Dimming and content recommendation method for sensing ambient light and human traffic of electronic screen

Abstract

The invention discloses an electronic screen ambient light and people flow rate perception dimming and content recommendation method, and relates to the technical field of electronic screen display, and the method comprises the steps: building a light intensity field and a people flow density field based on ambient light intensity data and people flow point cloud data; performing coupling processing on the light intensity field and the people flow density field to obtain coupling field intensity, and calculating a visual saliency field; constructing an energy functional based on the coupling field intensity and the visual saliency field, and obtaining an optimal brightness distribution field by solving an Euler-Lagrange equation of the energy functional; calculating an information flow potential well depth based on the visual saliency field, and obtaining a content matching degree matrix in combination with people flow density field features; and based on the optimal brightness distribution field and the content matching degree, establishing a collaborative dynamics system, and obtaining a brightness adjustment instruction and a content selection instruction through Lyapunov stability analysis. The brightness and content recommendation of the electronic screen are dynamically adjusted by sensing the ambient light intensity and the people flow density in real time.

Description

Technical Field

[0001] This invention relates to the field of electronic screen display technology, specifically to a method for dimming and content recommendation based on ambient light and pedestrian traffic perception of electronic screens. Background Technology

[0002] With the continuous development of intelligent display technology and digital advertising, electronic screens have been widely used in public places and commercial environments. To improve the display effect and user experience of electronic screens, especially in dynamic environments, real-time adjustment of display brightness and recommended content has become a key technical issue. Traditional methods for adjusting electronic screen brightness mostly rely on fixed time periods or ambient light sensing, lacking effective responses to changes in pedestrian traffic and dynamic environments. Meanwhile, existing content recommendation systems are mostly based on users' historical behavior or simple environmental data, lacking joint adjustment of displayed content with ambient light intensity and pedestrian density, and thus unable to provide accurate personalized displays and content recommendations in different scenarios.

[0003] This invention provides a method for adjusting the brightness of an electronic screen and recommending content based on ambient light intensity and pedestrian traffic. This method can sense changes in ambient light and pedestrian traffic in real time and adjust accordingly based on this data, thereby achieving intelligent brightness control and personalized content recommendation, improving the display effect and user experience of the electronic screen. Summary of the Invention

[0004] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a method for dimming and content recommendation of electronic screens based on ambient light and pedestrian flow perception, so as to solve the above-mentioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens, comprising: S1: Real-time ambient light intensity data is collected through an ambient light sensor, and point cloud data of pedestrian flow covering the monitoring area in front of the electronic screen is collected through a pedestrian flow monitoring sensor. S2: Based on ambient light intensity data and pedestrian flow point cloud data, construct a spatiotemporally continuous light intensity field and pedestrian flow density field respectively; S3: Couple the light intensity field and the crowd density field to obtain the coupled field strength, and calculate the visual saliency field based on the coupled field strength; S4: Construct an energy functional of the screen brightness distribution based on the coupling field strength and visual saliency field, and obtain the optimal brightness distribution field by solving the Euler-Lagrange equation of the energy functional; S5: Calculate the information flow potential trap depth based on the visual saliency field, and combine it with the characteristics of the crowd density field to obtain the content matching degree matrix; S6: Based on the optimal brightness distribution field and content matching degree, a collaborative dynamic system is established, and brightness adjustment instructions and content selection instructions are obtained through Lyapunov stability analysis.

[0006] The present invention is further configured such that S1 includes: Using the center of the electronic screen display plane as the spatial origin, and based on the ambient light intensity data collected by the ambient light sensor, a Gaussian distribution model is used to construct a spatiotemporally continuous light intensity field in which the ambient light intensity decays from the display plane to the surrounding space. Based on the pedestrian point cloud data collected by pedestrian monitoring sensors, the discrete pedestrian point cloud data is transformed into a spatiotemporally continuous pedestrian density field using a kernel density estimation algorithm, and a corresponding weighting coefficient is preset according to the movement speed of each data point.

