An advertising monitoring method and system based on image AI analysis

Abstract

The present invention discloses a method and system for monitoring advertising delivery based on image AI analysis. The method comprises: integrating a shooting device on an advertising delivery device to obtain video data; using a target detection model to identify people in the video data, determining whether the people's eyes are looking at the advertising area, and counting the number of people looking at the advertising area; obtaining people's facial features and body posture features through a feature extraction model; classifying people according to age and gender; identifying and obtaining users' interactive behaviors when watching advertisements, and recording the number of interactive behaviors; and calculating advertising scores for different ages and genders based on the number of people who watched the advertisements, the number of interactive behaviors, and the duration of advertisement viewing. The present invention can accurately monitor the effectiveness of advertising delivery and provide advertisers with detailed audience behavior data so as to optimize advertising content, delivery channels, and strategies, effectively improve the attractiveness and delivery efficiency of advertisements, and enhance the return on investment and communication value of advertisements.

Description

Technical Field

[0001] The present invention relates to the field of advertising delivery technology, and in particular to an advertising delivery monitoring method and system based on image AI analysis. Background Art

[0002] In the current field of outdoor advertising monitoring, with the development of technology, camera monitoring technology has been widely used to monitor the effectiveness of advertising. By installing cameras in the advertising area, the situation around the advertisement can be recorded in real time, which to a certain extent improves the efficiency of monitoring and the ability to obtain data.

[0003] However, existing camera monitoring technology still has significant limitations. On the one hand, while cameras can capture image information of the advertising area, they cannot accurately identify the number of people viewing the ad. In real-world scenarios, due to factors such as crowd density, the dynamic movement of people, and the shooting angle, it is difficult for cameras to accurately distinguish each individual viewing the ad, making it prone to double counting or omissions, resulting in a significant deviation between the counted number of viewers and the actual number of viewers. On the other hand, current monitoring technology lacks in-depth analysis of the audience. Relying solely on camera footage, it is impossible to classify and study the audience based on different characteristics such as age and gender. This makes it impossible for advertisers to understand the reactions and acceptance levels of different audience groups to advertisements, making it difficult to adjust advertising content and delivery strategies in a targeted manner, and thus unable to achieve accurate advertising delivery. This results in a waste of advertising resources and reduces the effectiveness and return on investment of advertising. Summary of the Invention

[0004] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of this application to avoid obscuring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0005] In view of the above-mentioned existing problems, the present invention is proposed. Therefore, the present invention provides an advertising placement monitoring method based on image AI analysis to solve the problems in the existing technology, such as the inability to accurately identify the number of people viewing advertisements and the lack of analysis based on the characteristics of different audience groups, resulting in incomplete and inaccurate evaluation of advertising effectiveness.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an advertising delivery monitoring method based on image AI analysis, comprising:

[0008] Integrate a camera into the advertising device, and when a person is detected entering the visible area of ​​the advertisement, activate the camera function to record and obtain video data;

[0009] Use the object detection model to identify people in the video data, determine whether the people's eyes are looking at the advertising area, and count the number of people looking at the advertising area;

[0010] Build a feature extraction model to obtain facial features and body posture features of characters; classify characters by age and gender based on facial features; build a mapping relationship model between posture features and interactive behaviors based on body posture features, identify and obtain interactive behaviors of users when watching ads, and record the number of interactive behaviors;

[0011] Ad scores for different ages and genders are calculated based on the number of people who viewed the ad, the number of interactive behaviors, and the ad viewing time.

[0012] As a preferred solution of the advertising placement monitoring method based on image AI analysis described in the present invention, the method of using a target detection model to identify people in video data includes:

[0013] Preprocess the video data, including video frame extraction, resolution adjustment, image normalization and data enhancement operations;

[0014] Collect a large amount of image data containing people to build an object detection dataset, annotate the people with bounding boxes and category labels, and divide the annotated dataset into training, validation, and test sets in appropriate proportions.

