Low-latency Mini LED display calibration system for virtual pixels
Patent Information
- Application Number
- CN202611173056.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-04
- Publication Date
- 2026-10-23
AI Technical Summary
[0003]然而，这种静态的离线校正模式在实际应用中，因Mini LED显示屏在工作时，其驱动IC和LED芯片会产生大量热量，且屏体中心区域散热条件差、温度高，而边缘区域散热快、温度低，这种非均匀的温度分布会直接导致LED芯片的发光效率和波长发生漂移，进而使得屏体中心与边缘产生肉眼可见的色偏和亮度差异，最终在拼接处形成顽固的网格残影，然而，现有系统的补偿系数是在室温或特定温度下一次性标定并固化的静态值，其无法感知并补偿这种随工作时间动态变化的温度场差异，导致刚校正完成时画面均匀，但随着屏体温度逐渐升高并趋于稳定，画面反而变得越来越不均匀，校正效果随时间急剧衰减，严重影响了显示屏的长期观看体验和产品质量一致性
[0036]与现有技术相比，本发明的有益效果是：通过部署于Mini LED显示屏背板及拼接缝处的温度传感器阵列，实时采集各测温点的温度值，以生成表征Mini LED显示屏全屏面温度分布状态的实时温度场矩阵，提取屏体中心区域与边缘区域的温度梯度特征，结合预设的热时间常数模型外推预测未来控制周期内各像素区域的温度变化趋势，生成热漂移预补偿量矩阵，结合实时温度场矩阵，根据预存的每一像素的温度与光衰映射关系计算当前温度下的光效偏移量，结合热漂移预补偿量矩阵生成温度补偿系数，并将温度补偿系数与预存的静态校正系数进行加权融合，生成动态校正系数包，并缓存至各接收卡的片内高速存储器中，在实时显示过程中，同步获取输入视频流中各像素点的原始灰阶值，并行调用对应像素的动态校正系数包对原始灰阶值进行乘加修正，生成经温度补偿修正后的目标灰阶值，并根据目标灰阶值输出脉宽调制占空比控制信号以驱动Mini LED显示屏进行实时显示；能够从根本上克服传统静态离线校正系数无法感知和补偿屏体动态热漂移的技术缺陷，有效消除因温度非均匀分布导致的色偏、亮度差异及拼接处网格残影问题，确保MiniLED显示屏在全生命周期、全工况条件下始终保持优异的亮色度一致性和长期观看体验。
Smart Images
Figure CN122967582A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of LED display technology, and more specifically to a low-latency Mini LED display calibration system for virtual pixels. Background Technology
[0002] The core working principle of existing Mini LED display calibration systems is usually an offline calibration mode that first acquires data and then compensates. It uses a high-precision industrial camera to acquire images of the display screen at a specific grayscale level, accurately identifies the brightness and chromaticity information of each virtual pixel, calculates the differential compensation coefficient for each pixel through host computer software, and then compresses and packages these coefficients before sending them down and storing them in the memory of each receiving card in the screen. During normal display, the system calls these coefficients in real time to perform pixel-level correction on the input video stream to improve the brightness and color consistency of the entire screen.
[0003] However, in practical applications, this static offline calibration mode suffers from several drawbacks. Mini LED displays generate significant heat during operation due to the high temperatures and poor heat dissipation in the central area compared to the edges. This uneven temperature distribution directly causes a shift in the luminous efficiency and wavelength of the LED chips, resulting in visible color shifts and brightness differences between the center and edges of the screen. Ultimately, this leads to persistent grid-like afterimages at the splicing points. Furthermore, the compensation coefficients of existing systems are static values calibrated and fixed at room temperature or a specific temperature. These static values cannot detect or compensate for the dynamic temperature field differences that change over time. Consequently, while the image appears uniform immediately after calibration, it becomes increasingly uneven as the screen temperature gradually rises and stabilizes. The calibration effect rapidly diminishes over time, severely impacting the long-term viewing experience and product quality consistency.
[0004] Therefore, in view of this, the present invention proposes a low-latency Mini LED display calibration system for virtual pixels to make up for and improve the shortcomings of the prior art. Summary of the Invention
[0005] To address the aforementioned technical problems, this invention provides a low-latency Mini LED display calibration system for virtual pixels, thereby resolving the corresponding technical issues raised in the background section.
[0006] To achieve the above objectives, the technical solution adopted by this invention is: a low-latency Mini LED display calibration system for virtual pixels, comprising a temperature detection unit, a temperature prediction unit, a coefficient generation unit, and a calibration execution unit, wherein:
[0007] The temperature detection unit is used to collect the temperature values of each temperature measurement point in real time through an array of temperature sensors deployed on the back panel and splicing seams of the Mini LED display, so as to generate a real-time temperature field matrix that characterizes the temperature distribution of the entire Mini LED display.
