Internet-based advertisement putting management system and method

Abstract

The invention discloses an internet-based advertisement putting management system and method, and relates to the technical field of advertisement putting. The system comprises an environment monitoring module, an advertisement management design module, a scene adjustment module and a dynamic adaptation module which run in sequence. The technical key points are that the operation of the management system can help to identify children in real time and switch audio modes, children protection and business demands are considered, and the advertisement input cost is reduced to a certain extent; the volume is adjusted in real time based on the environment noise and the crowd sound absorption effect, the advertisement volume definition is improved, a correction sub-strategy for the volume sensitivity degree is triggered according to the crowd density, the volume change smoothness is improved when the crowd density exceeds the critical threshold value, discomfort caused by sudden volume is avoided, and the system operation emphasizes clear transmission and auditory comfort.

Description

Technical Field

[0001] The present invention relates to the technical field of advertising delivery, and specifically provides an Internet-based advertising delivery management system and method. Background Art

[0002] Advertising delivery is a marketing behavior in which advertisers convey brand information to target audiences through media. Its core lies in precise reach and high-efficiency conversion. In the new media era, advertising delivery relies on digital channels such as social media, short video platforms, and KOL collaborations, combined with data algorithms to analyze user behaviors (such as browsing habits and interest tags) to achieve personalized content push. The design of advertising content focuses on dynamic creativity: generating customized visuals (such as replacing products / scenes with dynamic templates) and interactive elements (such as AR try-on and click voting) through user portraits, and integrating emotional narratives (story-based short videos and user UGC secondary creation). Data-driven content optimization runs through the whole process, using A / B testing to adjust material versions in real time and iterating on creativity based on click-through rate and conversion rate feedback. New media advertising delivery breaks through traditional one-way communication, amplifying the dissemination power of high-quality content through algorithms to form a "content-data-conversion" closed loop.

[0003] Currently, advertising delivery mostly relies on advertising terminals (such as LED displays) for dissemination.

[0004] On the one hand, traditional advertising terminals are prone to problems of weak environmental adaptation ability and insufficient differentiation of user groups in dynamic closed scenarios. For example, elevator advertising media terminals are in continuous advertising delivery with a single delivery mode. If an infant (child) enters the elevator, this special group of infants is easily attracted by the advertising sound and light, and close viewing may affect their vision development.

[0005] On the other hand, the static volume setting of traditional advertising terminals can lead to problems of fragmented experience (unable to hear clearly or noise interference). For example, the playback volume of elevator advertising media terminals is fixed. If the elevator is crowded with people, the played sound will be greatly covered, affecting the advertising delivery effect. Additionally, even with dynamic setting (dynamically adjusting the playback volume according to the ambient volume), there will still be a problem that people close to the advertising media terminal are easily affected by sudden volume due to the narrow elevator space. Summary of the Invention

[0006] (1) Technical Problems to be Solved

[0007] Aiming at the deficiencies of the prior art, the present invention provides an Internet-based advertising delivery management system and method, and the operation of this system effectively solves the problems raised in the background art.

[0008] (2) Technical Solutions

[0009] To achieve the above objectives, the present invention is realized through the following technical solutions:

[0010] An advertising placement management system based on the Internet, the system includes:

[0011] An environmental monitoring module, which captures video streams in the target scenario, identifies whether there are children in the target scenario, and analyzes and obtains environmental data; wherein, the environmental data at least includes crowd density and environmental noise;

[0012] An advertising management design module, when the condition of identifying children is met, the advertising terminal in the target scenario is switched to the audio mode; otherwise, the advertising terminal is switched to the normal mode;

[0013] A scene adjustment module, when in the normal mode, extracts and judges according to the crowd density to determine whether it is a multi-person scene or a single-person scene; when it is determined to be a multi-person scene, the normal mode runs with a vertical multi-line layout combined with a parallel scrolling carousel mechanism; when it is determined to be a single-person scene, the normal mode runs with a full-screen free layout combined with a queue-style sequential carousel mechanism;

[0014] A dynamic adaptation module, under the condition of executing the parallel scrolling carousel mechanism, triggers an environmental data call instruction, executes a volume dynamic setting strategy, and adjusts the playback volume of the advertising terminal in real time; wherein, the volume dynamic setting strategy has a built-in correction sub-strategy to correct the relationship between the playback volume and the environmental noise according to the constraint conditions.