[0015] After setting the number of training rounds and batch size for the target detection model, use the training set to train the target detection model, calculate the loss function, and update the target detection model parameters;

[0016] The preprocessed video frames are input into the trained target detection model for inference. The target detection model outputs the bounding boxes and confidence scores of the areas where people may exist. The bounding boxes are post-processed using the non-maximum suppression algorithm to remove redundant and erroneous detection results, and the bounding boxes with confidence scores higher than the set threshold are screened out. By analyzing their position and size information, the specific position and range of the person in the video frame are determined to complete the recognition of the person in the video data.

[0017] As a preferred solution of the advertising placement monitoring method based on image AI analysis described in the present invention, the method of determining whether a person's eyes are looking at the advertising area and counting the number of people looking at the advertising area includes:

[0018] Decomposing the video data into multiple frames of images, obtaining an image of the eye area of ​​the person, gray-scaling the extracted image of the eye area to calculate the brightness value of the image pixel points, and obtaining the pupil position of the person based on the brightness value;

[0019] Calculate the offset vector between the pupil center and the person's eye area. If the horizontal and vertical components of the offset vector are within a preset pixel range, it is preliminarily determined that the person is looking at the advertising area. Analyze five consecutive image frames. If the calculated offset vector for each of the five consecutive frames is within the preset pixel range, it is finally determined that the person is looking at the advertising area.

[0020] In each frame of the video, the person who is determined to be looking at the advertising area is marked, and a feature matching algorithm is used to track the person, record the person's outline and clothing color characteristics, and compare the features of the newly appeared suspected person with the recorded features to avoid double counting. The number of different people that have been marked and accurately matched is counted.

[0021] As a preferred solution of the advertising placement monitoring method based on image AI analysis described in the present invention, wherein: constructing a feature extraction model, and obtaining facial features of a person through the feature extraction model, including:

[0022] Design a feature extraction model using convolutional neural networks and human pose estimation algorithms;

[0023] The video data is decomposed into multiple frames, and the facial image of the person is extracted as input and input into the convolutional neural network. The facial image is subjected to sliding convolution in the convolutional layer to generate multiple feature maps. The feature maps output by the convolutional layer are transmitted to the activation layer and nonlinearly transformed using the ReLU function. The activated preliminary feature maps enter the pooling layer and are downsampled through the maximum pooling operation.

[0024] Repeat convolution, activation, and pooling operations to form a stack of multiple convolution-activation-pooling layers to obtain high-level feature maps, flatten the high-level feature maps into one-dimensional vectors, and input them into the fully connected layer to generate a fixed-length feature vector;

[0025] The feature vector is converted into a probability distribution for each category through a Softmax function, and the characters are classified according to age and gender based on the probability distribution.

[0026] As a preferred solution of the advertising placement monitoring method based on image AI analysis described in the present invention, wherein: constructing a feature extraction model, and obtaining the body posture features of the person through the feature extraction model, including:

[0027] The coordinates of the key points of the human body of the character in the video image are obtained, and based on the key point coordinates, the joint angles, limb length ratios and relative position relationships are calculated to obtain the character's body posture characteristics.

[0028] As a preferred solution of the advertising placement monitoring method based on image AI analysis described in the present invention, a mapping relationship model between posture features and interactive behaviors is constructed based on the body posture features of the person, the interactive behaviors of users when watching advertisements are identified and obtained, and the number of interactive behaviors is recorded, including:

[0029] Collect a large amount of video data containing different interactive behaviors, use the character body posture feature extraction method to obtain body posture feature data, and make detailed annotations of the interactive behaviors in the video to clarify the specific interactive behavior type corresponding to each posture feature;

[0030] A support vector machine is used to design a mapping relationship model. The support vector machine takes the collected and labeled posture feature data as input and the corresponding interactive behavior type as output. The mapping relationship model is trained to improve the classification accuracy of the model.

[0031] The body posture features of the person to be identified are input into the trained mapping relationship model. The mapping relationship model analyzes and judges the input posture features based on the learned mapping relationship and outputs the most likely corresponding interactive behavior type.

[0032] A state tracking mechanism is used to set a corresponding state identifier for each identified interactive behavior. When the mapping relationship model recognizes the start of an interactive behavior, the state identifier of the behavior is activated and the timing begins; when the behavior ends, the timing is stopped and the number of interactive behaviors is recorded.

