A dynamic visual identification system generation method suitable for multi-terminal screens

Abstract

This invention relates to a method for generating a dynamic visual identifier system adapted to multiple screens. The proposed reinforcement learning method belongs to the field of electronic digital data processing. The method includes: using any type of screen terminal as the target screen terminal, acquiring multiple terminal association information of the target screen terminal; acquiring various identifier-related parameters of the target visual identifier as an artistic design identifier; and traversing various quantitative contrast values ​​to intelligently search for the optimal quantitative contrast value of the target visual identifier under various screen terminals using an intelligent identification model of the display effect corresponding to the target visual identifier. This invention addresses the technical problem that artistic design identifiers cannot achieve their optimal display effect when displayed separately on multiple screens with irregular display areas. By employing an artificial intelligence model corresponding to the artistic design identifier and a method of traversing quantitative contrast values, it intelligently searches for the optimal quantitative contrast value for each of the multiple screens, thereby solving the aforementioned technical problem.

Description

Technical Field

[0001] The reinforcement learning proposed in this invention belongs to the field of electronic digital data processing, and in particular relates to a method for generating a dynamic visual identifier system that is adapted to multiple screens. Background Technology

[0002] Applying artificial intelligence models obtained through reinforcement learning to the field of electronic digital data processing can solve sophisticated and intelligent technical problems that were previously unresolved by existing solutions. For example, in the past, when displaying the same visual identifier on different types of display screens, the only methods used were stretching, denoising, and / or adaptively adjusting the position of the visual identifier to improve its display effect on different screens. This dynamic visual identifier system is suitable for display environments where sophisticated and intelligent requirements are not high.

[0003] For example, Chinese invention patent publication CN104281439A discloses a method and apparatus for displaying identifiers on a terminal. The method includes: acquiring the current display information of the terminal; generating an identifier display size based on the current display information of the terminal; generating an identifier with the identifier display size; and displaying the identifier with the identifier display size. This application adaptively adjusts the identifier display size based on the current display information of the terminal, particularly the size of the terminal's display area, enabling larger display areas to display high-resolution identifiers and smaller display areas to display low-resolution identifiers, thereby improving the identifier display efficiency and giving the terminal a better identifier display effect.

[0004] For example, Chinese invention patent publication CN107172305A discloses a terminal adaptive display method, device, terminal, and storage medium. The method includes: detecting the terminal's power-on operation; acquiring identity information and reading keyword data matching the identity information; forming display identification information based on the keyword data; and generating display data matching the identity information based on the display identification information and fixed display data. This application utilizes identity information to form display identification information containing operator-related information, and obtains display data corresponding to the operator based on fixed display data and display identification information, enabling adaptive power-on / off animations for different operators.

[0005] However, the dynamic visual signage system generated by the aforementioned existing technology only adopts a relatively coarse dynamic adjustment mechanism. Because it uses an exhaustive search method to find the optimal display parameters, the display effect obtained after adjustment is not actually optimal. Equally important, the number of terminal association information of the target screen terminal and the number of signage-related parameters of the target visual signage that are integrated during dynamic adjustment is limited, resulting in insufficient refinement and intelligence of the obtained adjustment parameters. This affects the display effect of the target visual signage, especially the diverse artistic design signage, on multi-terminal screens that may have irregular display areas. Summary of the Invention

[0006] To address technical challenges in related fields, this invention provides a method for generating a dynamic visual identifier system adaptable to multiple screen types. For various screen terminals that may have irregularly shaped display areas, when displaying the same target visual identifier as an artistic design identifier in different directions, a dynamic visual identifier system adaptable to multiple screen types is generated to ensure optimal clarity on each screen type. This system employs an artificial intelligence model specifically designed for the target visual identifier and uses a quantitative contrast ratio traversal method to intelligently search for the optimal quantitative contrast ratio that achieves the best clarity on each screen type. This provides a refined and intelligent solution for adapting artistic design identifiers to various screen terminals that may have irregularly shaped display areas, improving the display performance of multiple screen terminals in specific scenarios.

[0007] According to the present invention, a method for generating a dynamic visual identifier system adapted to multiple screens is provided, the method comprising: Using any type of screen terminal as the target screen terminal, the target screen terminal's calibration resolution, refresh rate, area ratio of irregular display area, direction and distance of the center pixel of the irregular display area relative to the center pixel of the target screen terminal's full screen, pixel density, display cache capacity, and processing chip computing performance data are output as multiple terminal-related information of the target screen terminal. The brightness gradient values ​​of each edge pixel of the original image of the target visual logo, the YUV three-component values ​​of each pixel, the horizontal coordinate value, and the vertical coordinate value are used as the logo-related parameters of the target visual logo. The target visual logo is an art design logo. Reinforcement learning operations are performed on the BP neural network to obtain the BP neural network after the reinforcement learning operations are performed, and the result is used as the output of the intelligent identification model for display effect. The reinforcement learning operations are learning operations that exceed a preset threshold number of times, and the number of learning operations has the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier. The system iterates through various contrast ratio values ​​and uses the intelligent display effect identification model to perform various sharpness ratio values ​​for each contrast ratio setting under the target screen terminal's full-screen central area display of the target visual logo. The system then selects the contrast ratio value corresponding to the sharpness ratio value with the highest value as the optimal contrast ratio value for the target screen terminal's full-screen central area display of the target visual logo.