[0007] The present invention is further configured such that S2 includes: Within the preset spatial integration domain, tensor product operations are performed on the light intensity field and the crowd density field. A Gaussian kernel function based on Euclidean distance is used to perform spatial convolution calculation on the tensor product operation results. The coupling field strength is obtained by double integration of the convolution result within the integration domain. Here, the Euclidean distance is the straight-line distance between the current calculation point and the integration variable point. The Laplacian operator for calculating the coupling field strength is used to detect the location of singular points in the field distribution by solving the absolute value of the Laplacian operator. These singular points correspond to the extreme regions of the visual attention distribution. The gradient magnitude of the crowd density field is calculated, and after a nonlinear transformation, it is multiplied by the absolute value of the Laplace operator to generate the visual saliency field.

[0008] The present invention is further configured such that S3 includes: An energy functional for screen brightness distribution is constructed, which is obtained by a weighted combination of brightness matching term, spatial smoothing term and visual saliency enhancement term; The brightness matching term is constructed based on the deviation between the actual brightness distribution and the ideal brightness distribution based on the coupling field strength; The spatial smoothing term is constructed based on the spatial gradient variation amplitude of the actual brightness distribution; The visual saliency enhancement term is constructed based on the nonlinear mapping relationship between actual brightness and the visual saliency field, and the mapping relationship is realized through an exponential decay function model; The Euler-Lagrange variational method is used to solve the extremum problem of the energy functional, and the partial differential equation describing the optimal brightness distribution is derived. By solving the partial differential equation, the optimal brightness distribution field that minimizes the energy functional is obtained.

[0009] The present invention is further configured such that S4 includes: The depth of the information flow potential trap is obtained by performing an exponential transformation on the integral result of the visual saliency field in the preset spatial integration domain and taking the negative logarithm. Based on the information flow potential well depth, the time rate of change of the information flow potential well depth, and the Laplace operator of the crowd density field, a field feature vector is constructed. Calculate the weighted Euclidean distance between the predefined content feature vector and the field feature vector, and perform an exponential mapping on the weighted Euclidean distance to obtain the content matching degree matrix.

[0010] The present invention is further configured such that S5 includes: Construct a collaborative dynamic system, which includes a brightness adjustment dynamic equation and a content matching degree evolution dynamic equation; The brightness adjustment dynamic equation defines a negative correlation between the rate of change of brightness over time and the energy functional gradient corresponding to the optimal brightness distribution location, so as to ensure a dynamic balance between brightness adjustment and energy minimization. The content matching degree evolution dynamic equation defines the functional relationship between the rate of change of content matching degree over time and the current state of content matching degree and the optimal brightness distribution field, so as to dynamically adjust the coupling between content and brightness; The steady-state solution of the cooperative dynamics system is determined based on the Lyapunov function, which is the sum of the weighted norms of the energy functional and the deviation between the content matching degree and the ideal matching degree. By solving for the minimum value of the Lyapunov function, the optimal steady-state solution of the cooperative dynamics system is obtained. This optimal steady-state solution corresponds to the final brightness adjustment command and content selection command to control the display of the electronic screen.

[0011] The present invention is further configured such that the brightness adjustment command and the content selection command are sent to the control system of the electronic screen via wireless communication.

[0012] The present invention is further configured such that the ambient light sensor and the pedestrian flow monitoring sensor transmit data wirelessly.

[0013] The present invention is further configured such that the ambient light intensity data is synchronously collected by ambient light sensors distributed at multiple preset positions around the electronic screen, and the collected data is obtained after fusion processing.

