Marketing advertisement recommendation method based on artificial intelligence
By connecting to the property access control system and elevator monitoring system, and combining advertising database information to personalize advertising recommendations, the problem that traditional elevator advertising systems cannot be dynamically adjusted is solved, accurate identification and personalized recommendations are achieved, and advertising effectiveness and reach rate are improved. .
Patent Information
- Application Number
- CN202411232262.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-12-13
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The traditional elevator advertising system adopts a fixed playback mode and cannot dynamically adjust according to the actual situation of the personnel in the elevator, resulting in the inability to match the actual needs of the target personnel and the inability to achieve accurate identification and personalized recommendation.
Access to the property access control system to obtain facial images of registered personnel, combine with the elevator monitoring system to collect personnel information, detect the number of target personnel and eye orientation in real time, personalize advertising recommendations based on the information in the advertising database, and dynamic rotations are performed based on the advertising weights broadcast.
It improves the pertinence and attractiveness of advertising, improves advertising effectiveness, ensures effective utilization of advertising resources, improves advertising reach and return on investment, and reduces costs.
Smart Images
Figure CN119130553A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of advertisement recommendation, and in particular to a marketing advertisement recommendation method based on artificial intelligence. Background Art
[0002] Elevator carousel advertising is a form of advertising that is played in a loop on the electronic screen in the elevator. It is a field-based marketing method that aims to use the enclosed space and forced viewing environment in the elevator to strengthen the impact of brand information on the audience in a short period of time. Elevator carousel advertising usually includes multiple short advertisements that are played on the screen in a rotating manner to ensure that the information of different advertisers can be equally displayed to the audience in the elevator.
[0003] Traditional marketing advertising recommendation methods often fail to accurately locate the target population, resulting in poor advertising results. At the same time, since the advertising broadcast time in the elevator car is limited, how to attract the attention of the target personnel in a short time and increase the viewing rate of the advertisement has become an urgent problem to be solved.
[0004] Most existing elevator advertising systems use a fixed playback mode, that is, they play according to a preset advertising playlist and cannot be dynamically adjusted according to the actual situation in the elevator. In this mode, the playback time and content of the advertisement cannot match the actual needs of the target personnel, resulting in a significant reduction in the advertising effect. It is impossible to accurately identify the target personnel and make personalized recommendations; it is impossible to make personalized advertising recommendations based on the target personnel's own intentions. Summary of the invention
[0005] The purpose of the present invention is to provide a marketing advertisement recommendation method based on artificial intelligence to solve the following technical problems:
[0006] Most existing elevator advertising systems use a fixed playback mode and cannot be dynamically adjusted according to the actual situation of people in the elevator. They cannot match the actual needs of the target personnel and cannot achieve accurate identification and personalized recommendations for the target personnel.
[0007] The purpose of the present invention can be achieved through the following technical solutions:
[0008] A marketing advertisement recommendation method based on artificial intelligence comprises the following steps:
[0009] Access the property access control system to obtain the facial image of the registered person, collect the image of the person in the elevator car through the elevator monitoring system, identify the registered person in the car, count the number of times each registered person takes the elevator and the floor number i, and mark the registered person who has taken the elevator more than the preset number of times as the target person;
[0010] Detect the number of target personnel in the car in real time. When the number of target personnel in the car is 1, obtain all the floor numbers selected by the current elevator button panel, determine the number of elevator stops before reaching the floor number i where the target personnel is located, and calculate the elevator running time t to the floor where the target personnel is located based on the number i, the elevator running speed and the number of stops;
[0011] According to the gender and age of the target person, select advertisements with a playing time shorter than t from the corresponding elevator advertisement library for rotation, and detect the eye direction of the target person in real time. When the target person's eye is directed toward the advertisement display screen for more than a preset time, mark the advertisement played at this time as an intended advertisement and store it in the personal advertisement library of the target person;
[0012] When the target person leaves the elevator car, if the playing time of the advertisement at this time exceeds 30%, the advertisement will be marked as a pending advertisement and stored in the target person's personal advertisement library. When the target person takes the elevator next time, the pending advertisement will be played again from the beginning. After the pending advertisement is finished, the intended advertisement will be played in rotation.