[0033] As a preferred embodiment of the advertising placement monitoring method based on image AI analysis described in the present invention, the advertising scores of different ages and genders are calculated based on the number of people who viewed the advertisement, the number of interactive behaviors, and the duration of advertisement viewing, including:

[0034] The number of people who viewed the ad, the total viewing time, and the number of interactive behaviors were weighted, with the sum of the weights being 1. The number of people who viewed the ad, the total viewing time, and the number of interactive behaviors within each age group were normalized. A weighted calculation was performed based on the weight coefficient, the number of people who viewed the ad, the total viewing time, and the number of interactive behaviors to obtain the ad scores for different age groups.

[0035] The number of people who watched the ad, the total viewing time, and the number of interactive behaviors were weighted separately, and the sum of the weights was 1. The number of people who watched the ad, the total viewing time, and the number of interactive behaviors in the gender group were standardized, and a weighted calculation was performed based on the weight coefficient, the number of people who watched the ad, the total viewing time, and the number of interactive behaviors to obtain the ad scores for different gender groups.

[0036] In a second aspect, the present invention provides a system for monitoring advertising delivery based on image AI analysis, comprising:

[0037] The data acquisition module is used to integrate a camera on the advertising device, and when a person is detected entering the advertisement viewing area, the camera function is activated to record and acquire video data;

[0038] A statistics module is used to use the target detection model to identify people in the video data, determine whether the people's eyes are looking at the advertising area, and count the number of people looking at the advertising area;

[0039] The feature extraction module is used to build a feature extraction model to obtain facial features and body posture features of characters; based on facial features, characters are classified according to age and gender; based on body posture features, a mapping relationship model between posture features and interactive behaviors is built to identify and obtain interactive behaviors of users when watching advertisements, and the number of interactive behaviors is recorded;

[0040] The evaluation and monitoring module is used to calculate the advertising scores of different ages and genders based on the number of people who watched the advertisement, the number of interactive behaviors and the advertising viewing time data.

[0041] In a third aspect, the present invention provides a computing device, comprising:

[0042] memory and processor;

[0043] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the advertising delivery monitoring method based on image AI analysis are implemented.

[0044] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the advertising delivery monitoring method based on image AI analysis.

[0045] Compared with the existing technology, the beneficial effects of the present invention are as follows: the advertising delivery monitoring method based on image AI analysis of the present invention integrates a shooting device on the advertising delivery device and obtains video data, uses a target detection model to accurately identify the characters in the video, judges whether the characters are looking at the advertising area based on the pupil position and counts the number of people, and constructs a feature extraction model to obtain the facial and body posture features of the characters, realizes classification by age and gender, and recognizes and counts interactive behaviors, and finally calculates the advertising scores of different ages and genders based on the number of people who watch the advertisement, the number of interactive behaviors and the length of time the advertisement is viewed. It can accurately monitor the effect of advertising delivery and provide advertisers with detailed audience behavior data, so as to optimize advertising content, delivery channels and strategies, effectively improve the attractiveness and delivery efficiency of advertising, realize the rational allocation of advertising resources, and improve the return on investment and communication value of advertising. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0047] Figure 1 This is a schematic diagram of the overall process of an advertising delivery monitoring method based on image AI analysis according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0049] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0050] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0051] The present invention is described in detail with reference to schematic diagrams. For ease of illustration, cross-sectional views of device structures may be partially enlarged and not to scale when describing embodiments of the present invention. Furthermore, the schematic diagrams are merely illustrative and should not limit the scope of the present invention. Furthermore, in actual production, the three-dimensional dimensions of length, width, and depth should be included.

[0052] Furthermore, in the description of the present invention, it should be noted that the terms "upper, lower, inner, and outer" and other references to orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the systems or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first, second, or third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0053] In this disclosure, unless otherwise specified or limited, the terms "mounted," "connected," and "connected" should be interpreted broadly. For example, they may refer to fixed, removable, or integral connections. They may also refer to mechanical, electrical, or direct connections, indirect connections through an intermediary, or internal communication between two components. Those skilled in the art will understand the specific meanings of these terms in this disclosure.

[0054] Example 1

[0055] Reference Figure 1 , as one embodiment of the present invention, provides an advertising delivery monitoring method based on image AI analysis, comprising:

[0056] S1: Integrate a camera on the advertising device. When a person is detected entering the visible area of ​​the advertisement, start the camera to record and obtain video data.