[0014] This invention provides a method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens. The method comprises the following steps: S1: Real-time acquisition of ambient light intensity data using an ambient light sensor and point cloud data of pedestrian flow covering the monitoring area in front of the electronic screen using a pedestrian flow monitoring sensor; S2: Construction of a spatiotemporally continuous light intensity field and pedestrian flow density field based on the ambient light intensity data and pedestrian flow point cloud data; S3: Coupling the light intensity field and pedestrian flow density field to obtain the coupling field strength, and calculating the visual saliency field based on the coupling field strength; S4: Construction of an energy functional for the screen brightness distribution based on the coupling field strength and visual saliency field, and obtaining the optimal brightness distribution field by solving the Euler-Lagrange equation of the energy functional; S5: Calculation of the information flow potential well depth based on the visual saliency field, and obtaining the content matching degree matrix by combining the characteristics of the pedestrian flow density field; S6: Establishment of a co-dynamic system based on the optimal brightness distribution field and content matching degree, and obtaining brightness adjustment instructions and content selection instructions through Lyapunov stability analysis. The beneficial effects include: 1. By dynamically adjusting the brightness and recommended content of the electronic screen in real time based on ambient light intensity and pedestrian density, this invention, compared to traditional static dimming methods, can accurately match the current environment and user needs, making the display effect of the electronic screen more comfortable and avoiding the impact of excessively strong or weak lighting on the viewer's visual experience. Simultaneously, by combining content recommendations with pedestrian density, it ensures that high-traffic areas display more eye-catching content, improving the effectiveness of advertising and the efficiency of information delivery, and enhancing the user experience. 2. By combining the analysis of light intensity field and crowd density field, the visual saliency field and content matching degree matrix are calculated. Based on the optimal brightness distribution field and user behavior pattern, personalized content recommendation is carried out. It can provide more accurate and customized content according to the needs of different users, attention distribution and current environmental conditions, thereby improving the advertising effect or information delivery efficiency.

[0015] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 The flowchart illustrates an exemplary embodiment of the present invention, showing a method for dimming and recommending content based on ambient light and pedestrian traffic perception of an electronic screen. Detailed Implementation

[0017] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0018] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0019] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0020] Methods for dimming and content recommendation based on ambient light and pedestrian traffic perception on electronic screens, such as... Figure 1 As shown, it includes: S1: Real-time ambient light intensity data is collected through an ambient light sensor, and point cloud data of pedestrian flow covering the monitoring area in front of the electronic screen is collected through a pedestrian flow monitoring sensor. S2: Based on ambient light intensity data and pedestrian flow point cloud data, construct a spatiotemporally continuous light intensity field and pedestrian flow density field respectively; S3: Couple the light intensity field and the crowd density field to obtain the coupled field strength, and calculate the visual saliency field based on the coupled field strength; S4: Construct an energy functional of the screen brightness distribution based on the coupling field strength and visual saliency field, and obtain the optimal brightness distribution field by solving the Euler-Lagrange equation of the energy functional; S5: Calculate the information flow potential trap depth based on the visual saliency field, and combine it with the characteristics of the crowd density field to obtain the content matching degree matrix; S6: Based on the optimal brightness distribution field and content matching degree, a collaborative dynamic system is established, and brightness adjustment instructions and content selection instructions are obtained through Lyapunov stability analysis.

[0021] The present invention is further configured such that S1 includes: Using the center of the electronic screen display plane as the spatial origin, and based on the ambient light intensity data collected by the ambient light sensor, a Gaussian distribution model is used to construct a spatiotemporally continuous light intensity field in which the ambient light intensity decays from the display plane to the surrounding space. Based on pedestrian point cloud data collected by pedestrian monitoring sensors, a kernel density estimation algorithm is used to transform discrete pedestrian point cloud data into a spatiotemporally continuous pedestrian density field, and a corresponding weighting coefficient is preset according to the movement speed of each data point. Specifically, an ambient light sensor continuously collects raw light intensity data, processes the raw data using a sliding window mean filtering algorithm to eliminate instantaneous fluctuations, and outputs calibrated ambient light intensity values ​​at a preset fixed frequency. Simultaneously, the pedestrian monitoring sensor scans the monitoring area in real time, distinguishes and identifies different pedestrian individuals through a point cloud clustering algorithm, and records the coordinates of each individual on a preset two-dimensional plane. The system uses the position coordinates and instantaneous velocity of the points in the coordinate system to generate a discrete point cloud dataset of pedestrian flow with a unified timestamp. A two-dimensional plane coordinate system is established with the center of the electronic screen display area as the origin. The calibrated ambient light intensity value at the current moment is obtained and used as the reference amplitude at the spatial origin. Based on a preset spatial attenuation coefficient, the relative light intensity value of any point on the screen is calculated according to a Gaussian distribution model based on the Euclidean distance between the point and the origin, causing the light intensity value to attenuate with increasing distance, thus generating an initial spatial distribution of light intensity. A preset time modulation factor is further introduced to modulate the initial distribution, generating a coverage... The entire screen area has a spatiotemporally continuous light intensity field. The time modulation factor dynamically adjusts its modulation frequency according to the local time. For example, a higher frequency periodic modulation is used during the daytime to simulate natural light changes, such as from 06:00 to 18:00, and a lower frequency periodic modulation is used during the nighttime, such as from 18:00 to 06:00 the next day. Discrete pedestrian point cloud data is converted into a continuous pedestrian density field. The conversion process uses the location coordinates of each pedestrian as the center point of the density distribution and employs the Epanechnikov kernel function model to diffuse the density contribution of each pedestrian from its location to the surrounding space. The bandwidth parameter of the kernel function is configurable and can be preset according to different application scenarios. For example, a larger bandwidth can be set in a commercial center scenario to obtain a smoother density distribution, while a smaller bandwidth can be set in a transportation hub scenario to retain more refined density variation features. At the same time, a weighting coefficient related to movement speed is introduced for the density contribution of each pedestrian. As an optional implementation, the weighting coefficient for the stationary state can be set to 1.0, the weighting coefficient for the walking state can be set to 0.8, and the weighting coefficient for the running state can be set to 0.6. By weighted superposition of the density contributions of all pedestrians, a spatiotemporally continuous pedestrian density field is generated.

