A control method and system for outdoor LED display screen

Abstract

The present application provides a control method and system for an outdoor LED display screen, which relates to the field of data processing technology. The method includes: obtaining the initial display parameters of the outdoor LED display screen and the image to be displayed; controlling the outdoor LED display screen to display the image according to the initial display parameters; photographing the outdoor LED display screen through an image acquisition device to obtain the display image of the outdoor LED display screen; applying a Gaussian mixture model to process the display image to obtain the ambient light image and foreground image corresponding to the display image; inputting the ambient light image, foreground image, initial display parameters and the current environmental parameters of the outdoor LED display screen into a display parameter prediction model to obtain the prediction results output by the display parameter prediction model, the prediction results including the target display parameters; controlling the outdoor LED display screen to display according to the target display parameters. The display parameters of the outdoor LED display screen can be dynamically adjusted to enhance the user viewing experience.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a control method and system for an outdoor LED display screen. Background Art

[0002] With the development of LED display technology, it has been widely used in advertising, information release, public display and other fields.

[0003] However, the display parameters (such as brightness, contrast, etc.) of most outdoor LED displays on the market are usually fixed or can only be changed by manual adjustment. This means that the display effect of the display is often unsatisfactory under different lighting conditions. For example, under strong sunlight, the display may appear too dim, while at night or in a darker indoor environment, the display may appear too dazzling. This not only affects the user's viewing experience, but may also cause eye fatigue. Summary of the invention

[0004] The technical problem to be solved by this application is to provide a control method and system for an outdoor LED display screen, which can dynamically adjust the display parameters of the outdoor LED display screen. The specific solution is as follows:

[0005] A control method for an outdoor LED display screen, comprising:

[0006] In response to the display screen adjustment instruction, obtaining initial display parameters of the outdoor LED display screen and an image to be displayed;

[0007] Controlling the outdoor LED display screen to display the image according to the initial display parameters;

[0008] The outdoor LED display screen is photographed by an image acquisition device to obtain a display image of the outdoor LED display screen;

[0009] Applying a pre-trained Gaussian mixture model to process the display image to obtain an ambient light image and a foreground image corresponding to the display image;

[0010] Inputting the ambient light image, the foreground image, the initial display parameters, and the current environmental parameters of the outdoor LED display screen into a pre-trained display parameter prediction model to obtain a prediction result output by the display parameter prediction model, wherein the prediction result includes a target display parameter;

[0011] The outdoor LED display screen is controlled to display according to the target display parameters.

[0012] In the above method, optionally, applying a preset Gaussian mixture model to process the display image to obtain an ambient light image and a foreground image corresponding to the display image includes:

[0013] Use the pre-trained Gaussian mixture model to classify each pixel in the displayed image;

[0014] Combining the pixels classified as background in the displayed image to form an ambient light image;

[0015] Pixels in the displayed image that are not classified as background are combined to form a foreground image.

[0016] In the above method, optionally, the process of training the display parameter prediction model includes:

[0017] Acquire a training data set and an initial display parameter prediction model to be trained; the training data set includes multiple training samples and a sample label corresponding to each training sample; each training sample includes a historical foreground image, a historical ambient light image, a historical display parameter and an environmental parameter of an outdoor LED display screen, and the sample label of each training sample includes a display parameter;

[0018] Selecting a target training sample currently used for training from the training data set;

[0019] Inputting the target training sample into the initial display parameter prediction model to obtain a recognition result corresponding to the target training sample;

[0020] Based on a preset loss function, the recognition result corresponding to the target training sample and the sample label, a loss function value of the initial display parameter prediction model is calculated; the loss function value represents the degree of difference between the recognition result corresponding to the target training sample and the sample label;

[0021] Update the model parameters of the initial display parameter prediction model according to the loss function value;

[0022] If the initial display parameter prediction model after updating the model parameters does not meet the preset training completion condition, returning to the step of selecting a target training sample currently used for training in the training data set;

[0023] When the initial display parameter prediction model after the model parameters are updated satisfies a preset training completion condition, the initial display parameter prediction model satisfying the training completion condition is determined as a trained display parameter prediction model.

[0024] The above method, optionally, after controlling the outdoor LED display screen to display according to the target display parameters, further comprises:

[0025] Receive the human eye test results of the outdoor LED display screen by the test subject;

[0026] If the human eye test result does not meet the preset application conditions, output prompt information, where the prompt information is used to prompt the test subject to feedback the required display parameters;

[0027] receiving display parameters input by the test subject;

[0028] Generate a loss function value corresponding to the prediction result according to the loss function, the prediction result and the display parameter input by the test subject;

[0029] The parameters of the display parameter prediction model are updated using the loss function value corresponding to the prediction result.