[0008] Compared with the prior art, the present invention has at least the following significant technical advancements: Significant Technological Advancement A: For various screen terminals that may have irregularly shaped display areas, when performing the directional display of the same target visual identifier as an artistic design identifier, in order to ensure that the display effect with the best clarity can be achieved on each type of screen terminal, a dynamic visual identifier system adapted to multiple screen terminals is generated. This system uses an artificial intelligence model specifically designed for the target visual identifier and a method of traversing quantitative contrast values ​​to complete the intelligent search for the optimal quantitative contrast value that can achieve the best clarity display effect on each type of screen terminal. This provides a refined and intelligent solution mechanism for adapting artistic design identifiers to various screen terminals that may have irregularly shaped display areas, and improves the display performance of various screen terminals in specific on-site scenarios. Significant Technological Advancement B: For intelligent identification of the sharpness quantitative value obtained by displaying the target visual identifier in the central area of ​​the full-screen terminal under the contrast setting of the contrast quantitative value obtained during the execution of the contrast quantitative value, a customized intelligent identification model for the display effect of the target visual identifier was designed. The intelligent identification model for the display effect of the target visual identifier is a BP neural network after performing reinforcement learning operations. The reinforcement learning operations are learning operations exceeding a preset threshold number. The number of learning operations has the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier. The BP neural network performing the reinforcement learning operations includes a single input layer, multiple hidden layers, and a single output layer connected in sequence. The number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier. The hidden layers use the ReLU function or the Softplus function as the activation function. Through the above customized structural design, the stability and effectiveness of the sharpness quantitative value obtained by intelligent identification are guaranteed. Significant technological advancement C: In each learning operation performed on the BP neural network, a certain quantitative contrast value, multiple terminal association information of a certain screen terminal, and various identification-related parameters of the target visual identifier are used as input data of the BP neural network. The known quantitative value of the clarity obtained by displaying the target visual identifier in the central area of ​​the full screen of a certain screen terminal under the contrast setting of the certain quantitative contrast value is used as the single output data of the BP neural network to complete the learning operation, thereby ensuring the learning effect of each learning operation of the BP neural network. Significant Technological Advancement D: Intelligent identification of the clarity quantitative value obtained by displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal under the contrast setting of the traversed contrast quantitative value. Various basic data were specifically selected, including the traversed contrast quantitative value, multiple terminal association information of the target screen terminal, and various identifier-related parameters of the target visual identifier. The contrast quantitative value of the target screen terminal is the difference between the maximum and minimum values ​​of the brightness values ​​corresponding to each pixel in the display screen of the target screen terminal. The clarity quantitative value of the target screen terminal is the minimum number of pixels required for the black-and-white boundary line in the display screen of the target screen terminal to transition from black to white. The comprehensive and sufficient selection of the above basic data further ensures the stability and effectiveness of the clarity quantitative value obtained by intelligent identification. Significant Technological Advancement E: More specifically, taking any type of screen terminal as the target screen terminal, the target screen terminal's calibration resolution, refresh rate, area ratio of irregular display areas, direction and distance of the center pixel of the irregular display area relative to the center pixel of the target screen terminal's full screen, pixel density, display cache capacity, and processing chip computing performance data are used as multiple terminal-related information of the target screen terminal. The brightness gradient values ​​of each edge pixel of the original image of the target visual identifier, the YUV three-component values ​​of each pixel, the horizontal coordinate values, and the vertical coordinate values ​​are used as the identifier-related parameters of the target visual identifier. Thus, a customized data structure design is completed for the various basic data used for intelligent identification. Attached Figure Description

[0009] The accompanying drawings of this invention employ a "general-specific" structure for technical disclosure: First layer: Overall overview ├─ Figure 1 A schematic diagram illustrating the working scenario of a dynamic visual identifier generation method adapted to multiple screen devices. │ Second layer: Method implementation (five examples) ├─ Figure 2 Flowchart of the method steps in the first embodiment ├─ Figure 3Flowchart of the method steps in the second embodiment ├─ Figure 4 Flowchart of the method steps in the third embodiment ├─ Figure 5 Flowchart of the method steps in the fourth embodiment ├─ Figure 6 Fifth Embodiment Method Step Flowchart The accompanying figures complement each other to form a complete system for disclosing the technical solution. Detailed Implementation

[0010] Figure 1 A schematic diagram illustrating a working scenario of a dynamic visual identifier generation method adapted to multiple screens according to the present invention is provided. The reinforcement learning proposed in this invention belongs to the field of electronic digital data processing.