[0022] The present invention is further configured such that S2 includes: Within the preset spatial integration domain, tensor product operations are performed on the light intensity field and the crowd density field. A Gaussian kernel function based on Euclidean distance is used to perform spatial convolution calculation on the tensor product operation results. The coupling field strength is obtained by double integration of the convolution result within the integration domain. Here, the Euclidean distance is the straight-line distance between the current calculation point and the integration variable point. The Laplacian operator for calculating the coupling field strength is used to detect the location of singular points in the field distribution by solving the absolute value of the Laplacian operator. These singular points correspond to the extreme regions of the visual attention distribution. The gradient magnitude of the crowd density field is calculated, and after a nonlinear transformation, it is multiplied by the absolute value of the Laplacian operator to generate a visual saliency field. Specifically, time-stamped light intensity field data and crowd density field data are read from a memory buffer, both stored in matrix form. Each matrix element stores the field intensity value at the corresponding grid cell, and the matrix dimensions are consistent with the preset spatial grid division. Element-wise tensor products are performed on the light intensity field matrix and the crowd density field matrix, and the light intensity value at each location is multiplied by the crowd density value. The values ​​are multiplied to generate an interaction intensity matrix. Each element of the interaction intensity matrix represents the nonlinear coupling effect between lighting conditions and pedestrian density at the corresponding spatial location. For each target grid point in the interaction intensity matrix, all source grid points within a preset spatial integration domain are determined with that target grid point as the center. The Euclidean distance between the target grid point and each source grid point within the preset spatial integration domain is calculated. Based on the Gaussian decay function, the Gaussian weighting coefficient corresponding to each Euclidean distance is calculated. The Gaussian weighting coefficient decreases exponentially with the increase of the Euclidean distance. The source grid points in the interaction intensity matrix are then... The interaction intensity value of each grid point is multiplied by its corresponding Gaussian weight coefficient and then subjected to double discrete integration in the integration domain. The integral result is used as the coupling field intensity value of the target point, ultimately generating the coupling field intensity matrix. The coupling field intensity matrix is ​​then discretized using the Laplacian operator. For each grid point, its second-order central difference in the orthogonal direction is calculated. The sum of these differences yields the Laplacian operator value for that grid point. The absolute value of each grid point's Laplacian operator value is used to highlight extreme regions in the field distribution, which correspond to hotspots where the visual attention of the crowd naturally focuses. Location; calculate the first-order partial derivatives of the crowd density field matrix in the X and Y directions, calculate the gradient magnitude of each grid point based on the obtained first-order partial derivatives, apply a nonlinear gain function, such as square root transformation, to the gradient magnitude to enhance the sensitivity of small gradient values ​​while suppressing the dominant effect of large gradient values; multiply the absolute value of the coupled field strength Laplace operator with the nonlinear transformation result of the crowd density field gradient magnitude element-wise to generate a visual saliency field, in which high-value regions characterize the visual attention focus that simultaneously satisfies the field structure singularity and the dynamic saliency of the crowd.