[0015] Further, the process of analyzing and obtaining the crowd density is as follows:

[0016] Real-time detection and statistics of human targets in the target scenario through a lightweight CNN model;

[0017] The lightweight CNN model is trained with a crowd counting data set;

[0018] After outputting the number of people, combined with the floor area of the target scenario, calculate the crowd density D = number of people / floor area of the target scenario.

[0019] Further, the process of analyzing and obtaining the environmental noise is as follows:

[0020] Signal processing: Real-time capture of the original sound pressure signal in the target scenario and A-weighting filtering;

[0021] Real-time calculation: Obtain the instantaneous sound pressure at equal intervals and output the environmental noise based on the sound pressure formula;

[0022] Calibration output: After triggering the calibration signal, the environmental noise is transmitted in real time through the Modbus protocol.

[0023] Further, the audio mode means that the advertising terminal only enables audio output, and the normal mode means that the advertising terminal enables audio-visual advertising.

[0024] Furthermore, the process of determining whether it is a multi-person scenario or a single-person scenario is as follows:

[0025] Based on the crowd density, the number of people is obtained. When the number of people = 1, it is determined as a single-person scenario;

[0026] When the number of people ≠ 1, it is determined as a multi-person scenario.

[0027] Furthermore, the content of the normal mode running in a multi-person scenario is as follows:

[0028] Vertical multi-line layout: The screen of the advertising terminal is vertically divided into several lines, and each line corresponds to different types of advertising content;

[0029] Parallel scrolling carousel mechanism: Each line independently executes an upward scrolling animation, and the scrolling speed is determined according to the occupancy duration of the top advertising space; among them, the top advertising space enables audible and visual advertisements, and the other advertising spaces except the top one are in a silent display state;

[0030] The content of the normal mode running in a single-person scenario is as follows:

[0031] Full-screen free layout: The advertising terminal displays a single advertising content full-screen and adds interactive buttons;

[0032] Queue-style sequential carousel mechanism: After the advertising content on the current page is played, it switches to the next advertising content, and so on in a loop.

[0033] Furthermore, the triggered environmental data call instruction is: Call the crowd density and environmental noise in the environmental data, and preprocess the environmental data; among them, the preprocessing includes at least data cleaning.

[0034] Furthermore, in the executed volume dynamic setting strategy:

[0035] Based on the preprocessed environmental data, the volume is restricted and adjusted in the following way:

[0036] V adjusted = min(V base + k * Ro + S * (Nr - T), V max );

[0037] Among them, V adjusted is the adjusted advertising volume, that is, the playing volume, V base is the base volume, k is the crowd sound absorption effect compensation coefficient, Ro is the number of people in the crowd density, S is the volume adjustment sensitivity, Nr is the environmental noise, T is the noise level threshold, and V max is the maximum volume.

[0038] Further, the content of the correction sub-strategy built into the volume dynamic setting strategy is as follows:

[0039] Perform correction processing on the volume adjustment sensitivity S used in the volume dynamic setting strategy, determine whether the population density in the constraint conditions exceeds the preset critical threshold, and obtain the corrected volume adjustment sensitivity S based on the established correction function model - , and replace the original volume adjustment sensitivity S with the corrected volume adjustment sensitivity S - ;

[0040] When running the correction function model, the function definition is as follows:

[0041]

[0042] Among them, S max is the maximum value of the volume adjustment sensitivity S, both α and β are constants, and D_th is the critical threshold.

[0043] An Internet-based advertising placement management method includes the following steps:

[0044] Capture the video stream in the target scene, identify whether there are children in the target scene, and analyze and obtain environmental data; among them, the environmental data at least includes population density and environmental noise;

[0045] When the condition of identifying children is met, switch the advertising terminal in the target scene to the audio mode; otherwise, switch the advertising terminal to the normal mode;

[0046] When in the normal mode, extract and judge according to the population density to determine whether it is a multi-person scene or a single-person scene; when it is determined to be a multi-person scene, the normal mode runs with a vertical multi-line layout combined with a parallel scrolling carousel mechanism; when it is determined to be a single-person scene, the normal mode runs with a full-screen free layout combined with a queue-based sequential carousel mechanism;

[0047] Under the condition of executing the parallel scrolling carousel mechanism, trigger the environmental data call instruction, execute the volume dynamic setting strategy, and adjust the playback volume of the advertising terminal in real time; among them, the volume dynamic setting strategy has a built-in correction sub-strategy to correct the relationship between the playback volume and the environmental noise according to the constraint conditions.