[0030] The above method, optionally, after controlling the outdoor LED display screen to display according to the target display parameters, further comprises:

[0031] The target display parameters are stored in a preset data table, and the data table is detected to determine whether there are historical target display parameters within a preset time period;

[0032] If there is a historical target display parameter within a preset time period, determining the difference between the target display parameter and the historical target display parameter;

[0033] When the difference is greater than a preset difference threshold, hardware fault prompt information of the outdoor LED display screen is generated and the hardware fault prompt information is output.

[0034] A control system for an outdoor LED display screen, comprising:

[0035] An acquisition unit, used for acquiring initial display parameters of the outdoor LED display screen and an image to be displayed in response to a display screen adjustment instruction;

[0036] A first control unit, used to control the outdoor LED display screen to display the image according to the initial display parameters;

[0037] A shooting unit, used to shoot the outdoor LED display screen through an image acquisition device to obtain a display image of the outdoor LED display screen;

[0038] A processing unit, used to process the display image using a pre-trained Gaussian mixture model to obtain an ambient light image and a foreground image corresponding to the display image;

[0039] An execution unit, used for inputting the ambient light image, the foreground image, the initial display parameters and the current environmental parameters of the outdoor LED display screen into a pre-trained display parameter prediction model to obtain a prediction result output by the display parameter prediction model, wherein the prediction result includes a target display parameter;

[0040] The second control unit is used to control the outdoor LED display screen to display according to the target display parameters.

[0041] In the above system, optionally, the processing unit includes:

[0042] A classification unit, used for classifying each pixel in the displayed image using a pre-trained Gaussian mixture model;

[0043] A first processing subunit, configured to combine pixels classified as background in the display image to form an ambient light image;

[0044] The second processing subunit is used to combine the pixel points in the display image that are not classified as background to form a foreground image.

[0045] In the above system, optionally, the execution unit includes:

[0046] An acquisition subunit is used to acquire a training data set and an initial display parameter prediction model to be trained; the training data set includes a plurality of training samples and a sample label corresponding to each training sample; each training sample includes a historical foreground image, a historical ambient light image, a historical display parameter and an environmental parameter of an outdoor LED display screen, and the sample label of each training sample includes a display parameter;

[0047] A selection subunit is used to select a target training sample currently used for training from the training data set;

[0048] A first execution subunit, configured to input the target training sample into the initial display parameter prediction model to obtain a recognition result corresponding to the target training sample;

[0049] A calculation subunit, configured to calculate a loss function value of the initial display parameter prediction model based on a preset loss function, the recognition result corresponding to the target training sample, and the sample label; the loss function value represents the degree of difference between the recognition result corresponding to the target training sample and the sample label;

[0050] An updating subunit, used for updating the model parameters of the initial display parameter prediction model according to the loss function value;

[0051] A second execution subunit is used to return to trigger the selection subunit to re-execute the selection of target training samples currently used for training in the training data set when the initial display parameter prediction model after updating the model parameters does not meet the preset training completion condition;

[0052] The determination subunit is used to determine the initial display parameter prediction model that meets the preset training completion condition as the trained display parameter prediction model when the initial display parameter prediction model after the model parameters are updated meets the preset training completion condition.

[0053] The above system may optionally further include:

[0054] A first receiving unit is used to receive a human eye test result of a test subject on the outdoor LED display screen;

[0055] A first output unit, configured to output prompt information when the human eye test result does not meet a preset application condition, wherein the prompt information is used to prompt the test subject to feedback the display parameters required;

[0056] A second receiving unit, configured to receive display parameters input by the test subject;

[0057] A generating unit, configured to generate a loss function value corresponding to the prediction result according to the loss function, the prediction result and a display parameter input by the test subject;

[0058] An updating unit is used to update the parameters of the display parameter prediction model using the loss function value corresponding to the prediction result.

[0059] The above system may optionally further include:

[0060] A detection unit, used to store the target display parameters in a preset data table, and detect whether there are historical target display parameters within a preset time period in the data table;

[0061] a determination unit, configured to determine a difference between the target display parameter and the historical target display parameter if there is a historical target display parameter within a preset time period;

[0062] The second output unit is used to generate hardware fault prompt information of the outdoor LED display screen when the difference is greater than a preset difference threshold, and output the hardware fault prompt information.