[0023] The present invention is further configured such that S3 includes: An energy functional for screen brightness distribution is constructed, which is obtained by a weighted combination of brightness matching term, spatial smoothing term and visual saliency enhancement term; The brightness matching term is constructed based on the deviation between the actual brightness distribution and the ideal brightness distribution based on the coupling field strength; The spatial smoothing term is constructed based on the spatial gradient variation amplitude of the actual brightness distribution; The visual saliency enhancement term is constructed based on the nonlinear mapping relationship between actual brightness and the visual saliency field, and the mapping relationship is realized through an exponential decay function model; The Euler-Lagrange variational method is used to solve the extremum problem of the energy functional, and the partial differential equation describing the optimal brightness distribution is derived. By solving the partial differential equation, the optimal brightness distribution field that minimizes the energy functional is obtained. Specifically, the coupled field strength data is mapped to the ideal brightness distribution, and the mapping relationship between the coupled field strength and the ideal brightness is established. When the coupled field strength is in the preset minimum range, the corresponding ideal brightness is 30% of the maximum brightness; when the coupled field strength is in the preset maximum range, the corresponding ideal brightness is 85% of the maximum brightness. An ideal brightness reference field covering the entire screen area is generated using piecewise linear interpolation. The difference between the actual brightness distribution and the ideal brightness reference field is calculated pixel-by-pixel. This difference is squared, summed, and then multiplied by a preset brightness matching weight coefficient to obtain a brightness matching term. The smaller the value of this brightness matching term, the closer the actual brightness distribution is to the ideal state. The spatial gradient of the brightness distribution is calculated using the central difference method. For each pixel, the sum of the squares of the brightness differences between it and its four adjacent pixels (up, down, left, and right) is calculated and then multiplied by a preset spatial smoothing weight coefficient to obtain a spatial smoothing term. The smaller the value of this spatial smoothing term, the smoother the brightness transition. A response function between brightness and visual saliency field is established. At low brightness values, the enhancement effect of the visual saliency field is linearly positively correlated with brightness. At high brightness values, the enhancement effect of the visual saliency field tends to saturate. The visual saliency field data is input into this response function, and the output is multiplied by a preset visual saliency weight coefficient to obtain a visual saliency enhancement term. The brightness matching term, spatial smoothing term, and visual saliency enhancement term are added together to obtain... The energy functional value is minimized to optimize the brightness distribution. The Euler-Lagrange variational method is used to derive the mathematical conditions that the optimal brightness distribution must satisfy. The energy functional is then variationally applied to the brightness distribution function to obtain the corresponding partial differential equation. This partial differential equation includes linear terms generated by the brightness matching term, Laplace terms generated by the spatial smoothing term, and nonlinear exponential terms generated by the visual saliency term. The partial differential equation is solved by first discretizing the screen display area into a computational domain composed of a predetermined number of grid points. The finite difference method is used to transform the partial differential equation to be solved into an algebraic equation at each grid point, forming a large system of linear algebraic equations. An iterative method is used for numerical solution. The system of linear algebraic equations is established and solved to obtain initial values ​​for the iteration. Then, the influence of nonlinear terms is introduced to construct and solve a modified system of linear algebraic equations to obtain an updated solution. This iterative process is repeated, with the relative error between two adjacent iterations being less than a predetermined threshold as the convergence criterion. The final iterative solution is the optimal brightness distribution field.