[0008] The temperature prediction unit is used to receive the real-time temperature field matrix, extract the temperature gradient features of the central and edge regions of the screen, and extrapolate the temperature change trend of each pixel region within the future control cycle by combining the preset thermal time constant model, and generate a thermal drift pre-compensation matrix.
[0009] The coefficient generation unit is used to obtain the thermal drift pre-compensation matrix and the real-time temperature field matrix, calculate the light effect offset at the current temperature based on the pre-stored temperature and light decay mapping relationship of each pixel, generate temperature compensation coefficients by combining the thermal drift pre-compensation matrix, and weight and fuse the temperature compensation coefficients with the pre-stored static correction coefficients to generate a dynamic correction coefficient package.
[0010] The calibration execution unit is used to acquire the dynamic calibration coefficient package and cache it in the on-chip high-speed memory of each receiving card. During real-time display, it synchronously acquires the original grayscale value of each pixel in the input video stream, calls the dynamic calibration coefficient package of the corresponding pixel in parallel to perform multiplication and addition correction on the original grayscale value, generates the target grayscale value after temperature compensation correction, and outputs the pulse width modulation duty cycle control signal according to the target grayscale value to drive the Mini LED display screen for real-time display.
[0011] As a preferred embodiment, the specific process for generating the real-time temperature field matrix characterizing the temperature distribution across the entire surface of the Mini LED display is as follows:
[0012] The raw temperature values at each temperature measurement point are synchronously collected at a preset sampling frequency by a distributed fiber optic temperature sensor array deployed on the back panel of the Mini LED display and at the splicing seams of each cabinet.
[0013] The raw temperature values of each temperature measurement point are processed by outlier removal and interpolation completion. Outlier points that deviate from the average temperature of the neighboring area by more than a preset multiple of the standard deviation range are removed, and the outlier points are replaced and completed by linear interpolation fitting of the temperature values of the neighboring area to obtain the preprocessed temperature matrix.
[0014] Bicubic spline interpolation is performed on the preprocessed temperature matrix to expand the temperature data from a sparse sensor array grid to a dense grid that matches the resolution of the physical pixel array of the Mini LED display, generating a real-time temperature field matrix.
[0015] As a preferred embodiment, the specific process for generating the thermal drift pre-compensation matrix is as follows:
[0016] Obtain the real-time temperature field matrix at the current moment, extract the average temperature of the geometric center region of the screen and the average temperature of the edge region of the screen respectively, and calculate the current temperature gradient feature of the center and edge of the screen.
[0017] The temperature field matrices of multiple historical moments are obtained to form a time-series temperature dataset. The thermal time constant model parameters of the Mini LED display are fitted based on the time-series temperature dataset. The thermal time constant model is mathematically represented by a first-order inertial element plus a pure delay element.
[0018] Based on the real-time temperature field matrix at the current moment, the temperature distribution of each pixel region is predicted after multiple control cycles by recursively extrapolating through the thermal time constant model, thus obtaining the future temperature prediction value of each pixel position.
[0019] Based on the difference between the current temperature and the predicted future temperature at each pixel location, and combined with the preset temperature and light effect drift conversion coefficient, the additional light effect offset expected to occur at each pixel location due to future temperature changes is calculated, and a thermal drift pre-compensation matrix is generated.
[0020] As a preferred embodiment, the specific process for generating the dynamic correction coefficient package is as follows:
[0021] Obtain the real-time temperature field matrix. Based on the pre-stored temperature and light effect offset mapping table for each pixel, query or interpolate to calculate the brightness light effect offset and chromaticity light effect offset vector of each pixel at the current temperature. The chromaticity light effect offset vector includes the red channel offset, green channel offset and blue channel offset.
[0022] Obtain the thermal drift pre-compensation matrix. Based on the temperature and light effect offset mapping table, query or interpolate to calculate the pre-compensation luminance value and pre-compensation chromaticity vector corresponding to each pixel under the thermal drift pre-compensation. Superimpose the light effect offset and the pre-compensation value to generate a temperature compensation coefficient. The temperature compensation coefficient includes luminance temperature compensation value and chromaticity temperature compensation value.
[0023] The static brightness correction coefficient and static chromaticity correction coefficient of each pixel are obtained from a preset static correction coefficient database. The static correction coefficients are obtained and stored in advance through offline calibration under the factory calibration temperature conditions of the Mini LED display.
[0024] The static brightness correction coefficient of each pixel is weighted and fused with the brightness temperature compensation value to generate the dynamic brightness correction coefficient.