[0063] Based on the above-mentioned implementation of the present application, a control method and system for an outdoor LED display screen is provided, the method comprising: in response to a display screen adjustment instruction, obtaining the initial display parameters of the outdoor LED display screen and the image to be displayed; controlling the outdoor LED display screen to display the image according to the initial display parameters; photographing the outdoor LED display screen through an image acquisition device to obtain the display image of the outdoor LED display screen; applying a pre-trained Gaussian mixture model to process the display image to obtain an ambient light image and a foreground image corresponding to the display image; inputting the ambient light image, the foreground image, the initial display parameters and the current environmental parameters of the outdoor LED display screen into a pre-trained display parameter prediction model to obtain the prediction results output by the display parameter prediction model, the prediction results including the target display parameters; controlling the outdoor LED display screen to display according to the target display parameters. The method provided by the embodiment of the present application is applied, by comprehensively utilizing the ambient light image, the display screen content image, the display parameters and the environmental parameters, combined with the machine learning model for prediction and optimization, to achieve dynamic adaptation to environmental changes, so that the display screen can provide the best viewing experience under any environmental conditions, improve user satisfaction, and have higher intelligence and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0065] Figure 1 A method flow chart of a control method for an outdoor LED display screen provided in this application;

[0066] Figure 2 An example diagram of a process for obtaining an ambient light image and a foreground image corresponding to a display image provided by the present application;

[0067] Figure 3 A flowchart of a process of training a display parameter prediction model provided in the present application;

[0068] Figure 4 A schematic diagram of the structure of a control system for an outdoor LED display screen provided in this application. DETAILED DESCRIPTION

[0069] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0070] In this application, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0071] Based on this, the embodiment of the present application provides a control method for an outdoor LED display screen, which can be applied to electronic devices, such as computers, controllers of LED display screens, tablets, smart phones, or smart wearable devices. The method flow chart of the method is as follows: Figure 1 As shown, specifically including:

[0072] S101: In response to a display screen adjustment instruction, initial display parameters of the outdoor LED display screen and an image to be displayed are acquired.

[0073] In this embodiment, the display adjustment instruction can be an instruction that is automatically triggered at a preset time interval, or it can be an instruction triggered by a user clicking a preset virtual control or button, or it can be an instruction triggered when receiving target experience information fed back by the user. The target experience information can indicate that the display experience score of the outdoor LED display is lower than a preset experience score threshold.

[0074] In this embodiment, the initial display parameters may be default values ​​or current display parameters of the LED display screen.

[0075] Optionally, the initial display parameters of the LED display screen may include one or more of initial brightness, contrast, color saturation and color temperature.

[0076] In this embodiment, the image to be displayed may be a test image, a promotional image, an advertisement image, a video frame, etc. that needs to be displayed.

[0077] S102: Control the outdoor LED display screen to display the image according to the initial display parameters.

[0078] In this embodiment, the outdoor LED display screen is controlled to display the image to be displayed under the initial display parameters.

[0079] S103: photographing the outdoor LED display screen by using an image acquisition device to obtain a display image of the outdoor LED display screen.

[0080] In this embodiment, the image acquisition device can be fixed in front of the outdoor LED display screen, and the field of view of the image acquisition device includes the area where the outdoor LED display screen is located.

[0081] Optionally, the displayed image includes at least the content displayed by the outdoor LED display screen and the surrounding environment information of the outdoor LED display screen.

[0082] S104: Apply a pre-trained Gaussian mixture model to process the display image to obtain an ambient light image and a foreground image corresponding to the display image.

[0083] In this embodiment, scene images including the outdoor LED display screen can be collected in advance, and then when the outdoor LED display screen is turned off or displays a pure color image, ambient light sample images can be collected for background modeling; specifically, a series of ambient light sample images can be collected and a Gaussian mixture model GMM can be used to estimate the distribution of background light; then, the model parameters of the Gaussian mixture model can be adjusted according to the collected samples to make the Gaussian mixture model adapt to different lighting conditions.

[0084] In this embodiment, the training process of the Gaussian mixture model includes: obtaining the historical display image of the outdoor LED display screen, initializing one or more Gaussian distribution parameters (mean, variance, weight) for each pixel of the historical display image, and setting a learning rate α for updating the parameters of the Gaussian distribution; and setting a threshold τ for distinguishing the background from the foreground, thereby obtaining an initial Gaussian mixture model. Then, a video sequence of the outdoor LED display screen is collected by an image acquisition device for background learning; further, for each pixel, the value of the current frame (any frame in the video sequence) is used to update the parameters of the Gaussian distribution, and the updating process is as follows:

[0085] Mean update:

[0086] Variance Update:

[0087] Weight update: ;

[0088] Among them, I(t) represents the pixel value of the current frame, , and Respectively represent the mean, variance and weight of the i-th Gaussian distribution;

[0089] For each pixel, a best-matching Gaussian distribution is selected as the background estimate, resulting in a trained Gaussian mixture model.