[0057] Preferably, the video data is preprocessed, the preprocessing including video frame extraction, resolution adjustment, image normalization and data enhancement operations;

[0058] Specifically, video frame extraction involves segmenting continuous video data into frames at regular intervals, providing the basic data units for subsequent image analysis. Because different shooting devices or conditions may result in inconsistent video resolution, resolution adjustment is performed to uniformly adjust the resolution of all video frames to a size suitable for subsequent processing to ensure accurate and consistent analysis. Normalizing the pixel values ​​of an image maps them to a specific range, eliminating the effects of differences in pixel value distribution between images and making subsequent image processing more stable and efficient.

[0059] Data enhancement includes rotation, flipping, cropping, adding noise, etc. By performing data enhancement operations on the original video frames, new image samples can be generated, increasing the amount of training data, thereby improving the generalization ability of subsequent models and enabling them to perform better in various practical scenarios.

[0060] S2: Use the object detection model to identify people in the video data, determine whether the people's eyes are looking at the advertising area, and count the number of people looking at the advertising area;

[0061] Preferably, identifying people in video data using an object detection model specifically includes the following steps:

[0062] Collect a large amount of image data containing people to build an object detection dataset, annotate the people with bounding boxes and category labels, and divide the annotated dataset into training, validation, and test sets in appropriate proportions.

[0063] Set the number of training rounds and batch size for the object detection model. The number of training rounds represents the number of times the model traverses the entire training dataset, while the batch size represents the number of samples used for each parameter update. Too few training rounds will result in insufficient model learning, while too many may lead to overfitting. Too large a batch size may cause the model to fall into a local optimum, while too small a batch size will reduce training efficiency. During training, the object detection model processes the training data in batches and performs forward propagation to produce predictions. To measure the difference between the predicted results and the true labels, the mean squared error (MSE) is used as the loss function. The MSE calculates the average of the sum of the squares of the differences between the predicted and true values, reflecting the degree of deviation from the prediction. Based on this loss function value, the model parameters are updated. After multiple rounds of training, the object detection model continuously learns the data characteristics and gradually improves its ability to detect people in videos.

[0064] The validation set is used to evaluate the performance of the object detection model during training and adjust its hyperparameters. After completing a certain number of training rounds, the validation set is fed into the currently trained model. The model then performs object detection on the images in the validation set, outputting bounding boxes and confidence scores. The model's detection results are then compared with the ground truth data labeled in the validation set. Metrics such as precision and recall are used to measure the model's performance on the validation set. If the model's performance on the validation set falls short of expectations, the model's parameters are adjusted.

[0065] After the target detection model has been adjusted and optimized multiple times through the validation set to achieve a relatively satisfactory performance, the trained target detection model is then input into the test set for inference. The target detection model outputs the bounding box and confidence score of the area where people may exist. The bounding box is post-processed using the non-maximum suppression algorithm to remove redundant and erroneous detection results, and the bounding box with a confidence level higher than the set threshold is screened out. By analyzing its position and size information, the specific position and range of the person in the video frame are determined, and the person in the video data is identified.

[0066] Preferably, counting the number of people looking at the advertising area specifically includes the following steps:

[0067] Decompose the video data into multiple frames of images, obtain the image of the eye area of ​​the person, perform grayscale processing on the extracted eye area image to calculate the brightness value of the image pixel, and obtain the pupil position of the person based on the brightness value;

[0068] Calculate the offset vector between the pupil center and the person's eye area. If the horizontal and vertical components of the offset vector are within a preset pixel range, it is preliminarily determined that the person is looking at the advertising area. Analyze five consecutive image frames. If the calculated offset vector for each of the five consecutive frames is within the preset pixel range, it is finally determined that the person is looking at the advertising area.

[0069] In each frame of the video, the person who is determined to be looking at the advertising area is marked, and a feature matching algorithm is used to track the person, record the person's outline and clothing color characteristics, and compare the features of the newly appeared suspected person with the recorded features to avoid double counting. The number of different people that have been marked and accurately matched is counted.