[0025] The static chromaticity correction coefficients of each pixel are weighted and fused with the chromaticity temperature compensation value to generate dynamic chromaticity correction coefficients.
[0026] The dynamic luminance correction coefficient and the dynamic chrominance correction coefficient are encapsulated into a dynamic correction coefficient package according to a preset data structure.
[0027] As a preferred embodiment, the specific process of outputting a pulse width modulation duty cycle control signal based on the target grayscale value to drive the Mini LED display for real-time display is as follows:
[0028] The dynamic correction coefficient package is obtained, and according to the pixel area range driven by each receiving card of the Mini LED display, the dynamic correction coefficient package is divided into several sub-coefficient packages. Each sub-coefficient package is then sent down and cached in the on-chip high-speed static random access memory of the corresponding receiving card.
[0029] During real-time display, the frame synchronization signal of the input video stream is received synchronously, and the original grayscale value and original chroma value of each pixel in the current frame are parsed in parallel within each display frame cycle. The original chroma value includes the original red value, the original green value and the original blue value.
[0030] Within the same clock cycle, the dynamic correction coefficient packets of the corresponding pixels are read in parallel from the on-chip high-speed static random access memory of each receiving card, and the dynamic brightness correction coefficients and dynamic chromaticity correction coefficients of the corresponding pixels are obtained by parsing.
[0031] The original grayscale value of each pixel is multiplied and added with the dynamic brightness correction coefficient to generate the target brightness grayscale value after temperature compensation correction.
[0032] The original chromaticity values of each primary color of each pixel are multiplied and added with the corresponding chromaticity correction components in the dynamic chromaticity correction coefficient to generate the target chromaticity grayscale value after temperature compensation correction.
[0033] Based on the target brightness grayscale value and the target chromaticity grayscale value, and based on the grayscale and brightness characteristic curves of each primary color sub-pixel of the Mini LED display, a pulse width modulation duty cycle control signal corresponding to each primary color sub-pixel at each pixel position is generated and output to the driving circuit of the Mini LED display to drive each primary color sub-pixel to emit light according to the corresponding duty cycle.
[0034] Preferably, in the temperature detection unit:
[0035] In the temperature sensor array, the temperature sensors deployed on the back panel of the Mini LED display are evenly distributed in a rectangular grid, and the temperature sensors deployed at the splicing seams of each cabinet of the Mini LED display are distributed in a linear, equally spaced manner along the splicing seam. Each temperature sensor in the temperature sensor array is a fiber Bragg grating sensor.
[0036] Compared with existing technologies, the beneficial effects of this invention are as follows: By deploying a temperature sensor array at the back panel and splicing seams of the Mini LED display, the temperature values of each temperature measurement point are collected in real time to generate a real-time temperature field matrix characterizing the temperature distribution state of the entire Mini LED display screen. The temperature gradient characteristics of the central and edge areas of the screen are extracted. Combined with a preset thermal time constant model, the temperature change trend of each pixel area within the future control cycle is extrapolated to generate a thermal drift pre-compensation matrix. Combined with the real-time temperature field matrix, the luminous efficacy offset at the current temperature is calculated according to the pre-stored temperature-light decay mapping relationship of each pixel. The temperature compensation coefficient is generated by combining the thermal drift pre-compensation matrix. The temperature compensation coefficient is weighted and fused with the pre-stored static correction coefficient to generate a dynamic correction coefficient package, which is cached in the on-chip high-speed memory of each receiving card. During real-time display, the original grayscale values of each pixel in the input video stream are acquired synchronously. The dynamic correction coefficient package of the corresponding pixel is called in parallel to perform multiplication and addition correction on the original grayscale values to generate the target grayscale value after temperature compensation correction. The pulse width modulation duty cycle control signal is output according to the target grayscale value to drive the Mini LED display. LED displays provide real-time display; they can fundamentally overcome the technical defects of traditional static offline correction coefficients that cannot sense and compensate for dynamic thermal drift of the screen, effectively eliminate color deviation, brightness differences and grid afterimage at splicing points caused by non-uniform temperature distribution, and ensure that MiniLED displays maintain excellent brightness and color consistency and long-term viewing experience throughout their entire life cycle and under all working conditions. Attached Figure Description
[0037] Figure 1 This is a schematic diagram of the overall structure of a preferred embodiment of the present invention;
[0038] Figure 2 This is a flowchart of the dynamic temperature compensation method for Mini LED displays according to the present invention. Detailed Implementation
[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0040] Embodiments of the present invention:
[0041] Please refer to Figures 1 to 2 As shown, the low-latency Mini LED display calibration system for virtual pixels includes a temperature detection unit, a temperature prediction unit, a coefficient generation unit, and a calibration execution unit, wherein:
[0042] The temperature detection unit is used to collect the temperature values of each temperature measurement point in real time through an array of temperature sensors deployed on the back panel and splicing seams of the Mini LED display, so as to generate a real-time temperature field matrix that characterizes the temperature distribution of the entire Mini LED display.
