Liquid crystal display screen dynamic compensation method and system based on pixel response characteristics

Abstract

The application relates to a liquid crystal display screen dynamic compensation method and system based on pixel response characteristics, and belongs to the technical field of liquid crystal display screens. The application constructs a node set and an edge set, initializes feature coding of the nodes, constructs a liquid crystal display screen compensation model based on a space-time graph neural network, trains the liquid crystal display screen compensation model by using the feature coding, outputs a driving voltage correction amount of each pixel and real brightness of each pixel in a current time stamp by using the liquid crystal display screen compensation model, and finally adjusts model parameters of the liquid crystal display screen compensation model in combination with the real brightness of each pixel. The application can fully consider and express a non-local coupling relationship related to a non-linear physical distance, model the display screen as a heterogeneous graph, the nodes contain pixel units and auxiliary nodes, the edges contain various physical relationships, the dynamic prediction and compensation of crosstalk between pixels can be realized, and the pixel compensation precision of the liquid crystal display screen is improved.

Description

Technical Field

[0001] This invention relates to the field of technology, and in particular to a method and system for dynamic compensation of liquid crystal displays based on pixel response characteristics. Background Technology

[0002] Liquid crystal displays (LCDs) have become the mainstream display technology due to their advantages such as thinness, low power consumption, and high resolution. However, unlike self-emissive display technologies such as cathode ray tubes (CRTs) or organic light-emitting diodes (OLEDs), LCDs rely on the physical rotation of liquid crystal molecules to control backlight transmittance, thereby achieving grayscale display. This physical mechanism determines a core challenge for LCDs in displaying dynamic images: pixel response characteristics. Pixel response characteristics are usually measured by "response time," which is the time required for liquid crystal molecules to transition from one grayscale level to another, generally including rise time and fall time. When the response time is too long, and the video content (especially fast-moving scenes) changes rapidly between frames, the liquid crystal molecules cannot complete a complete grayscale switch before the next frame arrives, resulting in afterimages of the previous frame. Visually, this phenomenon manifests as blurred motion edges, image trailing, or unclear outlines, severely reducing the clarity of dynamic images and the viewing experience. Especially on display devices with ever-increasing refresh rates (such as 120Hz, 240Hz or even higher), the display window for each frame is compressed to a few milliseconds, which places extremely demanding requirements on the response speed of the LCD.

[0003] However, traditional brightness uniformity correction is based solely on the brightness distribution of static images, generating fixed compensation coefficients. It cannot handle dynamic crosstalk, such as brightness drift caused by bright pixels heating adjacent dark pixels, data line voltage drop coupling, and common electrode voltage fluctuations. Moreover, it cannot express the non-local coupling relationship between pixels that is non-linearly related to physical distance, such as capacitive coupling between two pixels that are 10 pixels apart but share the same data line, or thermal diffusion paths between diagonal pixels. Summary of the Invention

[0004] This invention overcomes the shortcomings of the prior art and provides a dynamic compensation method and system for liquid crystal displays based on pixel response characteristics.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: The first aspect of this invention provides a dynamic compensation method for a liquid crystal display screen based on pixel response characteristics, comprising the following steps: Obtain multi-source heterogeneous parameter information from the panel design database, perform simulation based on the multi-source heterogeneous parameter information, and extract the coupling coefficient matrix; Based on the multi-source heterogeneous parameter information and coupling coefficient matrix, a node set and an edge set are constructed, and the nodes are initialized with feature encoding. A liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network. The liquid crystal display compensation model is trained using the feature encoding. The liquid crystal display compensation model outputs the driving voltage correction amount and the true brightness of each pixel in the current timestamp. The model parameters of the liquid crystal display compensation model are adjusted based on the actual brightness of each pixel, and the driving voltage correction amount of each pixel in the current timestamp is used to dynamically compensate the liquid crystal display.