[0070] It should be noted that the present invention utilizes a target detection model to identify people in video data, conducts scientific training based on a data set constructed from a large amount of labeled data, and through validation set optimization and test set reasoning, achieves precise positioning of people, effectively avoids insufficient learning and overfitting problems in the model, and can improve detection capabilities and generalization. On the basis of the target detection model detecting people, the video data is decomposed into multiple frames of images, and the image of the person's eye area is grayed to calculate the pixel brightness value to obtain the pupil position. Then, the calculated offset vector and multi-frame image analysis are combined to accurately determine whether the person is looking at the advertising area. And when counting the number of people looking at the advertising area, a feature matching algorithm is used to track the person, record the person's outline and clothing color features, avoid double counting, and achieve accurate statistics of the number of people looking at the advertising area.

[0071] S3 builds a feature extraction model to obtain facial features and body posture features of characters; based on facial features, it classifies characters by age and gender; based on body posture features, it builds a mapping relationship model between posture features and interactive behaviors, identifies and obtains users' interactive behaviors when watching ads, and records the number of interactive behaviors;

[0072] Preferably, a feature extraction model is designed by using a convolutional neural network and a human posture estimation algorithm;

[0073] Preferably, obtaining facial features of a person through a feature extraction model specifically includes the following steps:

[0074] The video data is decomposed into multiple frames, and the facial image of the person is extracted as input and input into the convolutional neural network. The facial image is subjected to sliding convolution in the convolutional layer to generate multiple feature maps. The feature maps output by the convolutional layer are transmitted to the activation layer and nonlinearly transformed using the ReLU function. The activated preliminary feature maps enter the pooling layer and are downsampled through the maximum pooling operation.

[0075] Repeat convolution, activation, and pooling operations to form a stack of multiple convolution-activation-pooling layers to obtain high-level feature maps, flatten the high-level feature maps into one-dimensional vectors, and input them into the fully connected layer to generate a fixed-length feature vector;

[0076] The feature vector is converted into a probability distribution for each category through the Softmax function, and the characters are classified according to age and gender based on the probability distribution;

[0077] Specifically, the convolutional layers of a convolutional neural network capture the shape and positional features of parts like the eyes, nose, and mouth. After multiple layers of convolution, activation, and pooling, facial bone structure and skin texture features are extracted. These features at different levels are combined and transformed to generate a fixed-length feature vector.

[0078] The feature vector is input into the Softmax function, which converts it into a probability distribution for each category. For age classification, the person is divided into different age groups, such as children, adolescents, young adults, middle-aged, and elderly, based on the principle of maximum probability. For gender classification, the person is determined to be male or female based on probability, and the category with the highest probability is the age or gender classification result for that person.

[0079] Preferably, obtaining the body posture features of a person through a feature extraction model specifically includes the following steps:

[0080] Obtain the coordinates of the key points of the human body of the person in the video image, calculate the joint angles, limb length ratios and relative position relationships based on the key point coordinates, and obtain the person's body posture characteristics;

[0081] The key points of the human body include the midpoint of the neck, left and right shoulders, left and right elbows, left and right wrists, left and right hip joints, left and right knee joints, left and right ankles and other key parts;

[0082] Collect a large amount of video data containing different interactive behaviors, use the character body posture feature extraction method to obtain body posture feature data, and make detailed annotations of the interactive behaviors in the video to clarify the specific interactive behavior type corresponding to each posture feature;

[0083] A support vector machine is used to design a mapping relationship model. The support vector machine takes the collected and labeled posture feature data as input and the corresponding interactive behavior type as output. The mapping relationship model is trained to improve the classification accuracy of the model.

[0084] The body posture features of the person to be identified are input into the trained mapping relationship model. The mapping relationship model analyzes and judges the input posture features based on the learned mapping relationship and outputs the most likely corresponding interactive behavior type.

[0085] A state tracking mechanism is used to set a corresponding state identifier for each identified interactive behavior. When the mapping relationship model recognizes the start of an interactive behavior, the state identifier of the behavior is activated and the timing begins; when the behavior ends, the timing is stopped and the number of interactive behaviors is recorded.