[0043] The temperature prediction unit is used to receive the real-time temperature field matrix, extract the temperature gradient features of the central and edge regions of the screen, and extrapolate the temperature change trend of each pixel region within the future control cycle by combining the preset thermal time constant model, and generate a thermal drift pre-compensation matrix.
[0044] The coefficient generation unit is used to obtain the thermal drift pre-compensation matrix and the real-time temperature field matrix, calculate the light effect offset at the current temperature based on the pre-stored temperature and light decay mapping relationship of each pixel, generate temperature compensation coefficients by combining the thermal drift pre-compensation matrix, and weight and fuse the temperature compensation coefficients with the pre-stored static correction coefficients to generate a dynamic correction coefficient package.
[0045] The calibration execution unit is used to acquire the dynamic calibration coefficient package and cache it in the on-chip high-speed memory of each receiving card. During real-time display, it synchronously acquires the original grayscale value of each pixel in the input video stream, calls the dynamic calibration coefficient package of the corresponding pixel in parallel to perform multiplication and addition correction on the original grayscale value, generates the target grayscale value after temperature compensation correction, and outputs the pulse width modulation duty cycle control signal according to the target grayscale value to drive the Mini LED display screen for real-time display.
[0046] The specific process for generating the real-time temperature field matrix characterizing the temperature distribution across the entire surface of a Mini LED display is as follows:
[0047] The raw temperature values at each temperature measurement point are synchronously collected at a preset sampling frequency by a distributed fiber optic temperature sensor array deployed on the back panel of the Mini LED display and at the splicing seams of each cabinet.
[0048] The raw temperature values of each temperature measurement point are processed by outlier removal and interpolation completion. Outlier points that deviate from the average temperature of the neighboring area by more than a preset multiple of the standard deviation range are removed, and the outlier points are replaced and completed by linear interpolation fitting of the temperature values of the neighboring area to obtain the preprocessed temperature matrix.
[0049] Bicubic spline interpolation is performed on the preprocessed temperature matrix to expand the temperature data from a sparse sensor array grid to a dense grid that matches the resolution of the physical pixel array of the Mini LED display, generating a real-time temperature field matrix.
[0050] The Mini LED display screen was placed in a programmable temperature control chamber and operated stably for no less than 30 minutes at six temperature points: -10℃, 0℃, 25℃, 40℃, 60℃, and 80℃. Then, the brightness and color coordinate values of each pixel were measured at each temperature point using an industrial-grade spectroradiometer. The measurement value at 25℃ was used as the reference value. The brightness attenuation ratio of each pixel relative to the reference value at each temperature point and the color offset of the red, green, and blue channels were calculated. The results were stored in a lookup table format according to a two-dimensional index structure of temperature and pixels. When the actual query temperature is between two calibrated temperature points, the luminous efficacy offset of the intermediate temperature point was calculated using linear interpolation.
[0051] The specific process for generating the thermal drift pre-compensation matrix is as follows:
[0052] Obtain the real-time temperature field matrix at the current moment, extract the average temperature of the geometric center region of the screen and the average temperature of the edge region of the screen respectively, and calculate the current temperature gradient feature of the center and edge of the screen.
[0053] The temperature field matrices of multiple historical moments are obtained to form a time-series temperature dataset. The thermal time constant model parameters of the Mini LED display are fitted based on the time-series temperature dataset. The thermal time constant model is mathematically represented by a first-order inertial element plus a pure delay element.
[0054] Based on the real-time temperature field matrix at the current moment, the temperature distribution of each pixel region is predicted after multiple control cycles by recursively extrapolating through the thermal time constant model, thus obtaining the future temperature prediction value of each pixel position.
[0055] Based on the difference between the current temperature and the predicted future temperature at each pixel location, and combined with the preset temperature and light effect drift conversion coefficient, the additional light effect offset expected to occur at each pixel location due to future temperature changes is calculated, and a thermal drift pre-compensation matrix is generated.
[0056] The thermal time constant τ in the thermal time constant model is dynamically obtained through online identification. When the Mini LED display is powered on for the first time or when the screen temperature change rate exceeds the preset threshold, the parameter identification program is triggered. The system continuously records the temperature rise curve of the screen from the current temperature to the thermal equilibrium temperature. The least squares method is used to fit the curve of the first-order inertial element transfer function to identify the τ value under the current operating conditions. Since the thermal characteristics of different display sizes, different installation environments, and different playback content are significantly different, the dynamic identification of the τ value can ensure that the temperature prediction model always matches the current actual operating conditions, thereby improving the prediction accuracy.