[0011] The specific technical process of this invention is as follows: Technical Process 1: Design a customized display effect intelligent identification model for the target visual logo, which serves as an artistic design identifier. This model is used to intelligently identify the quantitative value of the clarity obtained by displaying the target visual logo in the central area of ​​the full screen of the target screen terminal under the contrast setting of the traversed quantitative contrast values. More specifically, by taking any type of screen terminal as the target screen terminal, it is possible to intelligently search for the optimal quantitative value of contrast for each type of screen terminal, i.e., multiple screens, when displaying the target visual identifier. More specifically, the target visual identity is a diverse artistic design identity, and the target visual identity is displayed on multiple screens that may have irregular display areas. The diversity of the identity and the complexity of the display area caused by the irregular display areas make the display adjustment mechanism of the existing coarse dynamic visual identity more difficult to adapt. More specifically, the intelligent identification model for display effects, specifically designed for target visual identifiers, has the following customized structural designs: First: The intelligent identification model for the display effect of the target visual identifier is a BP neural network after the reinforcement learning operation is completed. The reinforcement learning operation is each learning operation exceeding the preset number threshold. The number of learning operations has the same numerical change trend as the total number of pixels occupied by the original image of the target visual identifier. For example, when the original image of the target visual identifier occupies a total of 10,000 pixels, the number of learning operations selected is 800; when the original image of the target visual identifier occupies a total of 20,000 pixels, the number of learning operations selected is 900; when the original image of the target visual identifier occupies a total of 30,000 pixels, the number of learning operations selected is 1,000; when the original image of the target visual identifier occupies a total of 40,000 pixels, the number of learning operations selected is 1,100, and so on. The second point: The BP neural network that performs reinforcement learning operations includes a single input layer, multiple hidden layers and a single output layer connected in sequence. The number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier, and the hidden layers use the ReLU function or the Softplus function as activation functions. For example, when the total number of edge pixels in the original image of the target visual identifier is on the order of 1,000, the number of hidden layers selected is 2; when the total number of edge pixels in the original image of the target visual identifier is on the order of 2,000, the number of hidden layers selected is 4; when the total number of edge pixels in the original image of the target visual identifier is on the order of 4,000, the number of hidden layers selected is 8, and so on. Thirdly: In each learning operation performed on the BP neural network, a certain quantitative contrast value, multiple terminal association information of a certain screen terminal, and each set of label-related parameters of the target visual label are used as input data of the BP neural network. The known quantitative value of the clarity obtained by displaying the target visual label in the central area of ​​the full screen of a certain screen terminal under the contrast setting of the certain quantitative contrast value is used as the single output data of the BP neural network to complete the learning operation, thereby ensuring the learning effect of each learning operation of the BP neural network. In this way, the stability and effectiveness of the quantitative values ​​of clarity obtained by intelligent identification are ensured through the above-mentioned customized structural designs. Technical Process 2: To intelligently identify the clarity of the target visual identifier displayed in the central area of ​​the full screen of the target screen terminal under the contrast setting of the traversed quantitative contrast values, various basic data were selected in a targeted manner. More specifically, such as Figure 1 As shown, the basic data includes the contrast ratio quantitative value obtained through the traversal, multiple terminal association information of the target screen terminal, and various identification-related parameters of the target visual identifier. The contrast ratio quantitative value comes from the big data storage node. The contrast ratio quantitative value of the target screen terminal is the difference between the maximum and minimum values ​​of the brightness values ​​corresponding to each pixel in the display screen of the target screen terminal. The sharpness quantitative value of the target screen terminal is the minimum number of pixels required for the black-and-white boundary line in the display screen of the target screen terminal to transition from black to white. More specifically, taking any type of screen terminal as the target screen terminal, the target screen terminal's calibration resolution, refresh rate, area ratio of irregular display area, direction and distance of the center pixel of the irregular display area relative to the center pixel of the target screen terminal's full screen, pixel density, display cache capacity, and processing chip computing performance data are used as multiple terminal-related information of the target screen terminal. The brightness gradient values ​​of each edge pixel of the original image of the target visual identifier, the YUV three-component values ​​of each pixel, the horizontal coordinate value, and the vertical coordinate value are used as the identifier-related parameters of the target visual identifier. Thus, a customized data structure design is completed for the various basic data used for intelligent identification. In this way, the stability and effectiveness of the quantitative values ​​of clarity obtained by intelligent identification are further guaranteed through the comprehensive and sufficient selection of the above-mentioned basic data. Technical Process Three: Using a customized display effect intelligent assessment model based on Technical Process One, and employing the comprehensive and thorough selection of fundamental data from Technical Process Two, the intelligent assessment determines the clarity of the target visual identifier displayed in the central area of ​​the target screen terminal under the contrast settings traversed. For example... Figure 1 As shown; In this way, by traversing all the quantitative contrast values, the intelligent identification of each quantitative value of sharpness obtained from the display of the target visual identifier in the central area of ​​the full screen terminal is completed. Technical Process 4: Based on the quantitative values ​​of sharpness obtained by traversing all the quantitative values ​​of contrast in Technical Process 3, analyze the optimal quantitative value of contrast for displaying the target visual logo in the central area of ​​the full screen terminal. More specifically, the contrast ratio corresponding to the largest value among the various resolution ratio values ​​is taken as the optimal contrast ratio for displaying the target visual logo in the central area of ​​the full screen terminal. More specifically, since the target screen terminal is any type of screen terminal, it can perform directional intelligent search for the optimal quantitative values ​​of contrast corresponding to the display of the same visual logo on various screen terminals, thereby obtaining the best display effect of the same visual logo on various screen terminals. Therefore, through the coordinated operation of the above-mentioned technical processes, when displaying the same target visual identifier as an artistic design identifier in different directions for multiple screen terminals that may have irregular display areas, a dynamic visual identifier system adapted to multiple screens is generated to ensure that the best clarity display effect can be achieved on each type of screen terminal. This system uses an artificial intelligence model specifically designed for the target visual identifier and a method of traversing quantitative contrast values ​​to complete the intelligent search for the optimal quantitative contrast value that can achieve the best clarity display effect on each type of screen terminal. This provides a refined and intelligent solution mechanism for adapting artistic design identifiers to multiple screen terminals that may have irregular display areas, and improves the display performance of multiple screen terminals in specific on-site scenarios.