[0090] Optionally, the number of frames of the video sequence used to update the parameters can be adjusted according to actual conditions to achieve the best background segmentation effect. For example, it can be set to 500 frames.

[0091] S105: Input the ambient light image, the foreground image, the initial display parameters and the current environmental parameters of the outdoor LED display screen into a pre-trained display parameter prediction model to obtain a prediction result output by the display parameter prediction model, wherein the prediction result includes a target display parameter.

[0092] In this embodiment, the environmental parameters may include data such as light intensity, temperature, humidity, etc. collected by the sensor collection component.

[0093] Optionally, the display parameter prediction model includes an input layer, a feature extraction layer, a fusion layer and an output layer; the feature extraction layer includes a first convolutional neural network, a second convolutional neural network, a first fully connected layer and a second fully connected layer; the first convolutional neural network is used to extract features of the ambient light image; the second convolutional neural network is used to extract features of the foreground image, the first fully connected layer is used to extract features of the initial display parameters, and the second fully connected layer is used to extract features of the environmental parameters.

[0094] In this embodiment, the features of the ambient light image may include object edge features, texture features, color features, local shapes of objects, local structures of objects, scene information in the ambient light image, color distribution of the scene in the ambient light image, light intensity distribution in the ambient light image, and environmental features (cloud, fog image features), etc.

[0095] In this embodiment, the features of the foreground image may include one or more of edge features, texture features, color features, local shapes of objects in the foreground image (corner points, curves), local structures of objects in the foreground image (texture blocks, patterns), overall shapes of objects in the foreground image, color distribution and semantic information of objects, etc.

[0096] S106: Control the outdoor LED display screen to display according to the target display parameters.

[0097] In this embodiment, the outdoor LED display screen can be controlled to display the image to be displayed according to the target display parameters, and can also display other subsequent images to be displayed based on the target display parameters until new target display parameters are obtained.

[0098] By applying the method provided in the embodiment of the present application, by comprehensively utilizing ambient light images, display content images, display parameters and environmental parameters, combined with machine learning models for prediction and optimization, dynamic adaptation to environmental changes is achieved, so that the display screen can provide the best viewing experience under any environmental conditions, improve user satisfaction, and have higher intelligence and adaptability.

[0099] In an embodiment provided in the present application, based on the above-mentioned solution, optionally, the process of applying a preset Gaussian mixture model to process the display image to obtain an ambient light image and a foreground image corresponding to the display image is as follows: Figure 2 As shown, specifically including:

[0100] S201: Classify each pixel in the displayed image using a pre-trained Gaussian mixture model.

[0101] In this embodiment, the process of classifying each pixel in the displayed image using the pre-trained Gaussian mixture model is as follows:

[0102] For each pixel of the displayed image, the likelihood of the pixel value and all Gaussian distributions corresponding to the pixel in the Gaussian mixture model is calculated; whether the likelihood of the pixel value and any Gaussian distribution is lower than a threshold value τ is determined to classify the pixel and obtain a classification result; specifically, if the likelihood of the pixel value and any Gaussian distribution is lower than the threshold value τ, the pixel is classified as foreground; otherwise, the pixel is classified as background.

[0103] S202: Combining the pixels classified as background in the display image to form an ambient light image.

[0104] S203: Combining the pixel points in the displayed image that are not classified as background to form a foreground image.

[0105] In some embodiments, a binary mask image can be generated based on the classification result of the displayed image, with foreground pixels marked as white (or 255) and background pixels marked as black (or 0); a logical AND operation is performed between the foreground mask and the displayed image to extract the background image, and the background image is used as the ambient light image; a logical OR operation is performed between the foreground mask and the displayed image to extract the foreground image.

[0106] In an embodiment provided in the present application, based on the above solution, optionally, the process of training the display parameter prediction model is as follows: Figure 3 As shown, including:

[0107] S301: Obtain a training data set and an initial display parameter prediction model to be trained; the training data set includes multiple training samples and a sample label corresponding to each training sample; each training sample includes a historical foreground image, a historical ambient light image, a historical display parameter and an environmental parameter of an outdoor LED display screen, and the sample label of each training sample includes a display parameter.

[0108] In this embodiment, the sample labels corresponding to the training samples are obtained by annotating by an expert system or manually annotating by an expert.