[0013] As a further solution of the present invention: when the number of target persons in the elevator car is greater than 1, the number of intended advertisements in the personal advertising library of each target person and the floor number of each target person are counted, the elevator running time is calculated according to all the floor numbers selected by the elevator button panel, and the elevator running time is divided into several time periods. In each time period, the co-existing target persons in the current time period are counted, and the ratio of the number of intended advertisements for each target person to the total amount of intended advertisements for co-existing target persons in the current time period is marked as the advertising weight of each target person, and the intended advertisements of the co-existing target persons are randomly rotated according to the advertising weights.
[0014] As a further solution of the present invention: when the weight of any co-existing target person is greater than 60%, the advertising display screen first plays the pending advertisement of the co-existing target person. If the weight of each co-existing person is lower than 60%, the intended advertisements of all target persons are randomly rotated according to the advertisement weights.
[0015] As a further solution of the present invention, the specific steps of randomly rotating the intention advertisements of all target persons according to the advertisement weights are as follows:
[0016] Obtain the advertising weights a1, a2, ..., an, n of the co-existing target persons in sequence, obtain the duration T of the current time period, allocate the initial advertising playback time of each person as t1, t2, ..., tn according to the advertising weights, select several advertisements with the highest viewing rates in the advertising library of each co-existing target person, mark them as priority advertisements, start with the co-existing target person with the highest weight, count the priority advertisements whose playback duration t1' is closest to t1 in the corresponding advertising library, and store them in the random rotation sequence of the current time period, then reallocate the remaining duration T-t1' according to the weights of the remaining persons, repeat the above steps until the total duration of the priority advertisements is greater than the duration T of the current time period, and then stop allocating.
[0017] As a further solution of the present invention: the viewing rate is the ratio of the time that the target person watches any advertisement to the total time of the advertisement.
[0018] As a further solution of the present invention: the positions of the facial features of the target person are detected based on the Haar feature classifier in the OpenCV library, and the positions of the eyeballs are identified, and the eye direction of the target person is determined according to the rotation state of the eyeballs.
[0019] As a further solution of the present invention: when there is no target person in the elevator car, advertisements in the elevator advertisement library are automatically and randomly played.
[0020] Beneficial effects of the present invention:
[0021] (1) The present invention selects advertisements with a playing time shorter than the elevator running time from the advertisement library for rotation based on the facial image of the registered person in the community, the number of times he / she has taken the elevator, the floor number, as well as the gender and age information; when the target person's eyes are directed toward the advertisement display screen for more than a preset time, the system will mark the advertisement as an intended advertisement and store it in the personal advertisement library; this personalized advertisement playing method can improve the pertinence and attractiveness of advertisements and enhance the advertising effect;
[0022] (2) The present invention detects the number of target persons in the elevator in real time, and adjusts the intended advertisement playback of the coexisting target persons according to the number. When the number of target persons is greater than 1, the number of intended advertisements in the personal advertisement library of each target person is counted, and the advertisement weight is allocated according to the number of intended advertisements. The advertisements to be played of the target persons with higher advertisement weights are played first, and the advertisement playback weight is allocated according to the elevator running time. Such real-time monitoring and dynamic adjustment can ensure the effective use of advertising resources and improve the reach rate of advertisements; it can also improve the return on investment of advertisements and reduce costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] The present invention will be further described below in conjunction with the accompanying drawings.
[0024] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION
[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0026] See also Figure 1 As shown, the present invention is a marketing advertisement recommendation method based on artificial intelligence, comprising the following steps:
[0027] Access the property access control system to obtain the facial image of the registered person, collect the image of the person in the elevator car through the elevator monitoring system, identify the registered person in the car, count the number of times each registered person takes the elevator and the floor number i, and mark the registered person who has taken the elevator more than the preset number of times as the target person;
[0028] Detect the number of target personnel in the car in real time. When the number of target personnel in the car is 1, obtain all the floor numbers selected by the current elevator button panel, determine the number of elevator stops before reaching the floor number i where the target personnel is located, and calculate the elevator running time t to the floor where the target personnel is located based on the number i, the elevator running speed and the number of stops;
[0029] According to the gender and age of the target person, select advertisements with a playing time shorter than t from the corresponding elevator advertisement library for rotation, and detect the eye direction of the target person in real time. When the target person's eye is directed toward the advertisement display screen for more than a preset time, mark the advertisement played at this time as an intended advertisement and store it in the personal advertisement library of the target person;
[0030] When the target person leaves the elevator car, if the playing time of the advertisement at this time exceeds 30%, the advertisement will be marked as a pending advertisement and stored in the target person's personal advertisement library. When the target person takes the elevator next time, the pending advertisement will be played again from the beginning. After the pending advertisement is finished, the intended advertisement will be played in rotation.