[0012] The key points of this invention are: directional intelligent search for the best display effect of the same artistic design logo on multiple screen terminals that may have irregular display areas; targeted calculation of quantitative values ​​of contrast and sharpness; intelligent search mode based on the traversal mechanism of each quantitative value of contrast; and customized structural design of intelligent identification model for different display effects of different artistic design logos.

[0013] The following will describe in detail, by way of an embodiment, a method for generating a dynamic visual identifier system adapted to multiple screens according to the present invention. Example

[0014] Figure 2 This is a flowchart illustrating the steps of a method for generating a dynamic visual identifier system adapted to multiple screens according to a first embodiment of the present invention.

[0015] like Figure 2 As shown, the method for generating a dynamic visual identifier system adapted to multiple screens includes the following specific steps: Step S1: Using any type of screen terminal as the target screen terminal, output the target screen terminal's calibration resolution, refresh rate, area ratio occupied by irregular display area, direction and distance of the center pixel of the irregular display area relative to the center pixel of the target screen terminal's full screen, pixel density, display cache capacity, and processing chip computing performance data as multiple terminal association information of the target screen terminal. Thus, since any type of screen terminal is used as the target screen terminal, the resulting dynamic visual identification system is adaptable to all types of screen terminals. For example, the irregular display area on the target screen terminal's display interface may be a notch, a curved screen on both sides, or other distorted display areas at the top; Step S2: Use the brightness gradient value of each edge pixel, the YUV three-component value of each pixel, the horizontal coordinate value and the vertical coordinate value of each pixel as the identification-related parameters of the target visual logo. The target visual logo is an art design logo. More specifically, in the YUV three-component values ​​of each pixel, the Y component value is the luminance component value, while the UV component values ​​are two different chrominance component values. Step S3: Perform reinforcement learning operations on the BP neural network to obtain the BP neural network after the reinforcement learning operations are performed and output as the intelligent identification model for display effect. The reinforcement learning operations are learning operations that exceed a preset threshold number of times, and the number of learning operations has the same numerical change trend as the total number of pixels occupied by the original image of the target visual identifier. For example, the number of learning operations follows the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier: when the original image of the target visual identifier occupies 10,000 pixels, the number of learning operations selected is 800; when the original image of the target visual identifier occupies 20,000 pixels, the number of learning operations selected is 900; when the original image of the target visual identifier occupies 30,000 pixels, the number of learning operations selected is 1,000; when the original image of the target visual identifier occupies 40,000 pixels, the number of learning operations selected is 1,100, and so on. Step S4: Iterate through each contrast quantitative value and use the display effect intelligent identification model to execute each contrast quantitative value of the target visual logo displayed in the central area of ​​the full screen terminal under the contrast setting of each contrast quantitative value. The contrast quantitative value corresponding to the largest clarity quantitative value is taken as the optimal contrast quantitative value for displaying the target visual logo in the central area of ​​the full screen terminal. It is worth noting that the quantitative values ​​of contrast and sharpness here are not the conventional values ​​of contrast and sharpness; they are calculated using a specific method. The specific calculation methods for both are as follows: The quantitative value of the contrast of the target screen terminal is the difference between the maximum and minimum values ​​of the brightness values ​​corresponding to each pixel in the display screen of the target screen terminal; the quantitative value of the sharpness of the target screen terminal is the minimum number of pixels required for the black-and-white boundary line in the display screen of the target screen terminal to transition from black to white. For each quantitative contrast value, the display effect intelligent identification model uses the quantitative contrast value, multiple terminal association information of the target screen terminal, and each set of identification-related parameters of the target visual identifier to intelligently identify the quantitative clarity value obtained by displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value. Therefore, for the target screen terminal, various quantitative values ​​of contrast were traversed to obtain various quantitative values ​​of sharpness. The BP neural network that performs reinforcement learning includes a single input layer, multiple hidden layers and a single output layer connected in sequence. The number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier. The hidden layers use ReLU or Softplus functions as activation functions. For example, the number of hidden layers and the total number of edge pixels in the original image of the target visual identifier have the same numerical trend, including: when the total number of edge pixels in the original image of the target visual identifier is on the order of 1,000, the number of hidden layers selected is 2; when the total number of edge pixels in the original image of the target visual identifier is on the order of 2,000, the number of hidden layers selected is 4; when the total number of edge pixels in the original image of the target visual identifier is on the order of 4,000, the number of hidden layers selected is 8, and so on. In each learning operation performed on the BP neural network, a certain quantitative contrast value, multiple terminal association information of a certain screen terminal, and various identification-related parameters of the target visual identifier are used as input data of the BP neural network. The known quantitative clarity value obtained by displaying the target visual identifier in the central area of ​​the full screen of a certain screen terminal under the contrast setting of the certain quantitative contrast value is used as the single output data of the BP neural network to complete the learning operation. Example

[0016] Figure 3 This is a flowchart illustrating the steps of a dynamic visual identifier system generation method adapted to multiple screens according to a second embodiment of the present invention.