[0006] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, multi-source heterogeneous parameter information is obtained from the panel design database, specifically as follows: At the time of manufacture, the design parameter data of the liquid crystal display screen is collected, and the pixel array topology, the set of pixel columns connected to each data line, the data line resistance, the parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer are obtained from the design parameter data of the liquid crystal display screen. Construct a panel design database and store the pixel array topology, the set of pixel columns connected to each data line, the data line resistance, the parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer into the panel design database. The panel design database is used to extract the pixel array topology of the current type of liquid crystal display, the set of pixel columns connected by each data line, the data line resistance, parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer, forming multi-source heterogeneous parameter information.

[0007] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, simulation is performed based on the multi-source heterogeneous parameter information to extract the coupling coefficient matrix, specifically as follows: Using the circuit simulation software HSPICE, a fine model of a single pixel and four adjacent pixels is established based on the multi-source heterogeneous parameter information, and the fine model of the single pixel and four adjacent pixels is simulated. Through simulation, the proportion of electrode potential fluctuations in one pixel caused by a change in the driving voltage of one pixel is statistically analyzed to obtain the capacitive coupling coefficient matrix. Through simulation, using a low-pass filter model, the transfer function of the power consumption of one pixel leading to the temperature rise of another pixel is statistically analyzed, and the thermal diffusivity matrix is ​​obtained. Through simulation, the proportion by which one pixel causes a drop in VDD voltage of another pixel is statistically analyzed to obtain the power supply voltage drop coefficient matrix. Based on the capacitance coupling coefficient matrix, thermal diffusion coefficient matrix, and power supply voltage drop coefficient matrix, a coupling coefficient matrix is ​​formed.

[0008] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, a node set and an edge set are constructed according to the multi-source heterogeneous parameter information and the coupling coefficient matrix, specifically as follows: Pixel nodes, data line nodes, and scan line nodes are constructed based on the multi-source heterogeneous parameter information. A pixel is connected to its four neighboring pixels above, below, left, and right. The edge features are Euclidean distance and capacitive coupling coefficient, forming spatially adjacent edges. Connect all pixel nodes in the same column with the corresponding data line nodes. The edge features are the segmented values ​​of data line resistance and parasitic capacitance, forming a shared edge for data lines. Connect all pixel nodes in the same row to the corresponding scan line node. The edge feature is gate line RC delay, forming a scan line shared edge. Connect the pixel node to the nearest VDD grid node. The edge feature is power supply resistance, forming a power grid edge. Connect the pixel node to all nodes centered at the pixel node with a radius of ≤5 pixels, and use the thermal diffusion coefficient as the edge feature to form thermal diffusion edges, ultimately constructing a set of nodes and a set of edges.

[0009] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, the nodes are initialized with feature encoding, specifically as follows: An input feature vector is constructed for each pixel node. The input feature vector includes the target gray level of the current frame, the actual gray level of the previous frame, the current temperature of the pixel, the accumulated lighting time of the pixel, the initial threshold voltage offset of the pixel, and the average brightness of the local area. For data line nodes, the initial characteristics are the average target voltage of all pixels in the current frame, the actual total current in the previous frame, and the output impedance of the data line driver IC. For scan line nodes, the initial characteristics are scan timing offset and gate drive voltage amplitude.

[0010] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, a liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network, and the liquid crystal display compensation model is trained using the feature encoding, specifically as follows: A liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network, using the feature-encoded features as input. The spatial graph convolution module is used to perform normalization attention processing, node feature update and multi-head attention processing on the features encoded by the feature, and the output is concatenated and dimensionality reduced. The time dimension is modeled as a sequence of frames. For the nth frame, the graph convolution output is a dimension-reduced vector. The hidden state is updated using a graph-gated loop unit, and the final hidden state is passed through a two-layer fully connected network via the output layer to obtain the driving voltage correction for each pixel and the true brightness of each pixel.

[0011] Furthermore, in the dynamic compensation method for a liquid crystal display based on pixel response characteristics, the driving voltage correction amount for each pixel in the current timestamp is output using the liquid crystal display compensation model, specifically as follows: The image data information generated by the LCD screen within a preset time period is obtained, and the image data information generated by the LCD screen within the preset time period is input into the LCD screen compensation model for prediction. By prediction, the driving voltage correction amount and the true brightness of each pixel in the current timestamp are obtained, and the driving voltage correction amount and the true brightness of each pixel in the current timestamp are output.