[0024] The present invention is further configured such that S4 includes: The depth of the information flow potential trap is obtained by performing an exponential transformation on the integral result of the visual saliency field in the preset spatial integration domain and taking the negative logarithm. Based on the information flow potential well depth, the time rate of change of the information flow potential well depth, and the Laplace operator of the crowd density field, a field feature vector is constructed. The weighted Euclidean distance between predefined content feature vectors and field feature vectors is calculated, and an exponential mapping is performed on this weighted Euclidean distance to obtain the content matching degree matrix. Specifically, the visual saliency field data is spatially integrated within a preset spatial integration domain, and Gaussian quadrature is used to ensure numerical accuracy. The integration result reflects the global intensity of visual attention within the screen area at the current moment. The above integration result is subjected to exponential transformation processing, and the transformation coefficient is controlled by the information temperature parameter. The information temperature parameter is dynamically adjusted by the environmental information entropy. When in a high information entropy environment, i.e., when the environmental information complexity is high, the information temperature parameter value is increased to make the transformed numerical distribution smoother. When in a low information entropy environment, i.e., when the environmental information is relatively simple, the information temperature parameter is decreased to enhance the numerical discrimination. Finally, the negative logarithm of the exponential transformation result is taken to obtain the information flow potential depth. This information flow potential depth is positively correlated with the information delivery capacity. A high information flow potential depth value indicates that the environment is suitable for delivering high information density content. Field features are extracted from three independent dimensions. Dimension one obtains the information flow potential depth at the current moment. Dimension 2 obtains the rate of change of the information flow potential well depth, calculated using the first-order difference of a sliding window of preset length. Dimension 3 obtains the Laplace operator value of the pedestrian density field. The three field feature values ​​are linearly normalized to the [0,1] interval and weighted according to preset weight coefficients to generate a three-dimensional field feature vector. The weight coefficients are set according to the scene type. For example, in a commercial display scene, the information flow potential well depth has the highest weight, while in a traffic guidance scene, the rate of change of the information flow potential well depth has the highest weight. The predefined content feature vectors in the content feature library are converted into three-dimensional vectors through a preset projection matrix to make them consistent with the dimensions of the field feature vectors. The projection matrix is ​​obtained by principal component analysis training on historical data to ensure that more than 90% of the original feature information is retained after projection. The weighted Euclidean distance between the projected content feature vector and the field feature vector is calculated, and the weighted Euclidean distance is subjected to exponential mapping to obtain the matching degree score of each content item in the content feature vector with the current feature scene. The content matching degree matrix is ​​obtained based on the matching degree scores of all content items.

[0025] The present invention is further configured such that S5 includes: Construct a collaborative dynamic system, which includes a brightness adjustment dynamic equation and a content matching degree evolution dynamic equation; The brightness adjustment dynamic equation defines a negative correlation between the rate of change of brightness over time and the energy functional gradient corresponding to the optimal brightness distribution location, so as to ensure a dynamic balance between brightness adjustment and energy minimization. The content matching degree evolution dynamic equation defines the functional relationship between the rate of change of content matching degree over time and the current state of content matching degree and the optimal brightness distribution field, so as to dynamically adjust the coupling between content and brightness; The steady-state solution of the cooperative dynamics system is determined based on the Lyapunov function, which is the sum of the weighted norms of the energy functional and the deviation between the content matching degree and the ideal matching degree. By solving for the minimum value of the Lyapunov function, the optimal steady-state solution of the co-dynamic system is obtained. This optimal steady-state solution corresponds to the final brightness adjustment command and content selection command to control the display of the electronic screen. Specifically, a co-dynamic system is constructed based on the optimal brightness distribution field and the content matching degree matrix. The brightness adjustment dynamic equation ensures that the brightness distribution dynamically converges to an energy-saving and comfortable state by keeping the brightness change rate negatively correlated with the energy functional gradient. The preset brightness adjustment rate parameter controls the response agility of the co-dynamic system. The brightness adjustment rate parameter is set according to the screen type; for example, the brightness adjustment rate parameter is 0.6 for LCD screens and 0.8 for LED screens. The content matching degree evolution dynamic equation describes the natural correlation between content through the state transition matrix obtained by statistical calculation using a Markov chain model based on historical content playback sequence data, utilizing brightness-content coupling. The function incorporates the characteristics of the optimal brightness distribution field into the content recommendation process. To ensure the global stability of the co-dynamic system, a Lyapunov function is constructed, consisting of a brightness energy functional and a weighted norm of the content matching degree deviation. The weight coefficients are used to balance the relative importance of the two optimization objectives and are set through expert experience. Finally, a numerical iterative algorithm is used to solve for the minimum value of the function. Taking the current optimal brightness field and content matching degree matrix as the initial state, the brightness distribution and content matching degree are updated in parallel using the gradient descent method. The brightness update step size is controlled by the brightness adjustment rate parameter, and the content matching degree update considers both state transition and brightness field driving effects. When the relative change of the Lyapunov function value is less than a preset threshold, the co-dynamic system is determined to have reached a stable state. The corresponding brightness distribution and the content with the highest matching degree at this time are output as the final brightness adjustment command and content selection command, respectively, thus completing the co-control of the electronic screen display.