[0109] Optionally, the display parameters in the sample tag include at least one of brightness, contrast, color saturation and color temperature.

[0110] S302: Select target training samples currently used for training from the training data set.

[0111] In this embodiment, any n samples can be selected from the training samples that have not participated in the training in the training data set as the target training samples currently used for training.

[0112] S303: Input the target training sample into the initial display parameter prediction model to obtain a recognition result corresponding to the target training sample.

[0113] In this embodiment, the recognition result corresponding to the target training sample may include the display parameters of the outdoor LED display screen output by the initial display parameter prediction model.

[0114] S304: Based on a preset loss function, the recognition results corresponding to the target training samples and the sample labels, the loss function value of the initial display parameter prediction model is calculated; the loss function value represents the degree of difference between the recognition results corresponding to the target training samples and the sample labels.

[0115] In this embodiment, the loss function may be a mean square error loss function.

[0116] The process of calculating the loss function value is as follows:

[0117]

[0118] Where f(x;θ) represents the initial display parameter prediction model, θ represents the model parameters, and x represents the input variable. is the predicted value of the initial display parameter prediction model for the i-th target training sample; Represents the true label of the i-th sample; the loss function L(θ) represents the average of the squares of the differences between the predicted values ​​and the true values ​​of n samples.

[0119] S305: Update the model parameters of the initial display parameter prediction model according to the loss function value.

[0120] In this embodiment, the model parameters of the initial display parameter prediction model may be updated by a gradient descent method.

[0121] S306: Determine whether the initial display parameter prediction model after updating the model parameters meets the preset training completion conditions; if not, return to execute S302; if so, execute S307.

[0122] In this embodiment, by selecting n target training samples for training each time, memory usage can be effectively reduced. When using small batch gradient descent, the gradient of each parameter update is calculated based on a small number of samples, which can reduce the fluctuation of the gradient and make the training process more stable. A certain degree of randomness is introduced through group training, which helps the model escape from the local optimal solution and find the global optimal solution. It can also help the model better generalize to unseen data. By training on different sample groups, the model can learn a wider range of data distributions and reduce the risk of overfitting. And each sample group may contain different data distributions, which helps the model learn more diverse features and improve its generalization ability.

[0123] S307: Determine the initial display parameter prediction model that meets the training completion condition as the trained display parameter prediction model.

[0124] Optionally, the training completion conditions may include that the number of training times is greater than a preset training number threshold, the prediction accuracy of the initial display parameter prediction model is greater than a preset accuracy threshold, or the loss function value converges, etc.

[0125] In an embodiment provided in the present application, based on the above solution, optionally, after controlling the outdoor LED display screen to display according to the target display parameters, the method further includes:

[0126] Receive the human eye test results of the outdoor LED display screen by the test subject;

[0127] If the human eye test result does not meet the preset application conditions, output prompt information, where the prompt information is used to prompt the test subject to feedback the required display parameters;

[0128] receiving display parameters input by the test subject;

[0129] Generate a loss function value corresponding to the prediction result according to the loss function, the prediction result and the display parameter input by the test subject;

[0130] The parameters of the display parameter prediction model are updated using the loss function value corresponding to the prediction result.

[0131] In this embodiment, the human eye test results may include the test subject's clarity evaluation results, comfort evaluation results, color reproduction evaluation results and satisfaction evaluation results of the outdoor LED display screen; the clarity evaluation results, comfort evaluation results, color reproduction evaluation results and satisfaction evaluation results are weighted and summed based on preset weight values ​​to obtain an experience score; if the experience score is less than a preset score threshold, it is determined that the preset application conditions are not met; if the experience score is not less than the score threshold, it is determined that the application conditions are met.

[0132] Optionally, the display parameters of the test object input can be used as sample labels to generate loss function values.

[0133] In an embodiment provided in the present application, based on the above solution, optionally, after controlling the outdoor LED display screen to display according to the target display parameters, the method further includes:

[0134] The target display parameters are stored in a preset data table, and the data table is detected to determine whether there are historical target display parameters within a preset time period;

[0135] If there is a historical target display parameter within a preset time period, determining the difference between the target display parameter and the historical target display parameter;

[0136] When the difference is greater than a preset difference threshold, hardware fault prompt information of the outdoor LED display screen is generated and the hardware fault prompt information is output.

[0137] In this embodiment, the time period may be 30 seconds, 2 minutes, 5 minutes, or 10 minutes.