[0012] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, the model parameters of the liquid crystal display compensation model are adjusted in conjunction with the actual brightness of each pixel, specifically including: Based on the true brightness of each pixel, a predicted brightness distribution is constructed. A random test pattern is injected using a driver chip. At the same time, the actual brightness distribution is read back using an optical sensor array. The error value between the predicted brightness distribution and the actual brightness distribution is calculated. Set an error threshold and determine whether the error value between the predicted brightness distribution and the actual brightness distribution is greater than the error threshold; When the error between the predicted brightness distribution and the actual brightness distribution is greater than the error threshold, backpropagation is used to adjust the model parameters of the liquid crystal display compensation model and update the liquid crystal display compensation model. When the error between the predicted brightness distribution and the actual brightness distribution is not greater than the error threshold, the model parameters of the liquid crystal display compensation model remain unchanged.

[0013] Furthermore, the dynamic compensation method for liquid crystal displays based on pixel response characteristics also includes: Calculate the temperature gradient and brightness gradient between each pixel node and its neighboring pixels. When the temperature gradient is greater than a preset temperature gradient threshold or the brightness gradient is greater than a preset brightness gradient threshold, the corresponding heat diffusion edge is retained; otherwise, it is considered uncoupled. For the retained edges, probability sampling is performed according to the magnitude of the thermal diffusion coefficient, and each pixel retains a maximum of a preset number of thermal diffusion neighbors.

[0014] A second aspect of the present invention provides a dynamic compensation system for a liquid crystal display screen based on pixel response characteristics, comprising a memory and a processor. The memory includes a program for a dynamic compensation method for a liquid crystal display screen based on pixel response characteristics. When the program for a dynamic compensation method for a liquid crystal display screen based on pixel response characteristics is executed by the processor, it implements the steps of the dynamic compensation method for a liquid crystal display screen based on pixel response characteristics as described in any one of the present invention.

[0015] This invention addresses the shortcomings of the prior art and has the following beneficial effects: This invention obtains multi-source heterogeneous parameter information from a panel design database, performs simulations based on this information, extracts the coupling coefficient matrix, and then constructs a node set and an edge set based on the multi-source heterogeneous parameter information and the coupling coefficient matrix. The nodes are then initialized with feature encoding, and a liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network. The model is trained using the feature encoding, and outputs the driving voltage correction amount and the true brightness of each pixel at the current timestamp. Finally, the model parameters are adjusted based on the true brightness of each pixel. This invention can fully consider and express the non-local coupling relationships between pixels that are non-linearly related to physical distance. It models the display screen as a heterogeneous graph, where nodes include pixel units and auxiliary nodes, and edges contain various physical relationships. This enables dynamic prediction and compensation of crosstalk between pixels, improving the pixel compensation accuracy of the liquid crystal display. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0017] Figure 1 A flowchart of the overall process for dynamic compensation of liquid crystal displays based on pixel response characteristics is shown. Figure 2 A system block diagram of a dynamic compensation system for a liquid crystal display based on pixel response characteristics is shown. Detailed Implementation

[0018] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0019] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0020] like Figure 1 As shown, the first aspect of the present invention provides a dynamic compensation method for a liquid crystal display screen based on pixel response characteristics, comprising the following steps: Obtain multi-source heterogeneous parameter information from the panel design database, perform simulation based on the multi-source heterogeneous parameter information, and extract the coupling coefficient matrix; Based on the multi-source heterogeneous parameter information and coupling coefficient matrix, a node set and an edge set are constructed, and the nodes are initialized with feature encoding. A liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network. The model is trained using feature encoding. The model outputs the driving voltage correction amount and the true brightness of each pixel in the current timestamp. The model parameters of the liquid crystal display compensation model are adjusted based on the actual brightness of each pixel, and the driving voltage correction amount of each pixel in the current timestamp is used to dynamically compensate the liquid crystal display.