[0086] Specifically, a mapping relationship model is constructed based on the calculation results of joint angles, limb length ratios and relative positions. The mapping relationship model is pre-trained with a large amount of labeled human posture data. These annotations clearly define the specific interactive behaviors corresponding to different posture feature combinations. When the body posture features calculated in real time are input, the mapping relationship model will make judgments based on the learned patterns, identify the specific interactive behaviors of the characters when watching advertisements, and record the number of times each interactive behavior occurs.

[0087] It should be noted that the present invention can comprehensively and deeply capture facial features from basic parts to high-level abstractions through multi-layer operations of convolutional neural networks. The generated feature vectors can be accurately classified by age and gender through the Softmax function, providing a data basis for subsequent advertising ratings of different audiences. In the extraction of character body posture features, the present invention obtains posture feature data by obtaining key point coordinates and calculating correlations. Combined with a large amount of annotated video data, the mapping relationship model trained with a support vector machine can accurately identify interactive behaviors, and at the same time, a state tracking mechanism is used to ensure accurate counting.

[0088] S4: Calculate the ad scores for different ages and genders based on the number of people who viewed the ad, the number of interactive behaviors, and the ad viewing time.

[0089] Preferably, weights are assigned to the number of people who viewed the ad, the total viewing time, and the number of interactive behaviors, and the sum of the weights is 1. The number of people who viewed the ad, the total viewing time, and the number of interactive behaviors in each age group are normalized, and a weighted calculation is performed based on the weight coefficient, the number of people who viewed the ad, the total viewing time, and the number of interactive behaviors to obtain the ad scores for different age groups.

[0090] Preferably, weights are assigned to the number of people who watch the advertisement, the total viewing time, and the number of interactive behaviors, and the sum of the weights is 1. The number of people who watch the advertisement, the total viewing time, and the number of interactive behaviors in the gender group are standardized, and a weighted calculation is performed based on the weight coefficient, the number of people who watch the advertisement, the total viewing time, and the number of interactive behaviors to obtain the advertisement scores for different gender groups.

[0091] Based on the calculation results, comprehensive and in-depth monitoring and analysis of advertising effectiveness can be conducted. The ad scores for different age groups can clearly indicate the level of acceptance and interest of audiences in each age group. The ad scores for different gender groups can determine the performance differences between male and female audiences. If a particular gender scores poorly, relevant ad elements, such as the visual style and copywriting, can be specifically monitored to determine if they lack appeal to that gender. By continuously monitoring these calculation results, advertisers and publishers can promptly identify problems in the advertising process and dynamically adjust advertising strategies to improve the effectiveness of advertising among audiences of different ages and genders, achieving precise and efficient advertising. This, in turn, enhances the overall communication value and commercial benefits of advertising, ensuring that advertising reaches the target audience to the greatest extent possible and generates positive feedback.

[0092] The above is a schematic scheme of an advertising delivery monitoring method based on image AI analysis in this embodiment. It should be noted that the technical solution of the system for monitoring advertising delivery based on image AI analysis and the technical solution of the above-mentioned advertising delivery monitoring method based on image AI analysis are of the same concept. For details not described in detail in the technical solution of the advertising delivery monitoring system based on image AI analysis in this embodiment, please refer to the description of the technical solution of the above-mentioned advertising delivery monitoring method based on image AI analysis.

[0093] The advertising delivery monitoring system based on image AI analysis in this embodiment includes:

[0094] The data acquisition module is used to integrate a camera on the advertising device, and when a person is detected entering the advertisement viewing area, the camera function is activated to record and acquire video data;

[0095] A statistics module is used to use the target detection model to identify people in the video data, determine whether the people's eyes are looking at the advertising area, and count the number of people looking at the advertising area;

[0096] The feature extraction module is used to build a feature extraction model to obtain facial features and body posture features of characters; based on facial features, characters are classified according to age and gender; based on body posture features, a mapping relationship model between posture features and interactive behaviors is built to identify and obtain interactive behaviors of users when watching advertisements, and the number of interactive behaviors is recorded;

[0097] The evaluation and monitoring module is used to calculate the advertising scores of different ages and genders based on the number of people who watched the advertisement, the number of interactive behaviors and the advertising viewing time data.