[0048] (III) Beneficial effects

[0049] The present invention provides an Internet-based advertising placement management system and method, which have the following beneficial effects:

[0050] This solution addresses the problems of weak environmental adaptation ability and insufficient user group differentiation of traditional advertising terminals in dynamic closed scenarios through a multi-modal environmental perception and real-time computing architecture. Based on CNN-based child recognition and crowd density calculation, it realizes dynamic and precise switching of advertising modes; for noise decibel monitoring combined with A-weighting filtering, it reduces mechanical low-frequency interference to a certain extent and provides a basis for adaptive adjustment of advertising volume;

[0051] This solution achieves a dynamic balance between information transmission efficiency and user experience through a crowd density-driven adaptive strategy, solves the problem of rigid placement of traditional advertising terminals, real-time identifies children and switches audio modes, takes into account both child protection and commercial needs, and also reduces advertising investment costs to a certain extent; the information reach efficiency in multi-person scenarios is improved, and the interaction conversion rate in single-person scenarios is increased, realizing the efficient use of resources and enhancing the rationality of advertising content design and placement;

[0052] This solution addresses the problem of experience fragmentation (unclear hearing or noise interference) caused by static volume settings of traditional advertising terminals through environmental-crowd two-dimensional regulation. Based on environmental noise and crowd sound absorption effects, it adjusts the volume in real time, improving the clarity of advertising volume. According to the crowd density, it triggers a correction sub-strategy for volume sensitivity. When the crowd density exceeds the critical threshold, the smoothness of volume change is improved, avoiding discomfort caused by sudden volume (reducing the complaint rate), and the system operation focuses on both clear communication and auditory comfort. Brief Description of the Drawings

[0053] Figure 1 It is a modular operation schematic diagram of an Internet-based advertising placement management system in the present invention;

[0054] Figure 2 It is an overall process schematic diagram of an Internet-based advertising placement management method in the present invention. Detailed Embodiments

[0055] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0056] Embodiment 1:

[0057] Please refer to Figure 1, this embodiment provides an Internet-based advertising placement management system. The design of this system aims to create an intelligent and flexible Internet advertising placement management system that focuses on dynamically adjusting advertising placement according to environmental conditions (such as whether there are children or infants present) in specific scenarios (such as in an elevator). This system will combine computer vision technology, real-time data analysis, and intelligent advertising strategy design to achieve precise and user-friendly advertising placement. Especially when children are present, only voice advertisements will be played without turning on video advertisements to protect children's eyesight;

[0058] System modular design:

[0059] 1. Environmental monitoring module

[0060] Capture the video stream in the target scenario, identify whether there are children in the target scenario, and analyze to obtain environmental data. Among them, the environmental data at least includes crowd density and environmental noise;

[0061] Among them, the target scenario is the area where the advertising terminal is located. For example, if the advertising terminal is in an elevator, the target scenario is the area inside the elevator; if the advertising terminal is in a media screening hall, the target scenario is the promotional radiation area (front hall area) of the advertising terminal;

[0062] For this embodiment, the target scenario is mostly set as a closed space such as an elevator, specifically: inside the elevator;

[0063] The video stream capture is carried out using the camera in the elevator. Whether there are children in the elevator (with child characteristics, such as height, clothing, and age, etc.) is identified through computer vision algorithms. Technical implementation: Use a deep learning model (such as YOLOv5 or SSD) for object detection, and train a dedicated model to identify child characteristics. The model output is a binary signal (0 means no children, 1 means there are children). This model will be trained based on a large dataset containing images of children and adults to learn the unique characteristics of children, such as height ratio, clothing style (such as children's clothing, cartoon patterns, etc.), and the age range that can be indirectly inferred through facial features. When the model is detecting, it will extract these key features in the image and comprehensively judge whether there are children. If child characteristics are detected, it will output 1; if not detected, it will output 0;

[0064] YOLOv5 and SSD are suitable for this task because of their efficient multi-scale detection ability and real-time performance, and can quickly and accurately identify children in images or videos;

[0065] Based on the video stream captured by the camera, through real-time analysis by a convolutional neural network (CNN), the crowd density: the number of people in the current target scene, and the crowd density D = number of people / floor area of the target scene; specifically, a lightweight CNN model (such as MobileNet or a YOLO-derived architecture) is used to detect and count human targets in the elevator in real time; the model is trained with a crowd counting dataset (such as ShanghaiTech), and after outputting the number of people, combined with the floor area of the elevator (a known fixed value), the crowd density D = number of people / floor area of the target scene is calculated; the real-time performance is ensured through model compression (quantization, pruning) and edge computing devices (such as Jetson Nano).