[0138] In this embodiment, the difference value may include at least one of an absolute value of a difference between the target display parameter and the historical target display parameter, and a Euclidean distance between a vector of the target display parameter and a vector of the historical target display parameter.

[0139] In some embodiments, the historical target display parameters and the target display parameters may be vectorized to obtain a historical target display parameter vector and a target display parameter vector; and then the Euclidean distance between the historical target display parameter vector and the target display parameter vector is calculated, as follows:

[0140]

[0141] Where E represents the Euclidean distance; Indicates the qth display parameter value of the historical target display parameter. Represents the qth display parameter value of the target display parameter, where n is the number of display parameters.

[0142] In this embodiment, if at least one of the brightness, color temperature and contrast changes significantly in a short period of time, it is judged that the hardware components (such as power supply, driver board, LED lamp beads, etc.) may fail, resulting in a sudden change in display parameters, or the ambient light sensor or other sensors may fail, resulting in inaccurate reading of environmental parameters, thereby affecting the adjustment of display parameters.

[0143] If multiple display parameters (such as brightness, color temperature, contrast, etc.) show abnormal changes at the same time, it indicates that the mainboard or processor of the control system is faulty, causing multiple parameters to be abnormal at the same time.

[0144] and Figure 1 Corresponding to the method described above, the embodiment of the present application also provides a control system for an outdoor LED display screen, which is used to Figure 1 The specific implementation of the method in the system is shown in the following figure. Figure 4 As shown, including:

[0145] The acquisition unit 401 is used to acquire the initial display parameters of the outdoor LED display screen and the image to be displayed in response to the display screen adjustment instruction;

[0146] A first control unit 402, used to control the outdoor LED display screen to display the image according to the initial display parameters;

[0147] The shooting unit 403 is used to shoot the outdoor LED display screen through an image acquisition device to obtain a display image of the outdoor LED display screen;

[0148] The processing unit 404 is used to process the display image using a pre-trained Gaussian mixture model to obtain an ambient light image and a foreground image corresponding to the display image;

[0149] An execution unit 405 is used to input the ambient light image, the foreground image, the initial display parameters and the current environmental parameters of the outdoor LED display screen into a pre-trained display parameter prediction model to obtain a prediction result output by the display parameter prediction model, wherein the prediction result includes a target display parameter;

[0150] The second control unit 406 is used to control the outdoor LED display screen to display according to the target display parameters.

[0151] In an embodiment provided in the present application, based on the above solution, optionally, the processing unit includes:

[0152] A classification unit, used for classifying each pixel in the displayed image using a pre-trained Gaussian mixture model;

[0153] A first processing subunit, configured to combine pixels classified as background in the display image to form an ambient light image;

[0154] The second processing subunit is used to combine the pixel points in the display image that are not classified as background to form a foreground image.

[0155] In an embodiment provided in the present application, based on the above solution, optionally, the execution unit includes:

[0156] An acquisition subunit is used to acquire a training data set and an initial display parameter prediction model to be trained; the training data set includes a plurality of training samples and a sample label corresponding to each training sample; each training sample includes a historical foreground image, a historical ambient light image, a historical display parameter and an environmental parameter of an outdoor LED display screen, and the sample label of each training sample includes a display parameter;

[0157] A selection subunit is used to select a target training sample currently used for training from the training data set;

[0158] A first execution subunit, configured to input the target training sample into the initial display parameter prediction model to obtain a recognition result corresponding to the target training sample;

[0159] A calculation subunit, configured to calculate a loss function value of the initial display parameter prediction model based on a preset loss function, the recognition result corresponding to the target training sample, and the sample label; the loss function value represents the degree of difference between the recognition result corresponding to the target training sample and the sample label;

[0160] An updating subunit, used for updating the model parameters of the initial display parameter prediction model according to the loss function value;

[0161] A second execution subunit is used to return to trigger the selection subunit to re-execute the selection of target training samples currently used for training in the training data set when the initial display parameter prediction model after updating the model parameters does not meet the preset training completion condition;

[0162] The determination subunit is used to determine the initial display parameter prediction model that meets the preset training completion condition as the trained display parameter prediction model when the initial display parameter prediction model after the model parameters are updated meets the preset training completion condition.

[0163] In an embodiment provided in the present application, based on the above solution, optionally, it further includes:

[0164] A first receiving unit is used to receive a human eye test result of a test subject on the outdoor LED display screen;

[0165] A first output unit, configured to output prompt information when the human eye test result does not meet a preset application condition, wherein the prompt information is used to prompt the test subject to feedback the display parameters required;

[0166] A second receiving unit, configured to receive display parameters input by the test subject;

[0167] A generating unit, configured to generate a loss function value corresponding to the prediction result according to the loss function, the prediction result and a display parameter input by the test subject;

[0168] An updating unit is used to update the parameters of the display parameter prediction model using the loss function value corresponding to the prediction result.