[0098] This embodiment further provides a computing device suitable for use in monitoring advertising delivery based on image AI analysis, including:

[0099] Memory and processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the advertising delivery monitoring method based on image AI analysis as proposed in the above embodiment.

[0100] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for monitoring advertising delivery based on image AI analysis proposed in the above embodiment is implemented.

[0101] The storage medium proposed in this embodiment and the advertising delivery monitoring method based on image AI analysis proposed in the above embodiment belong to the same inventive concept. For technical details not fully described in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0102] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general-purpose hardware. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0103] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An advertising placement monitoring method based on image AI analysis, characterized in that: include: Integrate a camera into the advertising device, and when a person is detected entering the visible area of ​​the advertisement, activate the camera function to record and obtain video data; Use the object detection model to identify people in the video data, determine whether the people's eyes are looking at the advertising area, and count the number of people looking at the advertising area; Decomposing the video data into multiple frames of images, obtaining an image of the eye area of ​​the person, gray-scaling the extracted image of the eye area to calculate the brightness value of the image pixel points, and obtaining the pupil position of the person based on the brightness value; Calculate the offset vector between the pupil center and the person's eye area. If the horizontal and vertical components of the offset vector are within a preset pixel range, it is preliminarily determined that the person is looking at the advertising area. Analyze five consecutive frames of images. If the offset vector calculated for each of the five consecutive frames is within the preset pixel range, it is ultimately determined that the person is looking at the advertising area. In each frame of the video, the person determined to be looking at the ad area is marked and tracked using a feature matching algorithm, recording the person's outline and clothing color characteristics. The features of newly appeared suspected people looking at the ad area are compared with the recorded features to avoid double counting, and the number of different marked and accurately matched people is counted. Construct a feature extraction model to obtain facial features and body posture features of the character; Based on facial features, characters are classified by age and gender. Based on the characters' body posture features, a mapping relationship model between posture features and interactive behaviors is constructed to identify and obtain users' interactive behaviors when watching ads, and the number of interactive behaviors is recorded. Ad scores for different ages and genders are calculated based on the number of people who viewed the ad, the number of interactive behaviors, and the ad viewing time.

2. The advertising placement monitoring method based on image AI analysis according to claim 1, characterized in that: Use object detection models to identify people in video data, including: Preprocess the video data, including video frame extraction, resolution adjustment, image normalization and data enhancement operations; Collect a large amount of image data containing people to build an object detection dataset, annotate the people with bounding boxes and category labels, and divide the annotated dataset into training, validation, and test sets in appropriate proportions. After setting the number of training rounds and batch size for the target detection model, use the training set to train the target detection model, calculate the loss function, and update the target detection model parameters; The preprocessed video frames are input into the trained target detection model for inference. The target detection model outputs the bounding boxes and confidence scores of the areas where people may exist. The bounding boxes are post-processed using the non-maximum suppression algorithm to remove redundant and erroneous detection results, and the bounding boxes with confidence scores higher than the set threshold are screened out. By analyzing their position and size information, the specific position and range of the person in the video frame are determined to complete the recognition of the person in the video data.

3. The advertising placement monitoring method based on image AI analysis according to claim 2, characterized in that: Build a feature extraction model to obtain facial features of the person, including: Design a feature extraction model using convolutional neural networks and human pose estimation algorithms; The video data is decomposed into multiple frames, and the facial image of the person is extracted as input and input into the convolutional neural network. The facial image is subjected to sliding convolution in the convolutional layer to generate multiple feature maps. The feature maps output by the convolutional layer are transmitted to the activation layer and nonlinearly transformed using the ReLU function. The activated preliminary feature maps enter the pooling layer and are downsampled through the maximum pooling operation. Repeat convolution, activation, and pooling operations to form a stack of multiple convolution-activation-pooling layers to obtain high-level feature maps, flatten the high-level feature maps into one-dimensional vectors, and input them into the fully connected layer to generate a fixed-length feature vector; The feature vector is converted into a probability distribution for each category through a Softmax function, and the characters are classified according to age and gender based on the probability distribution.

4. The advertising placement monitoring method based on image AI analysis according to claim 3, characterized in that: Build a feature extraction model to obtain the body posture features of the character, including: The coordinates of the key points of the human body of the character in the video image are obtained, and based on the key point coordinates, the joint angles, limb length ratios and relative position relationships are calculated to obtain the character's body posture characteristics.