[0066] For environmental noise, it relies on the microphone array built into the advertising terminal to collect the audio stream, and is achieved through the following steps:

[0067] Definition of environmental noise: Directly measure the instantaneous sound pressure level (unit: dB) in the elevator, which reflects the superimposed volume of all sound sources (human voices, equipment sounds, etc.) at the current moment, not the average value.

[0068] Based on the sound pressure formula: Lp = 20 * log 10 (P / P0) (P0 = 20 μPa), calculate the decibel value of the instantaneous sound pressure level (i.e., environmental noise), where Lp: decibel value (dB); P: instantaneous sound pressure captured by the microphone array (Pascal, Pa);

[0069] P0: reference sound pressure (20 μPa, the minimum audible sound pressure of the human ear); this formula converts the physical quantity of sound pressure into a logarithmic decibel value, which conforms to the non-linear perception characteristics of the human ear to sound intensity.

[0070] The real-time measurement situation is as follows:

[0071] Hardware: Install a microphone array in the elevator (using an IEC61672 standard sound level meter), supporting A-weighting (simulating the frequency response of the human ear) and F mode (125 ms fast response); Signal processing: Capture the original sound pressure signal in real time at a sampling rate of 48 kHz; Only use A-weighting filtering to suppress low-frequency interference from the elevator motor (<100 Hz); Real-time calculation: Calculate the instantaneous sound pressure P every 1 ms and substitute it into the formula to output Lp; Without time integration (such as Leq), directly output the current decibel value; Calibration: Automatically trigger a 94 dB / 1 kHz calibration signal every day to ensure the error < ±1 dB; Output: The decibel value is transmitted in real time through the Modbus protocol, with a refresh rate of 1 kHz, and can monitor volume fluctuations at the millisecond level (such as sudden screams).

[0072] The Internet is mainly reflected in data interaction and remote collaboration in the solution:

[0073] Cloud model training and update: The child recognition and crowd density calculation models rely on large-scale cloud datasets (such as ShanghaiTech) for training, and the optimized model parameters are pushed to edge devices (such as Jetson Nano) through the network for iterative upgrades; Dynamic adjustment of advertising strategies: Environmental data (such as the presence of children and crowd density) needs to be transmitted to the advertising server in real time, triggering the cloud decision engine to adjust the advertising content (such as blocking inappropriate content for children or dynamic volume control), realizing a closed-loop of "environment-cloud-terminal"; Remote device management: The calibration signal (94dB / 1kHz) of the microphone array and the status monitoring of the advertising terminal can be connected to the cloud platform through Internet of Things protocols (such as Modbus to MQTT), supporting remote operation and maintenance and data analysis;

[0074] This solution solves the problems of weak environmental adaptation ability and insufficient user group differentiation of traditional advertising terminals in dynamic closed scenarios (such as elevators) through multi-modal environmental perception (video + audio) and real-time computing architecture. Based on lightweight CNN for child recognition and crowd density calculation, it realizes dynamic and precise switching of advertising modes (such as blocking visual advertisements when children are present); Real-time noise response: Millisecond-level decibel monitoring (refresh rate 1kHz) combined with A-weighting filtering to avoid mechanical low-frequency interference, providing a basis for adaptive adjustment of advertising volume.

[0075] 2. Advertising management design module

[0076] When children are recognized, the advertising terminal in the target scenario is switched to audio mode;

[0077] When children are not recognized, the advertising terminal is switched to the normal mode;

[0078] Among them, conditional logic (if-else or switch-case) is used to select the playback content according to the child recognition result; If children are recognized, voice advertisements are selected (i.e., the corresponding audio mode, for example: only the advertising sound is played on the advertising terminal, without displaying video or text); If there are no children, enter the normal module (usually advertising is carried out in the form of sound + light on the advertising terminal);

[0079] Mode switching logic:

[0080] Condition judgment (code example):

[0081] if child_detected:

[0082] switch_to_audio_mode() # Only enable audio output

[0083] else:

[0084] switch_to_normal_mode() # Enable audio-visual advertising;

[0085] Mode definition:

[0086] Audio mode: Turn off the screen backlight / OLED display and call pure audio advertising files (.mp3 / .wav);

[0087] Normal mode: Play video ads (.mp4) or picture carousel (.jpg + background music);

[0088] In summary, the system can achieve scene-based adaptation of advertising placement through real-time condition judgment, taking into account both child protection and commercial needs.