[0021] It should be noted that the present invention can fully consider and express the non-local coupling relationship between pixels that is non-linearly related to physical distance, and model the display screen as a heterogeneous graph. The nodes include pixel units and auxiliary nodes, and the edges contain a variety of physical relationships. This enables dynamic prediction and compensation of crosstalk between pixels, thereby improving the pixel compensation accuracy of the liquid crystal display screen.

[0022] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, multi-source heterogeneous parameter information is obtained from the panel design database, specifically as follows: At the time of leaving the factory, the design parameter data of the LCD screen is collected, and the pixel array topology, the set of pixel columns connected to each data line, the data line resistance, parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer are obtained from the design parameter data of the LCD screen. Construct a panel design database and store the pixel array topology, the set of pixel columns connected to each data line, the data line resistance, parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer into the panel design database. By utilizing the panel design database, the pixel array topology of the current type of LCD screen, the set of pixel columns connected by each data line, the data line resistance, parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer are extracted to form multi-source heterogeneous parameter information.

[0023] It should be noted that this method can obtain the pixel array topology of the current type of LCD screen, the set of pixel columns connected by each data line, the data line resistance, parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer, thus providing a theoretical basis for describing the coupling relationship.

[0024] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, simulation is performed based on multi-source heterogeneous parameter information to extract the coupling coefficient matrix, specifically: Using the circuit simulation software HSPICE, a fine model of a single pixel and four adjacent pixels is established based on multi-source heterogeneous parameter information, and the fine model of a single pixel and four adjacent pixels is simulated. Through simulation, the proportion of electrode potential fluctuations in one pixel caused by a change in the driving voltage of one pixel is statistically analyzed to obtain the capacitive coupling coefficient matrix. Through simulation, using a low-pass filter model, the transfer function of the power consumption of one pixel leading to the temperature rise of another pixel is statistically analyzed, and the thermal diffusivity matrix is ​​obtained. Through simulation, the proportion by which one pixel causes a drop in VDD voltage of another pixel is statistically analyzed to obtain the power supply voltage drop coefficient matrix. Based on the capacitance coupling coefficient matrix, thermal diffusion coefficient matrix, and power supply voltage drop coefficient matrix, a coupling coefficient matrix is ​​formed.

[0025] It should be noted that the coupling coefficient matrix is ​​an important indicator that can describe the coupling relationship, thereby enabling the subsequent model to dynamically predict and compensate for inter-pixel crosstalk and improve the pixel compensation accuracy of the LCD screen.

[0026] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, a set of nodes and a set of edges are constructed according to multi-source heterogeneous parameter information and coupling coefficient matrix, specifically as follows: Pixel nodes, data line nodes, and scan line nodes are constructed based on multi-source heterogeneous parameter information. A pixel is connected to its four neighboring pixels (up, down, left, and right). The edge features are Euclidean distance and capacitive coupling coefficient, forming spatially adjacent edges. Connect all pixel nodes in the same column with the corresponding data line nodes. The edge features are the segmented values ​​of data line resistance and parasitic capacitance, forming a shared edge for data lines. Connect all pixel nodes in the same row to the corresponding scan line node. The edge feature is gate line RC delay, forming a scan line shared edge. Connect the pixel node to the nearest VDD grid node. The edge feature is power supply resistance, forming a power grid edge. Connect the pixel node to all nodes centered at the pixel node with a radius of ≤5 pixels, and use the thermal diffusion coefficient as the edge feature to form thermal diffusion edges, ultimately constructing a set of nodes and a set of edges.

[0027] It should be noted that the display screen is regarded as a spatiotemporal graph, where pixels are nodes and adjacency relationships are edges. Graph neural networks are used to capture crosstalk, thermal diffusion and electrical coupling effects between pixels to achieve global collaborative driving.

[0028] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, the nodes are initialized with feature encoding, specifically as follows: For each pixel node, an input feature vector is constructed. The input feature vector includes the target gray level of the current frame, the actual gray level of the previous frame, the current temperature of the pixel, the cumulative lit duration of the pixel, the initial threshold voltage offset of the pixel, and the average brightness of the local area. For a data line node, the initial characteristics are the average target voltage of all pixels in the column in the current frame, the actual total current of the column in the previous frame, and the output impedance of the data line driver IC. For scan line nodes, the initial characteristics are scan timing offset and gate drive voltage amplitude.