[0017] like Figure 3 As shown, with Figure 2 Unlike the embodiments described above, in the method for generating a dynamic visual identifier system adapted to multiple screens, after taking any type of screen terminal as the target screen terminal and outputting the target screen terminal's calibration resolution, refresh rate, area ratio of irregular display areas, direction and distance of the center pixel of the irregular display area relative to the center pixel of the target screen terminal's full screen, pixel density, display cache capacity, and processing chip computing performance data as multiple terminal association information of the target screen terminal, that is, after step S1, the method further includes: Step S5: Receive and store the original image of the target visual identifier, and perform image analysis on the original image of the target visual identifier to obtain the brightness gradient value of each edge pixel, the YUV three component values ​​of each pixel, the horizontal coordinate value and the vertical coordinate value of the original image of the target visual identifier. The process of receiving and storing the original image of the target visual identifier and performing image analysis on the original image of the target visual identifier to obtain the brightness gradient value of each edge pixel, the YUV three-component value of each pixel, the horizontal coordinate value and the vertical coordinate value of each pixel include: the YUV three-component value of each pixel in the original image of the target visual identifier is the Y component value, U component value and V component value of the pixel in the YUV space. More specifically, the YUV three-component values ​​of each pixel in the original image of the target visual identifier are the Y component value, U component value and V component value of the pixel in the YUV space, including: the Y component value, U component value and V component value of each pixel in the YUV space are all between 0 and 255. Example

[0018] Figure 4 This is a flowchart illustrating the steps of a method for generating a dynamic visual identifier system adapted to multiple screens according to a third embodiment of the present invention.

[0019] like Figure 4 As shown, with Figure 2 Unlike the previous embodiment, in the method for generating a dynamic visual identifier system adapted to multiple screens, after performing reinforcement learning operations on a BP neural network to obtain a BP neural network after the reinforcement learning operations are completed and output as a display effect intelligent identification model, the reinforcement learning operations are learning operations exceeding a preset threshold number, and the number of learning operations has the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier, that is, after step S3, the method further includes: Step S6: Receive the intelligent identification model of display effect, and complete the model storage of the intelligent identification model of display effect by storing the various model parameters of the intelligent identification model of display effect; The process of receiving the intelligent identification model of display effect and storing the model of the intelligent identification model of display effect by storing the various model parameters of the intelligent identification model of display effect includes: storing the various model parameters of the intelligent identification model of display effect using different physical storage addresses respectively; For example, different physical storage addresses can be used to store the various model parameters of the intelligent identification model of display effect. These include: different physical storage addresses within the MMC storage chip can be selected to store the various model parameters of the intelligent identification model of display effect separately. Example

[0020] Figure 5 This is a flowchart illustrating the steps of a method for generating a dynamic visual identifier system adapted to multiple screens according to a fourth embodiment of the present invention.

[0021] like Figure 5 As shown, with Figure 2 Unlike the previous embodiment, in the method for generating a dynamic visual identifier system adapted to multiple screens, after iterating through various contrast quantitative values ​​and using a display effect intelligent identification model to execute the contrast settings for each contrast quantitative value, and obtaining various sharpness quantitative values ​​for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal, and taking the contrast quantitative value corresponding to the sharpness quantitative value with the largest value as the optimal contrast quantitative value for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal, that is, after step S4, the method further includes: Step S7: Receive the optimal contrast ratio quantitative value of the target visual logo displayed in the central area of ​​the full screen of the target screen terminal, and display the optimal contrast ratio quantitative value on site; For example, an LED display array can be selected to receive the optimal contrast ratio of the target visual identifier displayed in the central area of ​​the full screen of the target screen terminal, and the optimal contrast ratio can be displayed on site. The process of receiving the optimal contrast ratio of the target visual identifier displayed in the central area of ​​the full screen of the target screen terminal and displaying the optimal contrast ratio on-site includes: determining that the target visual identifier is displayed in the central area of ​​the full screen of the target screen terminal when the center pixel of the central area of ​​the full screen of the target screen terminal overlaps with the center pixel of the display area occupied by the target visual identifier. Example

[0022] Figure 6 This is a flowchart illustrating the steps of a method for generating a dynamic visual identifier system adapted to multiple screens according to a fifth embodiment of the present invention.

[0023] like Figure 6 As shown, with Figure 2 Unlike the previous embodiment, in the method for generating a dynamic visual identifier system adapted to multiple screens, after iterating through various contrast quantitative values ​​and using a display effect intelligent identification model to execute the contrast settings for each contrast quantitative value, and obtaining various sharpness quantitative values ​​for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal, and taking the contrast quantitative value corresponding to the sharpness quantitative value with the largest value as the optimal contrast quantitative value for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal, that is, after step S4, the method further includes: Step S8: Receive the optimal contrast ratio quantitative value of the target visual logo displayed in the central area of ​​the full screen of the target screen terminal, and wirelessly transmit the optimal contrast ratio quantitative value to the remote screen display monitoring server through the mobile communication network. The process of receiving the optimal contrast quantitative value of the target visual identifier displayed in the central area of ​​the full screen of the target screen terminal and wirelessly transmitting the optimal contrast quantitative value to the remote screen display monitoring server via a mobile communication network includes: the mobile communication network is based on time-division duplex communication mode. For example, the remote screen display monitoring server is one of the following: a big data monitoring server, a cloud computing monitoring server, or a blockchain monitoring server.

[0024] Next, the various method embodiments of the present invention will be described in detail.