[0089] 3. Scene adjustment module

[0090] When in normal mode, extract and judge as a multi-person scene or a single-person scene based on the crowd density;

[0091] When it is determined to be a multi-person scene, the normal mode uses a vertical multi-line layout + parallel scrolling carousel mechanism to complete the advertising placement action; when it is determined to be a single-person scene, the normal mode uses a full-screen free layout + queue-based sequential carousel mechanism to complete the advertising placement action;

[0092] Among them, the process of judging as a multi-person scene or a single-person scene is as follows:

[0093] Obtain the number of people based on the crowd density. When the number of people = 1, it is determined to be a single-person scene;

[0094] When the number of people ≠ 1, it is determined to be a multi-person scene;

[0095] Specifically, multi-person scene → vertical multi-line layout + parallel scrolling carousel:

[0096] Layout design: The screen is vertically divided into multiple lines (e.g., 3 lines) to display promotional, brand, and public welfare ads respectively;

[0097] Parallel scrolling: Each line independently executes an upward scrolling animation, and the speed is determined according to the occupancy duration of the topmost ad. All ads in this part are short ads ≤ 10s; among them, the topmost ad is displayed in the form of sound + light. Except for the topmost ad position (i.e., other ads in the carousel state), they are all in silent + text or video (i.e., light) display, and the adjacent ads above and below are seamlessly connected;

[0098] Advantages: High-density information flow adapts to the rapid browsing needs of multiple people (such as peak elevator hours), can centrally display core promotional information (such as text ads + product main pictures), and improve the information transmission efficiency;

[0099] Specifically, single-person scenario → full-screen free layout + queue-style sequential carousel:

[0100] Layout design: Display a single advertisement full-screen (such as a car video + background music + interactive button), also presented in the form of sound + light, with the addition of an interactive button (which can be arranged in the bottommost area of the advertising terminal, and the touch screen area needs to be adaptively increased) to enhance interactivity. For the viewing of advertisements, manual selection is possible, and each advertisement in full-screen display is highlighted through a dynamic composition algorithm;

[0101] Sequential carousel: After the advertisement finishes playing, switch to the next advertisement content;

[0102] Advantages: The immersive experience adapts to the in-depth attention of a single person, enhancing visual and interactive attraction;

[0103] In summary, through the adaptive strategy driven by crowd density, the system achieves a dynamic balance between information transmission efficiency and user experience, solves the problem of rigid placement of traditional advertising terminals, can identify children in real time and switch the audio mode, taking into account both child protection and commercial needs, and also reduces the advertising investment cost to a certain extent; the information reach efficiency in the multi-person scenario is improved, and the interactive conversion rate in the single-person scenario is increased, realizing the efficient utilization of resources and enhancing the rationality of advertising content design and placement.

[0104] 4. Dynamic Adaptation Module

[0105] Under the condition of executing the parallel scrolling carousel mechanism, trigger the environmental data call instruction and execute the volume dynamic setting strategy to adjust the playback volume of the advertising terminal in real time; among them, the volume dynamic setting strategy has a built-in correction sub-strategy to correct the relationship between the playback volume and the environmental noise according to the constraint conditions;

[0106] Among them, the triggered environmental data call instruction is:

[0107] Call the crowd density and environmental noise in the environmental data, and preprocess the environmental data (used to organize the data for subsequent calculation or application), and the preprocessing includes: data cleaning, etc.;

[0108] In the executed volume dynamic setting strategy:

[0109] Restrict the adjusted volume according to the preprocessed environmental data, and the basis is as follows:

[0110] V adjusted = min(V base + k * Ro + S * (Nr - T), V max ) ;

[0111] Among them, V adjusted: Adjusted advertisement volume (playing volume) (a value between 0 and 100, representing the percentage of volume);

[0112] V base : Base volume (a value between 0 and 100, representing the percentage of volume);

[0113] k: Compensation coefficient for crowd sound absorption effect (the additional volume required for each additional person, in dB, with a value range: 0 to 1);

[0114] Ro: Number of people (i.e., the number of people in the target scenario (elevator));

[0115] S: Volume adjustment sensitivity, whose value can be adjusted according to the actual situation, used to control the impact of noise level (i.e., ambient noise) on volume, with a value range: 0 to 1;

[0116] Nr: Currently detected ambient noise (in decibels, dB);

[0117] T: Noise level threshold (in decibels, dB), when the ambient noise is lower than this value, the volume remains at the base volume;