[0057] The specific process for generating the dynamic correction coefficient package is as follows:
[0058] The real-time temperature field matrix is obtained. Based on the pre-stored temperature and light effect offset mapping table for each pixel, the brightness and chromaticity offset vectors of each pixel at the current temperature are calculated by querying or interpolation. The chromaticity offset vector includes the red channel offset, green channel offset, and blue channel offset. The temperature and light effect offset mapping table is obtained and stored in advance through high and low temperature environment calibration experiments before leaving the factory.
[0059] Obtain the thermal drift pre-compensation matrix. Based on the temperature and light effect offset mapping table, query or interpolate to calculate the pre-compensation luminance value and pre-compensation chrominance vector corresponding to each pixel under the thermal drift pre-compensation. Superimpose the light effect offset and the pre-compensation value to generate the temperature compensation coefficient, which includes luminance temperature compensation value and chrominance temperature compensation value.
[0060] The static brightness correction coefficient and static chromaticity correction coefficient of each pixel are obtained from the preset static correction coefficient database. The static correction coefficients are obtained and stored in advance through offline calibration under the factory calibration temperature of the Mini LED display.
[0061] The static brightness correction coefficient of each pixel is weighted and fused with the brightness temperature compensation value to generate the dynamic brightness correction coefficient.
[0062] The static chromaticity correction coefficients of each pixel are weighted and fused with the chromaticity temperature compensation value to generate dynamic chromaticity correction coefficients.
[0063] The dynamic luminance correction coefficient and the dynamic chrominance correction coefficient are encapsulated into a dynamic correction coefficient package according to a preset data structure.
[0064] The temperature sensor array uses fiber Bragg grating temperature sensors, which are deployed in series and multiplexed at the back panel of the Mini LED display and at the splicing seams of each cabinet. The fiber Bragg grating sensors deployed on the back panel are evenly distributed in a rectangular grid with a grid spacing of no more than 10cm. The fiber Bragg grating sensors deployed at the splicing seams are evenly distributed in a linear pattern along the splicing seam, with the spacing between adjacent sensors not exceeding 5cm. Compared with traditional thermistor or thermocouple sensors, fiber Bragg grating sensors have the advantages of being resistant to electromagnetic interference, small in size, high in accuracy, and allowing multiple grating points to be multiplexed in series on a single fiber. They are especially suitable for deployment in scenarios where the back panel space of the Mini LED display is limited and the electromagnetic environment of the driving circuit is complex.
[0065] The specific process of outputting a pulse width modulation duty cycle control signal based on the target grayscale value to drive the Mini LED display for real-time display is as follows:
[0066] The dynamic correction coefficient package is obtained, and according to the pixel area range driven by each receiving card of the Mini LED display, the dynamic correction coefficient package is divided into several sub-coefficient packages. Each sub-coefficient package is then sent down and cached in the on-chip high-speed static random access memory of the corresponding receiving card.
[0067] During real-time display, the frame synchronization signal of the input video stream is received synchronously, and the original grayscale value and original chroma value of each pixel in the current frame are parsed in parallel within each display frame cycle. The original chroma value includes the original red value, the original green value and the original blue value.
[0068] Within the same clock cycle, the dynamic correction coefficient packets of the corresponding pixels are read in parallel from the on-chip high-speed static random access memory of each receiving card, and the dynamic brightness correction coefficients and dynamic chromaticity correction coefficients of the corresponding pixels are obtained by parsing.
[0069] The original grayscale value of each pixel is multiplied and added with the dynamic brightness correction coefficient to generate the target brightness grayscale value after temperature compensation correction.
[0070] The original chromaticity values of each primary color of each pixel are multiplied and added with the corresponding chromaticity correction components in the dynamic chromaticity correction coefficient to generate the target chromaticity grayscale value after temperature compensation correction.
[0071] Based on the target brightness grayscale value and the target chromaticity grayscale value, and based on the grayscale and brightness characteristic curves of each primary color sub-pixel of the Mini LED display, a pulse width modulation duty cycle control signal corresponding to each primary color sub-pixel at each pixel position is generated and output to the driving circuit of the Mini LED display to drive each primary color sub-pixel to emit light according to the corresponding duty cycle.
[0072] The pulse width modulation duty cycle control signal is generated as follows: based on the grayscale and brightness characteristic curves of each primary color sub-pixel of the Mini LED display, the target brightness grayscale value and the target color grayscale value are mapped to the corresponding pulse width modulation duty cycle value respectively.