[0031] The present invention can select advertisements with a playing time shorter than the elevator running time from the advertisement library for rotation according to the facial image of the registered person, the number of times the elevator has been taken, the floor number, as well as the gender and age information. When the target person's eyes are directed toward the advertisement display screen for more than a preset time, the system will mark the advertisement as an intended advertisement and store it in the personal advertisement library. This personalized advertisement playing method can improve the pertinence and attractiveness of advertisements and enhance the advertising effect.
[0032] In a preferred embodiment of the present invention, when the number of target persons in the elevator car is greater than 1, the number of intended advertisements in the personal advertising library of each target person and the floor number of each target person are counted, the elevator running time is calculated according to all the floor numbers selected by the elevator button panel, and the elevator running time is divided into several time periods. In each time period, the coexisting target persons in the current time period are counted, and the ratio of the number of intended advertisements for each target person to the total amount of intended advertisements for coexisting target persons in the current time period is marked as the advertising weight of each target person, and the intended advertisements of the coexisting target persons are randomly rotated according to the advertising weights.
[0033] The present invention detects the number of target persons in the car in real time, and adjusts the intended advertisement playback of the coexisting target persons according to the number. When the number of target persons is greater than 1, the number of intended advertisements in the personal advertisement library of each target person is counted, and the advertisement playback weight is allocated according to the elevator running time. This real-time monitoring and dynamic adjustment can ensure the effective use of advertising resources and improve the reach rate of advertisements.
[0034] In a preferred case of this embodiment, when the weight of any co-existing target person is greater than 60%, the advertising display screen first plays the pending advertisement of the co-existing target person. If the weight of each co-existing person is lower than 60%, the intended advertisements of all target persons are randomly rotated according to the advertisement weights.
[0035] When the weight of any co-existing target person is greater than 60%, the present invention gives priority to playing the advertisement to be played by the person. If the weight of each co-existing person is less than 60%, the intended advertisements of all target persons are randomly rotated according to the advertisement weight. This precise delivery and resource-saving method can improve the return on investment of advertising and reduce costs.
[0036] In another preferred embodiment of the present invention, the specific steps of randomly rotating all the intended advertisements of the target persons according to the advertisement weights are as follows:
[0037] Obtain the advertising weights a1, a2, ..., an, n of the co-existing target persons in sequence, obtain the duration T of the current time period, allocate the initial advertising playback time of each person as t1, t2, ..., tn according to the advertising weights, select several advertisements with the highest viewing rates in the advertising library of each co-existing target person, mark them as priority advertisements, start with the co-existing target person with the highest weight, count the priority advertisements whose playback duration t1' is closest to t1 in the corresponding advertising library, and store them in the random rotation sequence of the current time period, then reallocate the remaining duration T-t1' according to the weights of the remaining persons, repeat the above steps until the total duration of the priority advertisements is greater than the duration T of the current time period, and then stop allocating.
[0038] The present invention can improve the overall viewing rate of advertisements and the attention of users by selecting the advertisements with the highest viewing rate for playback, thereby improving the effect of advertisements. The advertisement playback time is allocated according to the advertisement weights of the co-existing target persons, and when the advertisement duration is limited, the viewing experience of the target persons with high weights is prioritized.
[0039] Allocating broadcast time according to advertisement weight not only reflects the fairness of advertisement delivery, but also ensures the effective use of advertisement resources and improves the efficiency of the entire advertisement system.
[0040] It is worth noting that the viewing rate is the ratio of the time the target person watches any advertisement to the total time of the advertisement.
[0041] In another preferred embodiment of the present invention, the positions of the facial features of the target person are detected based on the Haar feature classifier in the OpenCV library, and the positions of the eyeballs are identified, and the eye direction of the target person is determined according to the rotation state of the eyeballs.