[0025] In a method for generating a dynamic visual identifier system adapted to multiple screens according to various embodiments of the present invention: The number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier, and the hidden layers use ReLU function or Softplus function as activation function, including: using a first numerical change curve to represent the numerical trend of the total number of edge pixels in the original image of the target visual identifier, and using a second numerical change curve to represent the numerical trend of the number of hidden layers. More specifically, different GAL chips can be used to test and simulate the implementation process of the first and second numerical change curves, respectively. Among them, the number of hidden layers has the same numerical change trend as the total number of edge pixels in the original image of the target visual identifier, and the use of ReLU function or Softplus function as activation function for hidden layers also includes: the curvature values ​​at uniform intervals on the second numerical change curve are equal to the curvature values ​​at uniform intervals on the second numerical change curve. Among them, the number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier, and the use of ReLU function or Softplus function as activation function for hidden layers also includes: the number of uniformly spaced points on the first numerical change curve is equal to the number of uniformly spaced points on the second numerical change curve.

[0026] In a method for generating a dynamic visual identifier system adapted to multiple screens according to various embodiments of the present invention: The area ratio occupied by the irregular display area of ​​the target screen terminal is the ratio of the number of pixels occupied by the irregular display area of ​​the target screen terminal to the total number of pixels of the target screen terminal. Among them, the pixel closest to the centroid of the irregular display area of ​​the target screen terminal is taken as the center pixel of the irregular display area, and the pixel closest to the centroid of the entire display area of ​​the target screen terminal is taken as the center pixel of the entire screen of the target screen terminal. More specifically, if the centroid of the irregular display area of ​​the target screen terminal or the centroid of the entire display area of ​​the target screen terminal is two pixels, then the center pixel of the irregular display area is the centroid of the irregular display area of ​​the target screen terminal, and the center pixel of the entire screen of the target screen terminal is the centroid of the entire display area of ​​the target screen terminal. Among them, the calibrated resolution of the target screen terminal is the default resolution of the target screen terminal in the factory state and includes both horizontal and vertical resolution; the display cache capacity of the target screen terminal is the maximum total amount of data that the display cache of the target screen terminal can store; and the computing performance data of the processing chip of the target screen terminal is the upper limit of the number of operations that the processing chip of the target screen terminal can perform per unit time. Specifically, the brightness gradient values ​​of each edge pixel in the original image of the target visual identifier, the YUV three-component values ​​of each pixel, the horizontal coordinate value, and the vertical coordinate value are used as the identifier-related parameters of the target visual identifier. The target visual identifier is an artistic design identifier, which includes: the brightness gradient value of each edge pixel is the mean square error of the brightness values ​​of the edge pixel and its surrounding pixels.

[0027] And in a method for generating a dynamic visual identifier system adapted to multiple screens according to various method embodiments of the present invention: For each quantitative contrast value, the intelligent display effect assessment model uses the quantitative contrast value, multiple terminal association information of the target screen terminal, and each set of label-related parameters of the target visual label to intelligently assess the quantitative clarity value obtained by displaying the target visual label in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value. This includes: synchronously inputting the quantitative contrast value, multiple terminal association information of the target screen terminal, and each set of label-related parameters of the target visual label into the intelligent display effect assessment model. Specifically, for each quantitative contrast value, the intelligent evaluation model of display effect uses the quantitative contrast value, multiple terminal association information of the target screen terminal, and various identification-related parameters of the target visual identifier to intelligently evaluate the quantitative clarity value obtained by displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value. This also includes: running the intelligent evaluation model of display effect to obtain the quantitative clarity value of displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value, as output by the intelligent evaluation model of display effect. Among them, programmable logic devices are used to realize the synchronous input of the quantitative value of contrast, multiple terminal association information of the target screen terminal, and various identification-related parameters of the target visual identifier to the intelligent identification model of display effect; For example, the synchronous input of the quantitative contrast value, multiple terminal association information of the target screen terminal, and various identification-related parameters of the target visual identifier to the intelligent identification model of the display effect using a programmable logic device includes: the programmable logic device is an FPGA device, and the FPGA device used is designed with VHDL language for logic programming. Among them, the quantitative value of contrast, multiple terminal association information of the target screen terminal, each set of label-related parameters of the target visual label, and the quantitative value of the clarity of intelligent identification are all expressed in a numerical representation after numerical normalization. For example, the quantitative value of contrast, multiple terminal association information of the target screen terminal, various identification-related parameters of the target visual identifier, and the quantitative value of clarity of intelligent identification are all expressed in a numerical normalization form, including: the numerical normalization process is an octal numerical conversion process.

[0028] In addition, the present invention may also refer to the following technical contents to further highlight the significant technical progress of the present invention: Reinforcement learning operations are performed on the BP neural network to obtain the BP neural network after the reinforcement learning operations are performed, and the result is used as the output of the intelligent identification model for display effect. The reinforcement learning operations are each learning operation exceeding a preset threshold number, and the number of learning operations has the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier. This includes: using an information conversion formula to represent the information conversion relationship where the number of learning operations has the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier. The information conversion relationship, in which the number of learning operations and the total number of pixels occupied by the original image of the target visual identifier have the same numerical trend, is expressed by the information conversion formula. In the information conversion formula, the total number of pixels occupied by the original image of the target visual identifier is the input information of the information conversion formula. The information conversion formula, which expresses the information conversion relationship where the number of learning operations and the total number of pixels occupied by the original image of the target visual identifier have the same numerical trend, further includes: in the information conversion formula, the number of learning operations that have the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier is the output information of the information conversion formula. For example, the information conversion relationship, which uses an information conversion formula to represent the information conversion relationship where the number of learning operations and the total number of pixels in the original image of the target visual identifier have the same numerical trend, also includes: simulation and testing of the data processing process using a SOC chip to complete the information conversion relationship where the number of learning operations and the total number of pixels in the original image of the target visual identifier have the same numerical trend.