[0169] In an embodiment provided in the present application, based on the above solution, optionally, it further includes:

[0170] A detection unit, used to store the target display parameters in a preset data table, and detect whether there are historical target display parameters within a preset time period in the data table;

[0171] a determination unit, configured to determine a difference between the target display parameter and the historical target display parameter if there is a historical target display parameter within a preset time period;

[0172] The second output unit is used to generate hardware fault prompt information of the outdoor LED display screen when the difference is greater than a preset difference threshold, and output the hardware fault prompt information.

[0173] The specific principles and execution processes of each unit and module in the control system of an outdoor LED display screen disclosed in the above-mentioned embodiment of the present application are the same as the control method of an outdoor LED display screen disclosed in the above-mentioned embodiment of the present application. Please refer to the corresponding parts of the control method of an outdoor LED display screen provided in the above-mentioned embodiment of the present application, and no further details will be given here.

[0174] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0175] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

[0176] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0177] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0178] The above is a detailed introduction to a control method for an outdoor LED display screen provided by the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A control method for an outdoor LED display screen, characterized in that: include: In response to the display screen adjustment instruction, initial display parameters of the outdoor LED display screen and an image to be displayed are acquired; the initial display parameters include an initial contrast ratio and an initial color temperature; Controlling the outdoor LED display screen to display the image according to the initial display parameters; The outdoor LED display screen is photographed by an image acquisition device to obtain a display image of the outdoor LED display screen; Applying a pre-trained Gaussian mixture model to process the display image to obtain an ambient light image and a foreground image corresponding to the display image; The ambient light image, the foreground image, the initial display parameters and the current environmental parameters of the outdoor LED display screen are input into a pre-trained display parameter prediction model to obtain a prediction result output by the display parameter prediction model, wherein the prediction result includes a target display parameter; the display parameter prediction model includes an input layer, a feature extraction layer, a fusion layer and an output layer; the feature extraction layer includes a first convolutional neural network, a second convolutional neural network, a first fully connected layer and a second fully connected layer; the first convolutional neural network is used to extract features of the ambient light image; the second convolutional neural network is used to extract features of the foreground image, the first fully connected layer is used to extract features of the initial display parameters, and the second fully connected layer is used to extract features of the environmental parameters; the target display parameters include contrast and color temperature; Controlling the outdoor LED display screen to display according to the target display parameters; The target display parameters are stored in a preset data table, and the data table is detected to determine whether there are historical target display parameters within a preset time period; If there is a historical target display parameter within a preset time period, determining the difference between the target display parameter and the historical target display parameter; When the difference is greater than a preset difference threshold, hardware fault prompt information of the outdoor LED display screen is generated and the hardware fault prompt information is output.

2. The method according to claim 1, characterized in that The applying a preset Gaussian mixture model to process the display image to obtain an ambient light image and a foreground image corresponding to the display image includes: Use the pre-trained Gaussian mixture model to classify each pixel in the displayed image; Combining the pixels classified as background in the displayed image to form an ambient light image; Pixels in the displayed image that are not classified as background are combined to form a foreground image.

3. The method according to claim 1, characterized in that The process of training the display parameter prediction model includes: Acquire a training data set and an initial display parameter prediction model to be trained; the training data set includes multiple training samples and a sample label corresponding to each training sample; each training sample includes a historical foreground image, a historical ambient light image, a historical display parameter and an environmental parameter of an outdoor LED display screen, and the sample label of each training sample includes a display parameter; Selecting a target training sample currently used for training from the training data set; Inputting the target training sample into the initial display parameter prediction model to obtain a recognition result corresponding to the target training sample; Based on a preset loss function, the recognition result corresponding to the target training sample and the sample label, a loss function value of the initial display parameter prediction model is calculated; the loss function value represents the degree of difference between the recognition result corresponding to the target training sample and the sample label; Update the model parameters of the initial display parameter prediction model according to the loss function value; If the initial display parameter prediction model after updating the model parameters does not meet the preset training completion condition, returning to the step of selecting a target training sample currently used for training in the training data set; When the initial display parameter prediction model after the model parameters are updated satisfies a preset training completion condition, the initial display parameter prediction model satisfying the training completion condition is determined as a trained display parameter prediction model.