5. The advertising placement monitoring method based on image AI analysis according to claim 4, characterized in that: Based on the body posture characteristics of the characters, a mapping relationship model between posture characteristics and interactive behaviors is constructed to identify and obtain the interactive behaviors of users when watching ads, and record the number of interactive behaviors, including: Collect a large amount of video data containing different interactive behaviors, use the character body posture feature extraction method to obtain body posture feature data, and make detailed annotations of the interactive behaviors in the video to clarify the specific interactive behavior type corresponding to each posture feature; A support vector machine is used to design a mapping relationship model. The support vector machine takes the collected and labeled posture feature data as input and the corresponding interactive behavior type as output. The mapping relationship model is trained to improve the classification accuracy of the model. The body posture features of the person to be identified are input into the trained mapping relationship model. The mapping relationship model analyzes and judges the input posture features based on the learned mapping relationship and outputs the most likely corresponding interactive behavior type. A state tracking mechanism is used to set a corresponding state identifier for each identified interactive behavior. When the mapping relationship model recognizes the start of an interactive behavior, the state identifier of the behavior is activated and the timing begins; when the behavior ends, the timing is stopped and the number of interactive behaviors is recorded.

6. The advertising placement monitoring method based on image AI analysis according to claim 5, characterized in that: Based on the number of people who viewed the ad, the number of interactive behaviors, and the ad viewing time, the ad scores for different ages and genders are calculated, including: The number of people who viewed the ad, the total viewing time, and the number of interactive behaviors were weighted, with the sum of the weights being 1. The number of people who viewed the ad, the total viewing time, and the number of interactive behaviors within each age group were normalized. A weighted calculation was performed based on the weight coefficient, the number of people who viewed the ad, the total viewing time, and the number of interactive behaviors to obtain the ad scores for different age groups. The number of people who watched the ad, the total viewing time, and the number of interactive behaviors were weighted separately, and the sum of the weights was 1. The number of people who watched the ad, the total viewing time, and the number of interactive behaviors in the gender group were standardized, and a weighted calculation was performed based on the weight coefficient, the number of people who watched the ad, the total viewing time, and the number of interactive behaviors to obtain the ad scores for different gender groups.

7. A system for monitoring advertising based on image AI analysis, characterized in that: include, The data acquisition module is used to integrate a camera on the advertising device, and when a person is detected entering the advertisement viewing area, the camera function is activated to record and acquire video data; A statistics module is used to use the target detection model to identify people in the video data, determine whether the people's eyes are looking at the advertising area, and count the number of people looking at the advertising area; decompose the video data into multiple frames, obtain images of the people's eye areas, grayscale the extracted eye area images, calculate the brightness values ​​of the image pixels, and obtain the pupil positions of the people based on the brightness values; Calculate the offset vector between the pupil center and the person's eye area. If the horizontal and vertical components of the offset vector are within a preset pixel range, it is preliminarily determined that the person is looking at the advertising area. Five consecutive frames are analyzed. If the calculated offset vector for each of the five consecutive frames is within a preset pixel range, the person is ultimately determined to be gazing at the ad area. In each frame of the video, the person determined to be gazing at the ad area is marked and tracked using a feature matching algorithm, recording their outline and clothing color. The features of any newly appeared suspected person are compared with the recorded features to avoid double counting, and the number of unique people that have been marked and accurately matched is counted. The feature extraction module is used to build a feature extraction model to obtain facial features and body posture features of characters; based on facial features, characters are classified according to age and gender; based on body posture features, a mapping relationship model between posture features and interactive behaviors is built to identify and obtain interactive behaviors of users when watching advertisements, and the number of interactive behaviors is recorded; The evaluation and monitoring module is used to calculate the advertising scores of different ages and genders based on the number of people who watched the advertisement, the number of interactive behaviors and the advertising viewing time data.

8. An electronic device, characterized in that: include: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the advertising delivery monitoring method based on image AI analysis as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that It stores computer-executable instructions, which, when executed by a processor, implement the steps of the advertising delivery monitoring method based on image AI analysis as described in any one of claims 1 to 6.

Images