[0029] It should be noted that the current pixel temperature is calculated from the infrared sensor or thermal model, the actual grayscale of the previous frame is read back from the optical sensor, the initial threshold voltage offset of the pixel is calibrated at the factory, and the average brightness of the local area is used to perceive the ambient contrast.

[0030] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, a liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network, and the model is trained using feature encoding, specifically as follows: A liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network, using the feature-encoded features as input. The spatial graph convolution module is used to perform normalization attention processing, node feature update and multi-head attention processing on the feature-encoded features, and the output is concatenated and then reduced in dimensionality. The time dimension is modeled as a sequence of frames. For the nth frame, the graph convolution output is a dimension-reduced vector. The hidden state is updated using a graph-gated loop unit, and the final hidden state is passed through a two-layer fully connected network via the output layer to obtain the driving voltage correction for each pixel and the true brightness of each pixel.

[0031] It should be noted that by using spatiotemporal graph neural networks to aggregate neighbor pixel features, the transverse electric field effect between pixels (common in LCDs) and current leakage path are learned, and 1D convolution is used along the time dimension to capture the dynamic propagation of crosstalk, thereby obtaining the driving voltage correction amount and the true brightness of each pixel, thus improving the accuracy of LCD compensation.

[0032] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, the driving voltage correction amount for each pixel in the current timestamp is output using the liquid crystal display compensation model, specifically as follows: Acquire image data information generated by the LCD screen within a preset time period, and input the image data information generated by the LCD screen within the preset time period into the LCD screen compensation model for prediction. By prediction, the driving voltage correction amount and the true brightness of each pixel at the current timestamp are obtained, and the driving voltage correction amount and the true brightness of each pixel at the current timestamp are output.

[0033] Furthermore, in the dynamic compensation method for liquid crystal displays based on pixel response characteristics, the model parameters of the liquid crystal display compensation model are adjusted in conjunction with the actual brightness of each pixel, specifically including: Based on the true brightness of each pixel, a predicted brightness distribution is constructed. Random test patterns are injected using a driver chip. At the same time, the actual brightness distribution is read back using an optical sensor array. The error value between the predicted brightness distribution and the actual brightness distribution is calculated. Set an error threshold and determine whether the error between the predicted brightness distribution and the actual brightness distribution is greater than the error threshold. When the error between the predicted brightness distribution and the actual brightness distribution is greater than the error threshold, backpropagation is used to adjust the model parameters of the LCD compensation model and update the LCD compensation model. When the error between the predicted brightness distribution and the actual brightness distribution is not greater than the error threshold, the model parameters of the liquid crystal display compensation model remain unchanged.

[0034] It should be noted that, since environmental temperature, humidity, and mechanical stress can alter coupling parameters in actual use, this method, when the error between the predicted and actual brightness distributions exceeds the error threshold, utilizes backpropagation to adjust the model parameters of the LCD compensation model and update it. This allows for further consideration of the impact of environmental temperature, humidity, and mechanical stress on coupling parameters, thus optimizing the prediction accuracy of the LCD compensation model. Specifically, when adjusting the model parameters of the LCD compensation model, only the graph attention parameters of the last two layers of the ST-GNN are fine-tuned.

[0035] Furthermore, the dynamic compensation method for liquid crystal displays based on pixel response characteristics also includes: Calculate the temperature gradient and brightness gradient between each pixel node and its neighboring pixels. When the temperature gradient is greater than a preset temperature gradient threshold or the brightness gradient is greater than a preset brightness gradient threshold, the corresponding heat diffusion edge is retained; otherwise, it is considered uncoupled. For the retained edges, probability sampling is performed according to the magnitude of the thermal diffusion coefficient, and each pixel retains a maximum of a preset number of thermal diffusion neighbors.