[0029] While only exemplary embodiments of the invention have been described in detail above, those skilled in the art will immediately understand that many variations can be made to these exemplary embodiments without substantially departing from the novel principles and advantages of the invention. Accordingly, all such variations are to be included within the scope of protection defined by the claims of the invention.

Claims

1. A method for generating a dynamic visual identifier system adapted to multiple screen devices, characterized in that, The method includes: Using any type of screen terminal as the target screen terminal, the target screen terminal's calibration resolution, refresh rate, area ratio of irregular display area, direction and distance of the center pixel of the irregular display area relative to the center pixel of the target screen terminal's full screen, pixel density, display cache capacity, and processing chip computing performance data are output as multiple terminal-related information of the target screen terminal. The brightness gradient values ​​of each edge pixel of the original image of the target visual logo, the YUV three-component values ​​of each pixel, the horizontal coordinate value, and the vertical coordinate value are used as the logo-related parameters of the target visual logo. The target visual logo is an art design logo. Reinforcement learning operations are performed on the BP neural network to obtain the BP neural network after the reinforcement learning operations are performed, and the result is used as the output of the intelligent identification model for display effect. The reinforcement learning operations are learning operations that exceed a preset threshold number of times, and the number of learning operations has the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier. The system iterates through various contrast ratio values ​​and uses the intelligent display effect identification model to perform various sharpness ratio values ​​for each contrast ratio setting under the target screen terminal's full-screen central area display of the target visual logo. The system then selects the contrast ratio value corresponding to the sharpness ratio value with the highest value as the optimal contrast ratio value for the target screen terminal's full-screen central area display of the target visual logo.

2. The method for generating a dynamic visual identifier system adapted to multiple screens as described in claim 1, characterized in that: For each quantitative contrast value, the display effect intelligent identification model uses the quantitative contrast value, multiple terminal association information of the target screen terminal, and each set of identification-related parameters of the target visual identifier to intelligently identify the quantitative clarity value obtained by displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value. Among them, the quantitative value of the contrast ratio of the target screen terminal is the difference between the maximum and minimum values ​​of the brightness values ​​corresponding to each pixel in the display screen of the target screen terminal, and the quantitative value of the sharpness of the target screen terminal is the minimum number of pixels required for the black-and-white boundary line in the display screen of the target screen terminal to transition from black to white.

3. The method for generating a dynamic visual identifier system adapted to multiple screens as described in claim 2, characterized in that: The BP neural network that performs reinforcement learning operations consists of a single input layer, multiple hidden layers and a single output layer connected in sequence. The number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier, and the hidden layers use the ReLU function or the Softplus function as activation functions. In each learning operation performed on the BP neural network, a certain quantitative contrast value, multiple terminal association information of a certain screen terminal, and various identification-related parameters of the target visual identifier are used as input data of the BP neural network. The known quantitative value of the clarity obtained by displaying the target visual identifier in the central area of ​​the full screen of a certain screen terminal under the contrast setting of the certain quantitative contrast value is used as the single output data of the BP neural network to complete the learning operation.

4. The method for generating a dynamic visual identifier system adapted to multiple screens as described in claim 3, characterized in that, After taking any type of screen terminal as the target screen terminal and outputting the target screen terminal's calibrated resolution, refresh rate, area ratio of irregular display areas, direction and distance of the center pixel of the irregular display area relative to the center pixel of the target screen terminal's full screen, pixel density, display cache capacity, and processing chip computing performance data as multiple terminal association information of the target screen terminal, the method further includes: Receive and store the original image of the target visual identifier, and perform image analysis on the original image of the target visual identifier to obtain the brightness gradient value of each edge pixel, the YUV three component values ​​of each pixel, the horizontal coordinate value and the vertical coordinate value of the original image of the target visual identifier. The process of receiving and storing the original image of the target visual identifier, and performing image analysis on the original image of the target visual identifier to obtain the brightness gradient value of each edge pixel, the YUV three-component value of each pixel, the horizontal coordinate value, and the vertical coordinate value of the original image of the target visual identifier, includes: the YUV three-component value of each pixel in the original image of the target visual identifier is the Y component value, U component value, and V component value of the pixel in the YUV space.

5. The method for generating a dynamic visual identifier system adapted to multiple screens as described in claim 3, characterized in that, After performing reinforcement learning operations on a BP neural network to obtain a post-reinforcement learning BP neural network, which is then used as the output of an intelligent identification model for display effects, the method further includes: (The reinforcement learning operations are defined as learning operations exceeding a preset threshold number, and the number of learning operations has the same numerical trend as the total number of pixels occupied by the original image of the target visual identifier.) Receive the intelligent identification model of display effect, and complete the model storage of the intelligent identification model of display effect by storing various model parameters of the intelligent identification model of display effect; The process of receiving the intelligent identification model of display effect and storing the model parameters of the intelligent identification model of display effect includes storing the various model parameters of the intelligent identification model of display effect using different physical storage addresses.