[0118] V max : Maximum volume (also a value between 0 and 100, representing the percentage of volume, the set value);

[0119] Logical brief description or explanation:

[0120] Base volume: When the noise level in the elevator is lower than the noise level threshold T, the advertisement volume only considers the base volume V base and the number of people; Crowd sound absorption effect compensation: As the number of people in the elevator increases, due to the absorption of sound by clothes and bodies, the volume needs to be appropriately increased to ensure that the advertisement content can be clearly heard. This additional volume is represented by k*Ro; Noise level impact: When the noise level in the elevator (i.e., ambient noise) is higher than the noise level threshold T, the volume will increase with the increase in the noise level; The increased amount is jointly determined by the part by which the noise level exceeds the noise level threshold (Nr - T) and the volume adjustment sensitivity S; Volume limit: To avoid discomfort caused by excessive volume, the adjusted advertisement volume V adjusted will not exceed the maximum volume V max ;

[0121] Further explanation:

[0122] Crowd sound absorption effect compensation coefficient k: This coefficient can be determined according to actual tests to reflect the sound absorption effect of different crowds; Volume adjustment sensitivity S: This parameter determines how fast the volume changes with the noise level. If the value of S is large, the volume will increase rapidly as the noise level increases. If the value of S is small, the rate of volume increase will be relatively slow; Maximum volume V max : This parameter ensures that the volume does not exceed a comfortable range, protecting the passengers' ears from excessive volume damage;

[0123] To prove the effect of the above model formula, a set of test data is created, referring to Table 1 below:

[0124] Table 1: Observe the adjusted advertisement volume under different conditions;

[0125]

[0126] As can be seen from the above table:

[0127] Base volume: When the noise level (i.e., ambient noise) is lower than the noise level threshold T, the adjusted advertisement volume only considers the base volume V base and the number of people; Crowd sound absorption effect compensation coefficient: As the number of people increases, the volume gradually increases to compensate for sound absorption. For example, when the noise level is 70 dB and the number of people is 10, the volume increases from 37.5% to 40%; Noise level impact: When the noise level exceeds the noise level threshold T, the volume increases with the increase of the noise level. For example, when the number of people is 10, the noise level increases from 70 dB to 75 dB, and the volume increases from 40% to 42.5%; Volume limit: When the calculated volume exceeds the maximum volume V max , the volume is limited to V max ; For example, when the noise level is 85 dB and the number of people is 10, the volume is limited to 46% (for example only).

[0128] These data indicate that the model can dynamically adjust the advertisement volume according to the noise level and the number of people in the elevator, ensuring that the advertisement content can be clearly heard while avoiding discomfort caused by excessive volume;

[0129] The content of the correction sub-strategy built into the volume dynamic setting strategy is as follows:

[0130] Perform correction processing on the volume adjustment sensitivity S used in the volume dynamic setting strategy, judge whether the crowd density in the constraint conditions exceeds the preset critical threshold, and obtain the corrected volume adjustment sensitivity S based on the built correction function model - , and replace the original volume adjustment sensitivity S with the corrected volume adjustment sensitivity S - ;

[0131] When running the correction function model, the function definition is as follows:

[0132]

[0133] Among them, S max : The maximum value of the volume adjustment sensitivity S, which is obtained when D = D_th, so S max = 1;

[0134] α: A first constant greater than 0, controlling the attenuation rate of the sensitivity with the increase of the crowd density;

[0135] D_th: The critical threshold, used to judge whether the target scene is crowded;

[0136] β: A second constant greater than 0, controlling the decrease rate of the sensitivity with the increase of the crowd density, and α > β;

[0137] Logical brief description or explanation:

[0138] When D > D_th (indicating that the target scene is crowded), as D increases, S - decreases exponentially to prevent people near the advertising terminal from being affected by the rapidly changing advertising sound; when 1 < D ≤ D_th (indicating that the target scene is loose), as D increases, S decreases linearly because the crowd is relatively loose at this time and people are at a certain distance from the advertising terminal (in the actual scene, people will not actively approach the advertising terminal closely), so the advertising sound of the advertising terminal can increase rapidly to a certain extent;

[0139] For example:

[0140] Suppose S max = 1, α = 1, D_th = 2 people per square meter; when D = 2.5 people per square meter;

[0141] S - = 1.0 * e^(-1 * (2.5 - 2)) = 1.0 * e^(-0.5) ≈ 0.61 (rounded to two decimal places);