[0036] It should be noted that, due to the thermal diffusion edge, the number of neighbors for each pixel reaches approximately The computational cost of the entire graph is too high; therefore, this method is used to further optimize the computational cost of the model.

[0037] In addition, this method also includes: A generator G and a discriminator D are used, where the generator G is a temporal super-resolution network, and paired data of low frame rate luminance map + high frame rate drive signal and high frame rate luminance map are acquired synchronously using a high-speed camera and a precision oscilloscope. Temporal super-resolution networks include Temporal SRCNN or Video SRGAN.

[0038] The pairing data of the low frame rate luminance map + high frame rate drive signal and the high frame rate luminance map is used as training data. The pixel luminance sequence collected at low temporal resolution (e.g., 60fps) and the corresponding drive voltage signal at high temporal resolution (e.g., 1000fps) are input. The generator G outputs a reconstructed high temporal resolution (e.g., 1000fps) luminance response curve, and the discriminator D is trained to distinguish between the real high frame rate response curve and the curve reconstructed by the generator G. The discriminator D uses 3D convolution to capture the authenticity of dynamic patterns. The generated high frame rate curve is downsampled over time to match the input low frame rate curve.

[0039] It should be noted that the sampling rate of industrial cameras or under-display cameras is much lower than the pixel clock frequency, making it difficult to capture the transient processes of pixel responses at the microsecond / nanosecond level. Existing interpolation methods lose high-frequency dynamic details. However, this method can utilize the adversarial generative capabilities of GANs to recover the true transient waveforms of pixel responses (such as overshoot peaks and oscillation decay processes) from sparse, low-temporal-resolution observations. This enables ordinary devices to have high-speed visual analysis capabilities, greatly reducing the hardware cost of high-precision dynamic response measurement.

[0040] In addition, this method also includes: By configuring a camera device in a preset range area of ​​the LCD screen, the camera device is used to collect image data information of the LCD screen within a preset time period, and feature extraction is performed on the image data information of the LCD screen within the preset time period; By feature extraction, the visual reflective area of ​​the LCD screen during display is obtained, and the position information of each block or pixel is obtained based on the visual reflective area of ​​the LCD screen during display. The position information of each corresponding block or pixel is used as the driving voltage uncorrected region, and the non-visual reflective region is used as the driving voltage corrected region. The driving voltage correction area is corrected by using the correction amount of the driving voltage. At the same time, the visual reflection area of ​​the liquid crystal screen during display is estimated based on the light reflection characteristics to obtain the light source area. Related lighting control prompts are generated based on the light source area.

[0041] It should be noted that, in reality, when affected by external lighting, the LCD screen will be affected by light, resulting in reflective areas. These reflective areas are often not within the area of ​​human visual focus. Therefore, this method combines the actual situation with the correction amount of the driving voltage to correct the driving voltage correction area, thereby improving the rationality of the correction.

[0042] In addition, this method also includes: The screen content data of the LCD screen is acquired and divided into multiple visual regions. The eye gaze score data of the multiple visual regions is collected from the big data network. The visual region corresponding to the eye gaze score data being greater than a preset eye gaze score data threshold is obtained, and the visual region corresponding to the eye gaze score data being greater than the preset eye gaze score data threshold is used as the priority compensation region of the liquid crystal display screen. The visual area corresponding to the eye gaze score data not exceeding the preset eye gaze score data threshold is designated as the non-priority compensation area of ​​the liquid crystal display screen. The priority compensation area of ​​the LCD screen is compensated by adjusting the driving voltage.

[0043] It should be noted that the visual gaze score for the visual region ranges from 0 to 100. A higher score indicates greater focus on the region. Therefore, visual regions with gaze scores exceeding a preset threshold are designated as priority compensation areas for the LCD screen. The compensation for these priority areas is then adjusted using a correction in the driving voltage, further improving the rationality of the compensation. This method combines reflective area recognition and visual gaze scoring, prioritizing computational resources for areas of genuine observer focus, thus enhancing the subjective visual experience and compensation efficiency.