6. The method for generating a dynamic visual identifier system adapted to multiple screens as described in claim 3, characterized in that, After iterating through various contrast ratio values ​​and using the intelligent identification model for display effects to execute the contrast settings for each contrast ratio value, the method further includes obtaining various sharpness ratio values ​​for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal, and taking the contrast ratio value corresponding to the sharpness ratio value with the largest value as the optimal contrast ratio value for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal. Receive the optimal contrast ratio quantitative value of the target visual logo displayed in the central area of ​​the full screen terminal, and display the optimal contrast ratio quantitative value on site; The process of receiving the optimal contrast ratio of the target visual identifier displayed in the central area of ​​the full screen of the target screen terminal and displaying the optimal contrast ratio on-site includes: determining that the target visual identifier is displayed in the central area of ​​the full screen of the target screen terminal when the center pixel of the central area of ​​the full screen of the target screen terminal overlaps with the center pixel of the display area occupied by the target visual identifier.

7. The method for generating a dynamic visual identifier system adapted to multiple screens as described in claim 3, characterized in that, After iterating through various contrast ratio values ​​and using the intelligent identification model for display effects to execute the contrast settings for each contrast ratio value, the method further includes obtaining various sharpness ratio values ​​for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal, and taking the contrast ratio value corresponding to the sharpness ratio value with the largest value as the optimal contrast ratio value for displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal. Receive the optimal contrast ratio quantitative value of the target visual logo displayed in the central area of ​​the full screen of the target screen terminal, and wirelessly transmit the optimal contrast ratio quantitative value to the remote screen display monitoring server through the mobile communication network. The process of receiving the optimal contrast ratio quantitative value of the target visual identifier displayed in the central area of ​​the full-screen terminal of the target screen terminal, and wirelessly transmitting the optimal contrast ratio quantitative value to the remote screen display monitoring server via a mobile communication network includes: the mobile communication network is based on time-division duplex communication mode.

8. A method for generating a dynamic visual identifier system adapted to multiple screens as described in any one of claims 3-7, characterized in that: The number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier, and the hidden layers use ReLU function or Softplus function as activation function, including: using a first numerical change curve to represent the numerical trend of the total number of edge pixels in the original image of the target visual identifier, and using a second numerical change curve to represent the numerical trend of the number of hidden layers. Among them, the number of hidden layers has the same numerical change trend as the total number of edge pixels in the original image of the target visual identifier, and the use of ReLU function or Softplus function as activation function for hidden layers also includes: the curvature values ​​at uniform intervals on the second numerical change curve are equal to the curvature values ​​at uniform intervals on the second numerical change curve. Among them, the number of hidden layers has the same numerical trend as the total number of edge pixels in the original image of the target visual identifier, and the use of ReLU function or Softplus function as activation function for hidden layers also includes: the number of uniformly spaced points on the first numerical change curve is equal to the number of uniformly spaced points on the second numerical change curve.

9. A method for generating a dynamic visual identifier system adapted to multiple screens as described in any one of claims 3-7, characterized in that: The area ratio occupied by the irregular display area of ​​the target screen terminal is the ratio of the number of pixels occupied by the irregular display area of ​​the target screen terminal to the total number of pixels of the target screen terminal. Among them, the pixel closest to the centroid of the irregular display area of ​​the target screen terminal is taken as the center pixel of the irregular display area, and the pixel closest to the centroid of the entire display area of ​​the target screen terminal is taken as the center pixel of the entire screen of the target screen terminal. Among them, the calibrated resolution of the target screen terminal is the default resolution of the target screen terminal in the factory state and includes both horizontal and vertical resolution; the display cache capacity of the target screen terminal is the maximum total amount of data that the display cache of the target screen terminal can store; and the computing performance data of the processing chip of the target screen terminal is the upper limit of the number of operations that the processing chip of the target screen terminal can perform per unit time. Specifically, the brightness gradient values ​​of each edge pixel in the original image of the target visual identifier, the YUV three-component values ​​of each pixel, the horizontal coordinate value, and the vertical coordinate value are used as the identifier-related parameters of the target visual identifier. The target visual identifier is an artistic design identifier, which includes: the brightness gradient value of each edge pixel is the mean square error of the brightness values ​​of the edge pixel and its surrounding pixels.

10. A method for generating a dynamic visual identifier system adapted to multiple screens as described in any one of claims 3-7, characterized in that: For each quantitative contrast value, the intelligent display effect assessment model uses the quantitative contrast value, multiple terminal association information of the target screen terminal, and each set of label-related parameters of the target visual label to intelligently assess the quantitative clarity value obtained by displaying the target visual label in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value. This includes: synchronously inputting the quantitative contrast value, multiple terminal association information of the target screen terminal, and each set of label-related parameters of the target visual label into the intelligent display effect assessment model. Specifically, for each quantitative contrast value, the intelligent evaluation model of display effect uses the quantitative contrast value, multiple terminal association information of the target screen terminal, and various identification-related parameters of the target visual identifier to intelligently evaluate the quantitative clarity value obtained by displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value. This also includes: running the intelligent evaluation model of display effect to obtain the quantitative clarity value of displaying the target visual identifier in the central area of ​​the full screen of the target screen terminal under the contrast setting of the quantitative contrast value, as output by the intelligent evaluation model of display effect. Among them, programmable logic devices are used to realize the synchronous input of the quantitative value of contrast, multiple terminal association information of the target screen terminal, and various identification-related parameters of the target visual identifier to the intelligent identification model of display effect; Among them, the quantitative value of contrast, multiple terminal association information of the target screen terminal, various identification-related parameters of the target visual identifier, and the quantitative value of clarity of intelligent identification are all expressed in a numerical representation after numerical normalization.

Images