[0073] The grayscale and brightness characteristic curves are pre-acquired using a pixel-by-pixel calibration method. That is, for each sub-pixel, its actual brightness value is measured at grayscale levels from 0 to 255, forming a unique grayscale and brightness mapping curve for that sub-pixel.
[0074] The duty cycle of the pulse width modulation duty cycle control signal has a positive nonlinear mapping relationship with the target gray level value to ensure that the final actual output brightness has a linear correspondence with the input gray level value, thereby avoiding brightness distortion caused by the nonlinear electro-optic conversion of the LED chip itself.
[0075] In the temperature detection unit:
[0076] In the temperature sensor array, the temperature sensors deployed on the back panel of the Mini LED display are evenly distributed in a rectangular grid, and the temperature sensors deployed at the splicing seams of each cabinet of the Mini LED display are distributed in a linear pattern with equal spacing along the splicing seam. Each temperature sensor in the temperature sensor array is a fiber Bragg grating sensor.
[0077] By setting up a temperature detection unit and deploying a temperature sensor array on the back panel and splicing seams of the Mini LED display, the temperature values of each temperature measurement point on the entire screen are collected in real time and a real-time temperature field matrix is generated. This overcomes the technical blind spot of traditional offline calibration systems that do not know the working temperature distribution of the screen, and provides an accurate data foundation for subsequent dynamic compensation, enabling the calibration system to sense non-uniform temperature fields.
[0078] By setting up a temperature prediction unit, the temperature gradient features of the central and edge regions of the screen are extracted. Combined with the thermal time constant model, the temperature change trend of each pixel region in the future control cycle is extrapolated and a thermal drift pre-compensation matrix is generated. The thermal drift development trend is predicted in advance, overcoming the lag defect of traditional schemes that can only compensate after the fact. This makes the temperature compensation forward-looking and proactive, and significantly improves the timeliness and accuracy of the compensation.
[0079] By setting up a coefficient generation unit, based on the real-time temperature field matrix and the thermal drift pre-compensation matrix, the light effect offset is calculated according to the temperature and light decay mapping relationship of each pixel, and a temperature compensation coefficient is generated. Then, the temperature compensation coefficient is weighted and fused with the pre-stored static correction coefficient to generate a dynamic correction coefficient package. This breaks through the technical limitations of traditional static coefficients and realizes the adaptive dynamic update of correction coefficients as the screen temperature changes. It eliminates the color shift, brightness difference and grid ghosting problems caused by non-uniform temperature distribution from the root, and ensures the long-term display uniformity and consistency of the screen during continuous operation.
[0080] By setting up a correction execution unit, the dynamic correction coefficient package is cached in the on-chip high-speed memory of each receiving card. During real-time display, the original grayscale value of each pixel in the input video stream is obtained synchronously, and the dynamic correction coefficient package of the corresponding pixel is called in parallel to perform multiplication and addition correction and output pulse width modulation duty cycle control signal. While completing dynamic temperature compensation, low-latency real-time processing of the input video stream is guaranteed, taking into account both correction accuracy and display real-time performance, and meeting the high standard display requirements of virtual pixel low-latency Mini LED display.
[0081] The size of the interval and threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value.
[0082] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.
[0083] In the two embodiments provided in this application, it should be understood that the disclosed apparatus and system can be implemented in other ways; for example, the apparatus embodiments described above are merely illustrative, and the division of modules is merely a logical functional division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed; furthermore, the coupling or direct coupling or communication connection between the shown or discussed mutuals can be through some interfaces, and the indirect coupling or communication connection between the apparatus or modules can be electrical, mechanical or other forms.
[0084] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A low-latency Mini LED display calibration system for virtual pixels, characterized in that, It includes a temperature detection unit, a temperature prediction unit, a coefficient generation unit, and a correction execution unit, wherein: The temperature detection unit is used to collect the temperature values of each temperature measurement point in real time through an array of temperature sensors deployed on the back panel and splicing seams of the Mini LED display, so as to generate a real-time temperature field matrix that characterizes the temperature distribution of the entire Mini LED display. The temperature prediction unit is used to receive the real-time temperature field matrix, extract the temperature gradient features of the central and edge regions of the screen, and extrapolate the temperature change trend of each pixel region within the future control cycle by combining the preset thermal time constant model, and generate a thermal drift pre-compensation matrix. The coefficient generation unit is used to obtain the thermal drift pre-compensation matrix and the real-time temperature field matrix, calculate the light effect offset at the current temperature based on the pre-stored temperature and light decay mapping relationship of each pixel, generate temperature compensation coefficients by combining the thermal drift pre-compensation matrix, and weight and fuse the temperature compensation coefficients with the pre-stored static correction coefficients to generate a dynamic correction coefficient package. The calibration execution unit is used to acquire the dynamic calibration coefficient package and cache it in the on-chip high-speed memory of each receiving card. During real-time display, it synchronously acquires the original grayscale value of each pixel in the input video stream, calls the dynamic calibration coefficient package of the corresponding pixel in parallel to perform multiplication and addition correction on the original grayscale value, generates the target grayscale value after temperature compensation correction, and outputs the pulse width modulation duty cycle control signal according to the target grayscale value to drive the Mini LED display screen for real-time display.