[0044] like Figure 2 As shown, the second aspect of the present invention provides a dynamic compensation system for a liquid crystal display screen based on pixel response characteristics, including a memory and a processor. The memory includes a program for a dynamic compensation method for a liquid crystal display screen based on pixel response characteristics. When the program for a dynamic compensation method for a liquid crystal display screen based on pixel response characteristics is executed by the processor, it implements the steps of any one of the dynamic compensation methods for a liquid crystal display screen based on pixel response characteristics.

[0045] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0046] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0047] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0048] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0049] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0050] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A dynamic compensation method for liquid crystal displays based on pixel response characteristics, characterized in that, Includes the following steps: Obtain multi-source heterogeneous parameter information from the panel design database, perform simulation based on the multi-source heterogeneous parameter information, and extract the coupling coefficient matrix; Based on the multi-source heterogeneous parameter information and coupling coefficient matrix, a node set and an edge set are constructed, and the nodes are initialized with feature encoding. A liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network. The liquid crystal display compensation model is trained using the feature encoding. The liquid crystal display compensation model outputs the driving voltage correction amount and the true brightness of each pixel in the current timestamp. The model parameters of the liquid crystal display compensation model are adjusted by combining the actual brightness of each pixel, and the driving voltage correction amount of each pixel in the current timestamp is used to dynamically compensate the liquid crystal display. Obtain multi-source heterogeneous parameter information from the panel design database, specifically: At the time of manufacture, the design parameter data of the liquid crystal display screen is collected, and the pixel array topology, the set of pixel columns connected to each data line, the data line resistance, the parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer are obtained from the design parameter data of the liquid crystal display screen. Construct a panel design database and store the pixel array topology, the set of pixel columns connected to each data line, the data line resistance, the parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer into the panel design database. Using the panel design database, the pixel array topology of the current type of liquid crystal display, the set of pixel columns connected by each data line, the data line resistance, parasitic capacitance, the resistance of each scan line, the power line grid resistance network, and the interlayer capacitance between each layer are extracted to form multi-source heterogeneous parameter information. Based on the multi-source heterogeneous parameter information and coupling coefficient matrix, a node set and an edge set are constructed, specifically as follows: Pixel nodes, data line nodes, and scan line nodes are constructed based on the multi-source heterogeneous parameter information. A pixel is connected to its four neighboring pixels (up, down, left, and right). The edge features are Euclidean distance, capacitive coupling coefficient, and spatially adjacent edges. Connect all pixel nodes in the same column with the corresponding data line nodes. The edge features are the segmented values ​​of data line resistance and parasitic capacitance, forming a shared edge for data lines. Connect all pixel nodes in the same row to the corresponding scan line node. The edge feature is gate line RC delay, forming a scan line shared edge. Connect the pixel node to the nearest VDD grid node. The edge feature is power supply resistance, forming a power grid edge. Connect the pixel node to all nodes centered at the pixel node with a radius of ≤5 pixels, and use the thermal diffusion coefficient as the edge feature to form thermal diffusion edges, ultimately constructing a set of nodes and a set of edges.

2. The dynamic compensation method for a liquid crystal display screen based on pixel response characteristics according to claim 1, characterized in that, Simulations were performed based on the multi-source heterogeneous parameter information, and the coupling coefficient matrix was extracted, specifically as follows: Using the circuit simulation software HSPICE, a fine model of a single pixel and four adjacent pixels is established based on the multi-source heterogeneous parameter information, and the fine model of the single pixel and four adjacent pixels is simulated. Through simulation, the proportion of electrode potential fluctuations in one pixel caused by a change in the driving voltage of one pixel is statistically analyzed to obtain the capacitive coupling coefficient matrix. Through simulation, using a low-pass filter model, the transfer function of the power consumption of one pixel leading to the temperature rise of another pixel is statistically analyzed, and the thermal diffusivity matrix is ​​obtained. Through simulation, the proportion by which one pixel causes a drop in VDD voltage of another pixel is statistically analyzed to obtain the power supply voltage drop coefficient matrix. Based on the capacitance coupling coefficient matrix, thermal diffusion coefficient matrix, and power supply voltage drop coefficient matrix, a coupling coefficient matrix is ​​formed.