[0142] Suppose S max = 1, β = 0.2, when D = 1 person per square meter;

[0143] S- = 1.0 - 0.2 * (2 - 1) = 0.8;

[0144] Summary: We have defined two different functions to calculate the corrected volume adjustment sensitivity S - , corresponding to the cases of a crowded target scene and a loose target scene respectively. Both of these functions consider the influence of the crowd density D on the volume adjustment sensitivity S, but the specific forms and parameters are different to reflect the actual needs in different situations;

[0145] Among them, the constraint conditions are: whether the crowd density exceeds a preset critical threshold;

[0146] In addition, based on the original constraint conditions, it can also include:

[0147] When the distance value of the person closest to the advertising terminal in the target scenario is detected to be less than 0.5 m (an infrared sensor can be installed in the advertising terminal in the target scenario for detection, and the specific method will not be elaborated here), then regardless of whether the crowd density exceeds the preset critical threshold, it is determined that the target scenario is crowded, and the function corresponding to the crowded situation in the target scenario is used for calculation to obtain the corrected volume adjustment sensitivity S - 。

[0148] This solution solves the core problem of the fragmented experience (unable to hear clearly or noise interference) caused by the static volume setting of traditional advertising terminals through dual-dimensional regulation of the environment and the crowd. Based on the ambient noise and the crowd sound absorption effect, the volume is adjusted in real time, improving the clarity of the advertising volume. According to the crowd density, a correction sub-strategy for the volume sensitivity is triggered. When the crowd density exceeds the critical threshold, the smoothness of the volume change is improved, avoiding discomfort caused by sudden volume (reducing the complaint rate), realizing the leap of the advertising volume from "fixed output" to "intelligent adaptive", and also focusing on clear communication and auditory comfort.

[0149] Embodiment 2:

[0150] Please refer to Figure 2 , based on Embodiment 1, this embodiment also provides an Internet-based advertising placement management method, including the following specific steps:

[0151] S101. Capture the video stream in the target scenario, identify whether there are children in the target scenario, and analyze to obtain environmental data; among them, the environmental data includes at least the crowd density and the ambient noise;

[0152] S102. When children are identified, switch the advertising terminal in the target scenario to the audio mode; otherwise, switch the advertising terminal to the normal mode;

[0153] S103. When in the normal mode, extract and judge according to the crowd density to determine whether it is a multi-person scenario or a single-person scenario; when it is determined to be a multi-person scenario, the normal mode runs with a vertical multi-line layout combined with a parallel scrolling carousel mechanism; when it is determined to be a single-person scenario, the normal mode runs with a full-screen free layout combined with a queue-based sequential carousel mechanism;

[0154] S104. Under the condition of executing the parallel scrolling carousel mechanism, trigger the environmental data call instruction, execute the volume dynamic setting policy, and adjust the playback volume of the advertising terminal in real time; among them, the volume dynamic setting policy incorporates a correction sub-policy to correct the relationship between the playback volume and the environmental noise according to the constraint conditions.

[0155] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those of ordinary skill in the art will appreciate that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution.

[0156] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0157] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all of them should be covered by the protection scope of the present application.

Claims

1. An Internet-based advertising placement management system, characterized in that, The system includes: An environmental monitoring module that captures video streams in a target scenario, identifies whether there are children in the target scenario, and analyzes to obtain environmental data; among them, the environmental data at least includes crowd density and environmental noise; An advertising management design module that switches the advertising terminal in the target scenario to the audio mode when children are identified; otherwise, it switches the advertising terminal to the normal mode; A scene adjustment module that, when in the normal mode, extracts and judges as a multi-person scene or a single-person scene based on the crowd density; when it is determined as a multi-person scene, the normal mode runs with a vertical multi-line layout combined with a parallel scrolling carousel mechanism; when it is determined as a single-person scene, the normal mode runs with a full-screen free layout combined with a queue-based sequential carousel mechanism; A dynamic adaptation module that, under the condition of executing the parallel scrolling carousel mechanism, triggers an environmental data call instruction, executes a volume dynamic setting strategy, and adjusts the playback volume of the advertising terminal in real time; among them, the volume dynamic setting strategy has a built-in correction sub-strategy to correct the relationship between the playback volume and the environmental noise according to the constraint conditions.

2. The advertising placement management system based on the Internet according to claim 1, wherein: The process of analyzing and obtaining the crowd density is as follows: Real-time detection and statistics of human targets in the target scenario through a lightweight CNN model; The lightweight CNN model is trained with a crowd counting dataset; After outputting the number of people, combined with the floor area of the target scenario, calculate the crowd density D = number of people / floor area of the target scenario.

