1 to keep the record always unique.

[0050] When camera 2 captures customer Z1, it generates a facial feature vector and performs facial matching based on the registered customer information sent by the management server. After a successful match, the recommendation information is not sent directly, but the time difference between the current time and the timestamp bound to customer Z1 is calculated. If the time difference is less than a certain threshold, such as 5 minutes, it will no longer be pushed to avoid frequent disturbing of members.

[0051] When the time difference exceeds this threshold, the sorted store information is pushed to customer Z1. At the same time, camera 2 also multicasts customer Z1 and the current timestamp (for example, November 19, 2024, 11:20) to the entire network, refreshes the timestamp of customer Z1 on other cameras (November 19, 2024, 11:20), and updates the local timestamp corresponding to customer Z1 (November 19, 2024, 11:20).

[0052] Embodiment 3: This embodiment aims at the situation where there are relatively concentrated cameras in one area. A customer may be captured by multiple cameras, resulting in repeated push of information, causing discomfort to the customer.

[0053] In this embodiment, the dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties further includes: After the first camera sends the customer queue information to the target customer's application, it sends the binding relationship between the target customer and the geographic location information of the first camera to other cameras; When the second camera recognizes the same target customer, it calculates whether the walking path between the second camera and the first camera is less than a preset distance threshold. If so, the customer queuing information is not sent to the target customer's application. Otherwise, the customer queuing information is sent to the target customer's application.

[0054] This embodiment implements processing to avoid repeated interruptions based on geographic differences. The camera contains an electronic map. When camera 1 captures customer Z1, it generates a facial feature vector, performs facial matching based on the registered customer information issued by the management server, and sends recommendation information to customer Z1 after a successful match. At this time, camera 1 needs to multicast customer Z1 to the entire network through the (*, G) multicast forwarding tree, and bring the geographic location information of camera 1, for example: longitude: 116.4074°E latitude: 39.9042°N. This message is called an "information push message", and camera 1 locally saves the binding relationship between customer Z1 and camera 1's geographic location information.

[0055] After receiving the information, other cameras, such as Camera 2, save the binding relationship between customer Z1 and the geographic location information of Camera 1 locally. When Camera 2 captures customer Z1 locally, it uses the map information contained in it to calculate the walking path difference between its current location and the geographic location information of Camera 1. If the walking path difference is less than a certain threshold, such as 100 meters, it will no longer push information to avoid frequent disturbance of members. When the walking path difference exceeds this threshold, the recommendation information will be pushed to the pedestrian, and the binding relationship between customer Z1 and the geographic location information of Camera 2 will be multicasted throughout the network, and the binding relationship between customer Z1 and the geographic location information of Camera 2 will be saved locally.

[0056] In the technical solution of this application, the camera gateway calculates the ranking index of the store's smart terminal relative to the camera, which is helpful to reduce the computing pressure of the camera; by allowing the camera to compare the faces of pedestrians and push information, it is helpful to avoid invalid information disturbing the management server and reduce the occupation of network bandwidth by invalid information. At the same time, the camera's face recognition is only performed locally, and facial features are not transmitted in the network, which can reduce the risk of privacy leakage. When multiple cameras capture the same customer, an "information push message" is sent, and whether to continue pushing information is determined based on the timestamp or walking path, avoiding frequent disturbances to customers.

[0057] Embodiment 4, as Figure 2 As shown, the present application also proposes a dynamic intelligent recommendation system based on changes in the characteristics of both supply and demand sides, including a management server, a store intelligent terminal and a camera, wherein: The management server sends the registered customer information to each camera and sends a unified multicast group address to the camera and the store smart terminal; When the store's smart terminals and cameras join the multicast group, they carry their own geographic location information, which is recorded by the smart terminal gateway and camera gateway respectively; The store smart terminal obtains the customer queue information of the store and sends it to its corresponding smart terminal gateway. The smart terminal gateway inserts the smart terminal's geographical location into the customer queue information and sends it to all camera gateways in the network; The camera gateway calculates the ranking index of the store smart terminal relative to the camera, adds it to the customer queue information of the store smart terminal and sends it to the camera; After the camera identifies the target customer, it finds the application information corresponding to the target customer and sends the customer queue information of all the store smart terminals recorded by the camera to the application corresponding to the target customer.

[0058] In a specific embodiment, the camera gateway calculates the ranking index of the smart terminal relative to the camera and performs the following operations: Set the initial value of the ranking index. When one group of customers is added to the smart terminal customer queue information, the ranking index decreases by 1, and when one group is reduced, the ranking index increases by 1. Before the first time, when the walking path between the smart terminal and the camera increases by a preset distance, the ranking index decreases by 1; After the first time, the ranking index increases by 1 for each additional preset distance between the walking path between the smart terminal and the camera.

[0059] In a specific embodiment, the dynamic intelligent recommendation system based on changes in characteristics of both supply and demand parties further includes: The first camera records the timestamp of sending the customer queue information to the target customer's application, and sends the timestamp of the target customer to other cameras; When the second camera recognizes the same target customer, it calculates the time difference between the current time and the timestamp of the target customer. If the time difference is less than the preset time threshold, the customer queue information is not sent to the target customer's application. Otherwise, the customer queue information is sent to the target customer's application.

[0060] In a specific embodiment, the dynamic intelligent recommendation system based on changes in characteristics of both supply and demand parties further includes: After the first camera sends the customer queue information to the target customer's application, it sends the binding relationship between the target customer and the geographic location information of the first camera to other cameras; When the second camera recognizes the same target customer, it calculates whether the walking path between the second camera and the first camera is less than a preset distance threshold. If so, the customer queuing information is not sent to the target customer's application. Otherwise, the customer queuing information is sent to the target customer's application.