3. The dynamic compensation method for a liquid crystal display screen based on pixel response characteristics according to claim 1, characterized in that, The nodes are initialized with feature encoding, specifically as follows: An input feature vector is constructed for each pixel node. The input feature vector includes the target gray level of the current frame, the actual gray level of the previous frame, the current temperature of the pixel, the accumulated lighting time of the pixel, the initial threshold voltage offset of the pixel, and the average brightness of the local area. For data line nodes, the initial characteristics are the average target voltage of all pixels in the current frame, the actual total current in the previous frame, and the output impedance of the data line driver IC. For scan line nodes, the initial characteristics are scan timing offset and gate drive voltage amplitude.

4. The dynamic compensation method for a liquid crystal display screen based on pixel response characteristics according to claim 1, characterized in that, A liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network, and the model is trained using the aforementioned feature encoding. Specifically: A liquid crystal display compensation model is constructed based on a spatiotemporal graph neural network, using the feature-encoded features as input. The spatial graph convolution module is used to perform normalization attention processing, node feature update and multi-head attention processing on the features encoded by the feature, and the output is concatenated and dimensionality reduced. The time dimension is modeled as a sequence of frames. For the nth frame, the graph convolution output is a dimension-reduced vector. The hidden state is updated using a graph-gated loop unit, and the final hidden state is passed through a two-layer fully connected network via the output layer to obtain the driving voltage correction for each pixel and the true brightness of each pixel.

5. The dynamic compensation method for a liquid crystal display screen based on pixel response characteristics according to claim 1, characterized in that, The driving voltage correction amount for each pixel in the current timestamp is output using the aforementioned liquid crystal display compensation model, specifically as follows: The image data information generated by the LCD screen within a preset time period is obtained, and the image data information generated by the LCD screen within the preset time period is input into the LCD screen compensation model for prediction. By prediction, the driving voltage correction amount and the true brightness of each pixel in the current timestamp are obtained, and the driving voltage correction amount and the true brightness of each pixel in the current timestamp are output.

6. The dynamic compensation method for a liquid crystal display screen based on pixel response characteristics according to claim 1, characterized in that, The model parameters of the liquid crystal display compensation model are adjusted based on the actual brightness of each pixel, specifically including: Based on the true brightness of each pixel, a predicted brightness distribution is constructed. A random test pattern is injected using a driver chip. At the same time, the actual brightness distribution is read back using an optical sensor array. The error value between the predicted brightness distribution and the actual brightness distribution is calculated. Set an error threshold and determine whether the error value between the predicted brightness distribution and the actual brightness distribution is greater than the error threshold; When the error between the predicted brightness distribution and the actual brightness distribution is greater than the error threshold, backpropagation is used to adjust the model parameters of the liquid crystal display compensation model and update the liquid crystal display compensation model. When the error between the predicted brightness distribution and the actual brightness distribution is not greater than the error threshold, the model parameters of the liquid crystal display compensation model remain unchanged.

7. The dynamic compensation method for a liquid crystal display screen based on pixel response characteristics according to claim 1, characterized in that, Also includes: Calculate the temperature gradient and brightness gradient of each pixel node and its neighboring pixels. When the temperature gradient is greater than a preset temperature gradient threshold or the brightness gradient is greater than a preset brightness gradient threshold, the corresponding heat diffusion edge is retained; otherwise, it is considered uncoupled. For the retained edges, probability sampling is performed according to the magnitude of the thermal diffusion coefficient, and each pixel retains a maximum of a preset number of thermal diffusion neighbors.

8. A dynamic compensation system for a liquid crystal display screen based on pixel response characteristics, characterized in that, The device includes a memory and a processor. The memory includes a program for a dynamic compensation method for a liquid crystal display screen based on pixel response characteristics. When the processor executes the program for the dynamic compensation method for a liquid crystal display screen based on pixel response characteristics, it implements the steps of the dynamic compensation method for a liquid crystal display screen based on pixel response characteristics as described in any one of claims 1-7.

Images