4. The method according to claim 3, characterized in that After controlling the outdoor LED display screen to display according to the target display parameters, the method further includes: Receive the human eye test results of the outdoor LED display screen by the test subject; If the human eye test result does not meet the preset application conditions, output prompt information, where the prompt information is used to prompt the test subject to feedback the required display parameters; receiving display parameters input by the test subject; Generate a loss function value corresponding to the prediction result according to the loss function, the prediction result and the display parameter input by the test subject; The parameters of the display parameter prediction model are updated using the loss function value corresponding to the prediction result.

5. A control system for an outdoor LED display screen, characterized in that: include: An acquisition unit, configured to acquire initial display parameters of the outdoor LED display screen and an image to be displayed in response to a display screen adjustment instruction; the initial display parameters include an initial contrast ratio and an initial color temperature; A first control unit, used to control the outdoor LED display screen to display the image according to the initial display parameters; A shooting unit, used to shoot the outdoor LED display screen through an image acquisition device to obtain a display image of the outdoor LED display screen; A processing unit, used to process the display image using a pre-trained Gaussian mixture model to obtain an ambient light image and a foreground image corresponding to the display image; An execution unit is used to input the ambient light image, the foreground image, the initial display parameters and the current environmental parameters of the outdoor LED display screen into a pre-trained display parameter prediction model to obtain a prediction result output by the display parameter prediction model, wherein the prediction result includes a target display parameter; the display parameter prediction model includes an input layer, a feature extraction layer, a fusion layer and an output layer; the feature extraction layer includes a first convolutional neural network, a second convolutional neural network, a first fully connected layer and a second fully connected layer; the first convolutional neural network is used to extract features of the ambient light image; the second convolutional neural network is used to extract features of the foreground image, the first fully connected layer is used to extract features of the initial display parameters, and the second fully connected layer is used to extract features of the environmental parameters; the target display parameters include contrast and color temperature; A second control unit, used to control the outdoor LED display screen to display according to the target display parameters; A detection unit, used to store the target display parameters in a preset data table, and detect whether there are historical target display parameters within a preset time period in the data table; a determination unit, configured to determine a difference between the target display parameter and the historical target display parameter if there is a historical target display parameter within a preset time period; The second output unit is used to generate hardware fault prompt information of the outdoor LED display screen when the difference is greater than a preset difference threshold, and output the hardware fault prompt information.

6. The system according to claim 5, characterized in that The processing unit comprises: A classification unit, used for classifying each pixel in the displayed image using a pre-trained Gaussian mixture model; A first processing subunit, configured to combine pixels classified as background in the display image to form an ambient light image; The second processing subunit is used to combine the pixel points in the display image that are not classified as background to form a foreground image.

7. The system according to claim 5, characterized in that The execution unit comprises: An acquisition subunit is used to acquire a training data set and an initial display parameter prediction model to be trained; the training data set includes a plurality of training samples and a sample label corresponding to each training sample; each training sample includes a historical foreground image, a historical ambient light image, a historical display parameter and an environmental parameter of an outdoor LED display screen, and the sample label of each training sample includes a display parameter; A selection subunit is used to select a target training sample currently used for training from the training data set; A first execution subunit, configured to input the target training sample into the initial display parameter prediction model to obtain a recognition result corresponding to the target training sample; A calculation subunit, configured to calculate a loss function value of the initial display parameter prediction model based on a preset loss function, the recognition result corresponding to the target training sample, and the sample label; the loss function value represents the degree of difference between the recognition result corresponding to the target training sample and the sample label; An updating subunit, used for updating the model parameters of the initial display parameter prediction model according to the loss function value; A second execution subunit is used to return to trigger the selection subunit to re-execute the selection of target training samples currently used for training in the training data set when the initial display parameter prediction model after updating the model parameters does not meet the preset training completion condition; The determination subunit is used to determine the initial display parameter prediction model that meets the preset training completion condition as the trained display parameter prediction model when the initial display parameter prediction model after the model parameters are updated meets the preset training completion condition.

8. The system according to claim 7, characterized in that Also includes: A first receiving unit is used to receive a human eye test result of a test subject on the outdoor LED display screen; A first output unit, configured to output prompt information when the human eye test result does not meet a preset application condition, wherein the prompt information is used to prompt the test subject to feedback the display parameters required; A second receiving unit, configured to receive display parameters input by the test subject; A generating unit, configured to generate a loss function value corresponding to the prediction result according to the loss function, the prediction result and a display parameter input by the test subject; An updating unit is used to update the parameters of the display parameter prediction model using the loss function value corresponding to the prediction result.

Images