[0061] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A dynamic intelligent recommendation method based on changes in characteristics of both supply and demand sides, characterized in that: The dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties includes: The management server sends the registered customer information to each camera and sends a unified multicast group address to the camera and the store smart terminal; When the store's smart terminals and cameras join the multicast group, they carry their own geographic location information, which is recorded by the smart terminal gateway and camera gateway respectively; The store smart terminal obtains the customer queue information of the store and sends it to its corresponding smart terminal gateway. The smart terminal gateway inserts the smart terminal's geographical location into the customer queue information and sends it to all camera gateways in the network; The camera gateway calculates the ranking index of the store smart terminal relative to the camera, adds it to the customer queue information of the store smart terminal and sends it to the camera; After the camera identifies the target customer, it finds the application information corresponding to the target customer and sends the customer queue information of all the store smart terminals recorded by the camera to the application corresponding to the target customer.

2. The dynamic intelligent recommendation method based on the change of characteristics of both supply and demand sides according to claim 1 is characterized in that: The calculating the ranking index of the intelligent terminal relative to the camera includes: Set the initial value of the ranking index. When one group of customers is added to the smart terminal customer queue information, the ranking index decreases by 1, and when one group is reduced, the ranking index increases by 1. Before the first time, when the walking path between the smart terminal and the camera increases by a preset distance, the ranking index decreases by 1; After the first time, the ranking index increases by 1 for each additional preset distance between the walking path between the smart terminal and the camera.

3. The dynamic intelligent recommendation method based on the change of characteristics of both supply and demand sides according to claim 1 is characterized in that: The dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties also includes: The first camera records the timestamp of sending the customer queue information to the target customer's application, and sends the timestamp of the target customer to other cameras; When the second camera recognizes the same target customer, it calculates the time difference between the current time and the timestamp of the target customer. If the time difference is less than the preset time threshold, the customer queue information is not sent to the target customer's application. Otherwise, the customer queue information is sent to the target customer's application.

4. The dynamic intelligent recommendation method based on the change of characteristics of both supply and demand sides according to claim 1 is characterized in that: The dynamic intelligent recommendation method based on changes in characteristics of both supply and demand parties also includes: After the first camera sends the customer queue information to the target customer's application, it sends the binding relationship between the target customer and the geographic location information of the first camera to other cameras; When the second camera recognizes the same target customer, it calculates whether the walking path between the second camera and the first camera is less than a preset distance threshold. If so, the customer queuing information is not sent to the target customer's application; otherwise, the customer queuing information is sent to the target customer's application.

5. A dynamic intelligent recommendation system based on changes in characteristics of both supply and demand sides, characterized in that: The dynamic intelligent recommendation system based on the changes in the characteristics of both supply and demand parties includes a management server, a store intelligent terminal and a camera, wherein: The management server sends the registered customer information to each camera and sends a unified multicast group address to the camera and the store smart terminal; When the store's smart terminals and cameras join the multicast group, they carry their own geographic location information, which is recorded by the smart terminal gateway and camera gateway respectively; The store smart terminal obtains the customer queue information of the store and sends it to its corresponding smart terminal gateway. The smart terminal gateway inserts the smart terminal's geographical location into the customer queue information and sends it to all camera gateways in the network; The camera gateway calculates the ranking index of the store smart terminal relative to the camera, adds it to the customer queue information of the store smart terminal and sends it to the camera; After the camera identifies the target customer, it finds the application information corresponding to the target customer and sends the customer queue information of all the store smart terminals recorded by the camera to the application corresponding to the target customer.

6. The dynamic intelligent recommendation system based on changes in characteristics of both supply and demand sides according to claim 5, characterized in that: The camera gateway calculates the ranking index of the smart terminal relative to the camera and performs the following operations: Set the initial value of the ranking index. When one group of customers is added to the smart terminal customer queue information, the ranking index decreases by 1, and when one group is reduced, the ranking index increases by 1. Before the first time, when the walking path between the smart terminal and the camera increases by a preset distance, the ranking index decreases by 1; After the first time, the ranking index increases by 1 for each additional preset distance between the walking path between the smart terminal and the camera.

7. The dynamic intelligent recommendation system based on changes in characteristics of both supply and demand sides according to claim 5, characterized in that: The dynamic intelligent recommendation system based on changes in characteristics of both supply and demand parties also includes: The first camera records the timestamp of sending the customer queue information to the target customer's application, and sends the timestamp of the target customer to other cameras; When the second camera recognizes the same target customer, it calculates the time difference between the current time and the timestamp of the target customer. If the time difference is less than the preset time threshold, the customer queue information is not sent to the target customer's application. Otherwise, the customer queue information is sent to the target customer's application.

8. The dynamic intelligent recommendation system based on changes in characteristics of both supply and demand sides according to claim 5, characterized in that: The dynamic intelligent recommendation system based on changes in characteristics of both supply and demand parties also includes: After the first camera sends the customer queue information to the target customer's application, it sends the binding relationship between the target customer and the geographic location information of the first camera to other cameras; When the second camera recognizes the same target customer, it calculates whether the walking path between the second camera and the first camera is less than a preset distance threshold. If so, the customer queuing information is not sent to the target customer's application; otherwise, the customer queuing information is sent to the target customer's application.

Images