[0042] The present invention detects the facial features of the target person, identifies the eyeball position, and determines the eye direction of the target person according to the rotation state of the eyeball. This user behavior analysis can more accurately determine the user's interest in the advertisement and provide more accurate data support for advertisers.
[0043] In another preferred embodiment of the present invention, when there is no target person in the elevator car, advertisements in the elevator advertisement library are automatically and randomly played.
[0044] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.
Claims
1. A marketing advertisement recommendation method based on artificial intelligence, characterized in that: The following steps are involved: Access the property access control system to obtain the facial image of the registered person, collect the image of the person in the elevator car through the elevator monitoring system, identify the registered person in the car, count the number of times each registered person takes the elevator and the floor number i, and mark the registered person who has taken the elevator more than the preset number of times as the target person; Detect the number of target personnel in the car in real time. When the number of target personnel in the car is 1, obtain all the floor numbers selected by the current elevator button panel, determine the number of elevator stops before reaching the floor number i where the target personnel is located, and calculate the elevator running time t to the floor where the target personnel is located based on the number i, the elevator running speed and the number of stops; According to the gender and age of the target person, select advertisements with a playing time shorter than t from the corresponding elevator advertisement library for rotation, and detect the eye direction of the target person in real time. When the target person's eye is directed toward the advertisement display screen for more than a preset time, mark the advertisement played at this time as an intended advertisement and store it in the personal advertisement library of the target person; When the target person leaves the elevator car, if the playing time of the advertisement at this time exceeds 30%, the advertisement will be marked as a pending advertisement and stored in the target person's personal advertisement library. When the target person takes the elevator next time, the pending advertisement will be played again from the beginning. After the pending advertisement is finished, the intended advertisement will be played in rotation.
2. The marketing advertisement recommendation method based on artificial intelligence according to claim 1, characterized in that: When the number of target persons in the elevator car is greater than 1, the number of intended advertisements in the personal advertising library of each target person and the floor number of each target person are counted, and the elevator running time is calculated according to all the floor numbers selected by the elevator button panel. The elevator running time is divided into several time periods. In each time period, the co-existing target persons in the current time period are counted, and the ratio of the number of intended advertisements for each target person to the total number of intended advertisements for co-existing target persons in the current time period is marked as the advertising weight of each target person, and the intended advertisements of the co-existing target persons are randomly rotated according to the advertising weights.
3. The marketing advertisement recommendation method based on artificial intelligence according to claim 2, characterized in that: When the weight of any co-existing target person is greater than 60%, the advertising display screen will first play the pending advertisement of the co-existing target person. If the weight of each co-existing person is less than 60%, the intended advertisements of all target persons will be randomly rotated according to the advertisement weights.
4. The marketing advertisement recommendation method based on artificial intelligence according to claim 3, characterized in that: The specific steps to randomly rotate the intent ads of all target users according to the ad weights are as follows: Obtain the advertising weights a1, a2, ..., an, n of the co-existing target persons in sequence, obtain the duration T of the current time period, allocate the initial advertising playback time of each person as t1, t2, ..., tn according to the advertising weights, select several advertisements with the highest viewing rates in the advertising library of each co-existing target person, mark them as priority advertisements, start with the co-existing target person with the highest weight, count the priority advertisements whose playback duration t1' is closest to t1 in the corresponding advertising library, and store them in the random rotation sequence of the current time period, then reallocate the remaining duration T-t1' according to the weights of the remaining persons, repeat the above steps until the total duration of the priority advertisements is greater than the duration T of the current time period, and then stop allocating.
5. The marketing advertisement recommendation method based on artificial intelligence according to claim 4, characterized in that: The viewing rate is the ratio of the time that the target person watches any advertisement to the total time of the advertisement.
6. The marketing advertisement recommendation method based on artificial intelligence according to claim 1, characterized in that: Based on the Haar feature classifier in the OpenCV library, the facial features of the target person are detected, the eye position is identified, and the eye direction of the target person is determined according to the rotation state of the eye.
7. The marketing advertisement recommendation method based on artificial intelligence according to claim 1, characterized in that: When there is no target person in the elevator car, advertisements in the elevator advertisement library will be automatically played randomly.