[0026] The invention is further configured such that the brightness adjustment command and content selection command are sent to the control system of the electronic screen via wireless communication; specifically, the generated brightness control command and content recommendation command are transmitted to the local controller of the electronic screen via a wireless communication link; the communication process is configured with protocol parameters according to the deployment scenario, such as prioritizing low-power Bluetooth or Zigbee protocols in indoor short-range scenarios, and using Wi-Fi or cellular network technologies in outdoor wide-area scenarios, and ensuring reliable delivery of commands through data frame verification and retransmission mechanisms; after parsing the command, the electronic screen controller adjusts the backlight brightness to the target value through a pulse width modulation circuit, and simultaneously retrieves the specified content from the memory for rendering and display, thereby achieving coordinated control of screen brightness and playback content; for example, in a commercial center environment, the system can issue commands in real time to increase the screen brightness to 500 nits and switch to playing promotional advertising content.

[0027] The present invention is further configured such that the ambient light sensor and the people flow monitoring sensor transmit data wirelessly. Specifically, the ambient light sensor and the people flow monitoring sensor transmit the collected data to the control system of the electronic screen in real time via a wireless communication protocol. The ambient light intensity data collected by the ambient light sensor and the people flow point cloud data collected by the people flow monitoring sensor are transmitted via protocols such as Bluetooth, Zigbee, Wi-Fi, or cellular networks. The appropriate protocol is selected based on the application scenario and signal coverage range. For example, in indoor scenarios, the low-power Bluetooth or Zigbee protocol is preferred, while in outdoor scenarios, Wi-Fi or cellular network protocols may be selected to ensure the stability of data transmission and wide-area coverage. This method enables the control system to receive and process ambient light and people flow density data in real time, thereby dynamically adjusting the brightness and content display of the electronic screen, thus improving the display effect and optimizing the user experience.

[0028] The present invention is further configured such that the ambient light intensity data is synchronously collected by ambient light sensors distributed at multiple preset locations around the electronic screen, and the collected data is fused to obtain the data. Specifically, ambient light sensors are arranged at multiple preset locations around the electronic screen, and these sensors synchronously collect ambient light intensity data. The collection range and angle of each sensor are optimized to ensure that light intensity information from different directions can be fully acquired. All collected light intensity data are transmitted to the control system for centralized processing via a wireless communication protocol. The control system fuses the ambient light intensity data from different locations, uses weighted averaging or other data processing methods, and comprehensively considers the collection accuracy and geographical location of different sensors to obtain a global ambient light intensity value. This fusion processing result is used to accurately reflect the overall illumination conditions of the area where the electronic screen is located.

[0029] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens, characterized in that, include: S1: Real-time ambient light intensity data is collected through an ambient light sensor, and point cloud data of pedestrian flow covering the monitoring area in front of the electronic screen is collected through a pedestrian flow monitoring sensor. S2: Based on ambient light intensity data and pedestrian flow point cloud data, construct a spatiotemporally continuous light intensity field and pedestrian flow density field respectively; S3: Couple the light intensity field and the crowd density field to obtain the coupled field strength, and calculate the visual saliency field based on the coupled field strength; S4: Construct an energy functional of the screen brightness distribution based on the coupling field strength and visual saliency field, and obtain the optimal brightness distribution field by solving the Euler-Lagrange equation of the energy functional; S5: Calculate the information flow potential trap depth based on the visual saliency field, and combine it with the characteristics of the crowd density field to obtain the content matching degree matrix; S6: Based on the optimal brightness distribution field and content matching degree, a collaborative dynamic system is established, and brightness adjustment instructions and content selection instructions are obtained through Lyapunov stability analysis.

2. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens according to claim 1, characterized in that, S1 includes: Using the center of the electronic screen display plane as the spatial origin, and based on the ambient light intensity data collected by the ambient light sensor, a Gaussian distribution model is used to construct a spatiotemporally continuous light intensity field in which the ambient light intensity decays from the display plane to the surrounding space. Based on the pedestrian point cloud data collected by pedestrian monitoring sensors, the discrete pedestrian point cloud data is transformed into a spatiotemporally continuous pedestrian density field using a kernel density estimation algorithm, and a corresponding weighting coefficient is preset according to the movement speed of each data point.

3. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens according to claim 1, characterized in that, S2 includes: Within the preset spatial integration domain, tensor product operations are performed on the light intensity field and the crowd density field. A Gaussian kernel function based on Euclidean distance is used to perform spatial convolution calculation on the tensor product operation results. The coupling field strength is obtained by double integration of the convolution result within the integration domain. Here, the Euclidean distance is the straight-line distance between the current calculation point and the integration variable point. The Laplacian operator for calculating the coupling field strength is used to detect the location of singular points in the field distribution by solving the absolute value of the Laplacian operator. These singular points correspond to the extreme regions of the visual attention distribution. The gradient magnitude of the crowd density field is calculated, and after a nonlinear transformation, it is multiplied by the absolute value of the Laplace operator to generate the visual saliency field.

4. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens according to claim 1, characterized in that, S3 includes: An energy functional for screen brightness distribution is constructed, which is obtained by a weighted combination of brightness matching term, spatial smoothing term and visual saliency enhancement term; The brightness matching term is constructed based on the deviation between the actual brightness distribution and the ideal brightness distribution based on the coupling field strength; The spatial smoothing term is constructed based on the spatial gradient variation amplitude of the actual brightness distribution; The visual saliency enhancement term is constructed based on the nonlinear mapping relationship between actual brightness and the visual saliency field, and the mapping relationship is realized through an exponential decay function model; The Euler-Lagrange variational method is used to solve the extremum problem of the energy functional, and the partial differential equation describing the optimal brightness distribution is derived. By solving the partial differential equation, the optimal brightness distribution field that minimizes the energy functional is obtained.

5. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens according to claim 1, characterized in that, S4 includes: The depth of the information flow potential trap is obtained by performing an exponential transformation on the integral result of the visual saliency field in the preset spatial integration domain and taking the negative logarithm. Based on the information flow potential well depth, the time rate of change of the information flow potential well depth, and the Laplace operator of the crowd density field, a field feature vector is constructed. Calculate the weighted Euclidean distance between the predefined content feature vector and the field feature vector, and perform an exponential mapping on the weighted Euclidean distance to obtain the content matching degree matrix.

6. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens according to claim 1, characterized in that, S5 includes: Construct a collaborative dynamic system, which includes a brightness adjustment dynamic equation and a content matching degree evolution dynamic equation; The brightness adjustment dynamic equation defines a negative correlation between the rate of change of brightness over time and the energy functional gradient corresponding to the optimal brightness distribution location, so as to ensure a dynamic balance between brightness adjustment and energy minimization. The content matching degree evolution dynamic equation defines the functional relationship between the rate of change of content matching degree over time and the current state of content matching degree and the optimal brightness distribution field, so as to dynamically adjust the coupling between content and brightness; The steady-state solution of the cooperative dynamics system is determined based on the Lyapunov function, which is the sum of the weighted norms of the energy functional and the deviation between the content matching degree and the ideal matching degree. By solving for the minimum value of the Lyapunov function, the optimal steady-state solution of the cooperative dynamics system is obtained. This optimal steady-state solution corresponds to the final brightness adjustment command and content selection command to control the display of the electronic screen.

7. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of an electronic screen according to claim 6, characterized in that, The brightness adjustment command and content selection command are sent to the electronic screen's control system via wireless communication.

8. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of electronic screens according to claim 1, characterized in that, The ambient light sensor and the pedestrian flow monitoring sensor transmit data wirelessly.

9. The method for dimming and content recommendation based on ambient light and pedestrian flow perception of an electronic screen according to claim 1, characterized in that, The ambient light intensity data is collected synchronously by ambient light sensors distributed at multiple preset locations around the electronic screen, and the collected data is then fused.

Images