3. An Internet-based advertising placement management system according to claim 1, characterized in that: The process of analyzing and obtaining the environmental noise is as follows: Signal processing: Real-time capture of the original sound pressure signal in the target scenario and filtering with A-weighting; Real-time calculation: Obtain the instantaneous sound pressure at equal intervals and output the environmental noise based on the sound pressure formula; Calibration output: After triggering the calibration signal, the environmental noise is transmitted in real time through the Modbus protocol.

4. An Internet-based advertising placement management system according to claim 1, characterized in that: The audio mode means that the advertising terminal only enables audio output, and the normal mode means that the advertising terminal enables visual and audio advertisements.

5. An Internet-based advertising placement management system according to claim 4, characterized in that: The process of judging as a multi-person scene or a single-person scene is as follows: Obtain the number of people based on the crowd density. When the number of people = 1, it is determined as a single-person scene; When the number of people ≠ 1, it is determined as a multi-person scene.

6. The Internet-based advertising placement management system according to claim 5, wherein: The content of the normal mode running in a multi-person scene is: Vertical multi-line layout: The screen of the advertising terminal is vertically divided into several lines, and each line corresponds to different types of advertising content; Parallel scrolling carousel mechanism: Each line independently executes an upward scrolling animation, and the scrolling speed is determined according to the occupancy duration of the top advertising space; among them, the top advertising space enables visual and audio advertisements, and the other advertising spaces except the top one are in a silent display state; The content of the normal mode running in a single-person scene is: Full-screen free layout: The advertising terminal full-screen displays a single advertising content and adds interactive buttons; Queue-based sequential carousel mechanism: After the advertising content on the current page is played, it switches to the next advertising content, and so on in a loop.

7. An Internet-based advertising placement management system according to claim 1, characterized in that: The triggered environmental data call instruction is: Call the crowd density and environmental noise in the environmental data and preprocess the environmental data; among them, the preprocessing at least includes data cleaning.

8. An Internet-based advertising placement management system according to claim 7, characterized in that: In the executed volume dynamic setting strategy: Based on the preprocessed environmental data, restrict the adjustment of the volume, and the basis is as follows: V adjusted = min(V base + k * Ro + S * (Nr - T), V max ) Among them, V adjusted is the adjusted advertisement volume, that is, the playing volume, V base is the base volume, k is the compensation coefficient for the crowd sound absorption effect, Ro is the number of people in the crowd density, S is the volume adjustment sensitivity, Nr is the ambient noise, T is the noise level threshold, V max is the maximum volume.

9. The advertising placement management system based on the Internet according to claim 8, characterized in that: The content of the correction sub-strategy built in the volume dynamic setting strategy is as follows: Perform a correction process on the volume adjustment sensitivity S used in the volume dynamic setting strategy, determine whether the population density in the constraint conditions exceeds a preset critical threshold, and obtain the corrected volume adjustment sensitivity S based on the established correction function model - , and replace the original volume adjustment sensitivity S with the corrected volume adjustment sensitivity S - ; When running the correction function model, the function definition is as follows: Among them, S max is the maximum value of the volume adjustment sensitivity S, both α and β are constants, and D_th is the critical threshold.

10. An Internet-based advertising placement management method, using the system described in any one of claims 1 to 9, characterized in that: It includes the following steps: Capture the video stream in the target scenario, identify whether there are children in the target scenario, and analyze to obtain environmental data; among them, the environmental data at least includes crowd density and environmental noise; When children are identified, the advertising terminal in the target scenario is switched to the audio mode; otherwise, the advertising terminal is switched to the normal mode; When in the normal mode, extract and judge according to the crowd density to determine whether it is a multi-person scenario or a single-person scenario; when it is determined to be a multi-person scenario, the normal mode runs with a vertical multi-line layout combined with a parallel scrolling carousel mechanism; when it is determined to be a single-person scenario, the normal mode runs with a full-screen free layout combined with a queue-based sequential carousel mechanism; Under the condition of executing the parallel scrolling carousel mechanism, trigger the environmental data call instruction, execute the volume dynamic setting strategy, and adjust the playback volume of the advertising terminal in real time; among them, the volume dynamic setting strategy has a built-in correction sub-strategy to correct the relationship between the playback volume and the environmental noise according to the constraint conditions.

Images