2. The low-latency Mini LED display calibration system for virtual pixels according to claim 1, characterized in that, The specific process for generating the real-time temperature field matrix characterizing the temperature distribution across the entire surface of a Mini LED display is as follows: The raw temperature values at each temperature measurement point are synchronously collected at a preset sampling frequency by a distributed fiber optic temperature sensor array deployed on the back panel of the Mini LED display and at the splicing seams of each cabinet. The raw temperature values of each temperature measurement point are processed by outlier removal and interpolation completion. Outlier points that deviate from the average temperature of the neighboring area by more than a preset multiple of the standard deviation range are removed, and the outlier points are replaced and completed by linear interpolation fitting of the temperature values of the neighboring area to obtain the preprocessed temperature matrix. Bicubic spline interpolation is performed on the preprocessed temperature matrix to expand the temperature data from a sparse sensor array grid to a dense grid that matches the resolution of the physical pixel array of the Mini LED display, generating a real-time temperature field matrix. The Mini LED display screen was placed in a programmable temperature control chamber and operated stably for no less than 30 minutes at six temperature points: -10℃, 0℃, 25℃, 40℃, 60℃, and 80℃. Then, the brightness and color coordinate values of each pixel were measured at each temperature point using an industrial-grade spectroradiometer. The measurement value at 25℃ was used as the reference value. The brightness attenuation ratio of each pixel relative to the reference value at each temperature point and the color offset of the red, green, and blue channels were calculated. The results were stored in a lookup table format according to a two-dimensional index structure of temperature and pixels. When the actual query temperature is between two calibrated temperature points, the luminous efficacy offset of the intermediate temperature point was calculated using linear interpolation.
3. The low-latency Mini LED display calibration system for virtual pixels according to claim 1, characterized in that, The specific process for generating the thermal drift pre-compensation matrix is as follows: Obtain the real-time temperature field matrix at the current moment, extract the average temperature of the geometric center region of the screen and the average temperature of the edge region of the screen respectively, and calculate the current temperature gradient feature of the center and edge of the screen. The temperature field matrices of multiple historical moments are obtained to form a time-series temperature dataset. The thermal time constant model parameters of the Mini LED display are fitted based on the time-series temperature dataset. The thermal time constant model is mathematically represented by a first-order inertial element plus a pure delay element. Based on the real-time temperature field matrix at the current moment, the temperature distribution of each pixel region is predicted after multiple control cycles by recursively extrapolating through the thermal time constant model, thus obtaining the future temperature prediction value of each pixel position. Based on the difference between the current temperature and the predicted future temperature at each pixel location, and combined with the preset temperature and light effect drift conversion coefficient, the additional light effect offset expected to occur at each pixel location due to future temperature changes is calculated, and a thermal drift pre-compensation matrix is generated. The thermal time constant τ in the thermal time constant model is dynamically obtained through online identification. When the Mini LED display is powered on for the first time or when the screen temperature change rate exceeds a preset threshold, the parameter identification program is triggered. The system continuously records the temperature rise curve of the screen from the current temperature to the thermal equilibrium temperature. The least squares method is used to fit the curve of the first-order inertial element transfer function to identify the τ value under the current operating conditions. Since the thermal characteristics of different display sizes, different installation environments, and different playback content are significantly different, the dynamic identification of the τ value can ensure that the temperature prediction model always matches the current actual operating conditions, thereby improving the prediction accuracy.
4. The low-latency Mini LED display calibration system for virtual pixels according to claim 1, characterized in that, The specific process for generating the dynamic correction coefficient package is as follows: The real-time temperature field matrix is obtained. Based on the pre-stored temperature-to-luminous effect offset mapping table for each pixel, the luminance and chrominance luminous effect offset vectors of each pixel at the current temperature are calculated by querying or interpolation. The chrominance luminous effect offset vector includes the red channel offset, green channel offset, and blue channel offset. The temperature-to-luminous effect offset mapping table is pre-obtained and stored through high and low temperature environment calibration experiments before leaving the factory. Obtain the thermal drift pre-compensation matrix. Based on the temperature and light effect offset mapping table, query or interpolate to calculate the pre-compensation luminance value and pre-compensation chromaticity vector corresponding to each pixel under the thermal drift pre-compensation. Superimpose the light effect offset and the pre-compensation value to generate a temperature compensation coefficient. The temperature compensation coefficient includes luminance temperature compensation value and chromaticity temperature compensation value. The static brightness correction coefficient and static chromaticity correction coefficient of each pixel are obtained from a preset static correction coefficient database. The static correction coefficients are obtained and stored in advance through offline calibration under the factory calibration temperature conditions of the Mini LED display. The static brightness correction coefficient of each pixel is weighted and fused with the brightness temperature compensation value to generate the dynamic brightness correction coefficient. The static chromaticity correction coefficients of each pixel are weighted and fused with the chromaticity temperature compensation value to generate dynamic chromaticity correction coefficients. The dynamic luminance correction coefficient and the dynamic chrominance correction coefficient are encapsulated into a dynamic correction coefficient package according to a preset data structure. The temperature sensor array employs fiber Bragg grating temperature sensors, which are deployed in series and multiplexed at the back panel of the Mini LED display and at the splicing seams of each cabinet. The fiber Bragg grating sensors deployed on the back panel are evenly distributed in a rectangular grid pattern with a grid spacing of no more than 10 cm. The fiber Bragg grating sensors deployed at the splicing seams are evenly distributed in a linear pattern along the splicing seam, with the spacing between adjacent sensors not exceeding 5 cm. Compared with traditional thermistor or thermocouple sensors, fiber Bragg grating sensors have advantages such as resistance to electromagnetic interference, small size, high accuracy, and the ability to multiple grating points in series on a single fiber. They are particularly suitable for deployment in scenarios where the back panel space of a Mini LED display is limited and the electromagnetic environment of the driving circuit is complex.
5. The low-latency Mini LED display calibration system for virtual pixels according to claim 1, characterized in that, The specific process of outputting a pulse width modulation duty cycle control signal based on the target grayscale value to drive the Mini LED display for real-time display is as follows: The dynamic correction coefficient package is obtained, and according to the pixel area range driven by each receiving card of the Mini LED display, the dynamic correction coefficient package is divided into several sub-coefficient packages. Each sub-coefficient package is then sent down and cached in the on-chip high-speed static random access memory of the corresponding receiving card. During real-time display, the frame synchronization signal of the input video stream is received synchronously, and the original grayscale value and original chroma value of each pixel in the current frame are parsed in parallel within each display frame cycle. The original chroma value includes the original red value, the original green value and the original blue value. Within the same clock cycle, the dynamic correction coefficient packets of the corresponding pixels are read in parallel from the on-chip high-speed static random access memory of each receiving card, and the dynamic brightness correction coefficients and dynamic chromaticity correction coefficients of the corresponding pixels are obtained by parsing. The original grayscale value of each pixel is multiplied and added with the dynamic brightness correction coefficient to generate the target brightness grayscale value after temperature compensation correction. The original chromaticity values of each primary color of each pixel are multiplied and added with the corresponding chromaticity correction components in the dynamic chromaticity correction coefficient to generate the target chromaticity grayscale value after temperature compensation correction. Based on the target brightness grayscale value and the target chromaticity grayscale value, and based on the grayscale and brightness characteristic curves of each primary color sub-pixel of the Mini LED display, a pulse width modulation duty cycle control signal corresponding to each primary color sub-pixel at each pixel position is generated and output to the driving circuit of the Mini LED display to drive each primary color sub-pixel to emit light according to the corresponding duty cycle. The pulse width modulation duty cycle control signal is generated as follows: based on the grayscale and brightness characteristic curves of each primary color sub-pixel of the Mini LED display, the target brightness grayscale value and the target color grayscale value are mapped to the corresponding pulse width modulation duty cycle value respectively. The grayscale and brightness characteristic curves are pre-acquired using a pixel-by-pixel calibration method. That is, for each sub-pixel, its actual brightness value is measured at grayscale levels from 0 to 255, forming a unique grayscale and brightness mapping curve for that sub-pixel. The duty cycle of the pulse width modulation duty cycle control signal has a positive nonlinear mapping relationship with the target gray level value, so as to ensure that the final actual output brightness has a linear correspondence with the input gray level value, thereby avoiding brightness distortion caused by the nonlinear electro-optic conversion of the LED chip itself.
6. The low-latency Mini LED display calibration system for virtual pixels according to claim 1, characterized in that, In the temperature detection unit: In the temperature sensor array, the temperature sensors deployed on the back panel of the Mini LED display are evenly distributed in a rectangular grid, and the temperature sensors deployed at the splicing seams of each cabinet of the Mini LED display are distributed in a linear, equally spaced manner along the splicing seam. Each temperature sensor in the temperature sensor array is a fiber Bragg grating sensor.