LED spliced screen cloud correction method, device and equipment and storage medium

Abstract

The invention discloses an LED spliced screen cloud correction method, device and equipment and a storage medium, and relates to the technical field of data processing, and the method comprises the steps: obtaining spliced screen data, and carrying out the single-screen basic correction of a screen module based on the spliced screen data; after the single-screen basic correction is completed, monitoring whether boundary mismatch exists or not through an edge computing node; the boundary mismatch comprises deviation of brightness and / or chromaticity of edge pixels of adjacent screen modules; if the boundary mismatch exists, uploading boundary data to the cloud through the edge computing node; and obtaining a boundary correction parameter based on a pre-trained boundary correction model in the cloud, and receiving the boundary correction parameter to carry out cross-screen compensation. Through cloud intelligent compensation, macroscopic splicing seams of the spliced screen are eliminated, the display quality is improved, the environmental factor influence is dynamically compensated, the adaptive capacity of the spliced screen is improved, the resource utilization efficiency is improved through combined utilization of edge calculation and a cloud model, and the operation and maintenance cost is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to an LED splicing screen cloud correction method, device and equipment and storage medium. BACKGROUND

[0002] In the field of LED display technology, splicing screen systems are widely used in high-performance scenarios such as performance projection and broadcast performance, and the display consistency directly affects the visual presentation effect. The existing technology usually adopts a localized single module basic correction scheme: the edge node collects module brightness / chromaticity data, generates correction parameters combined with local aging compensation algorithms, and burns them to the receiver card.

[0003] However, the local edge node is limited by hardware resources, resulting in limited pixel-level brightness and chromaticity compensation accuracy in the splicing seam area; single module correction only relies on its own historical data and cannot fuse the spatio-temporal correlation characteristics of multiple modules across screens, resulting in brightness jump and chromaticity fault in the splicing seam; at the same time, the local algorithm lacks the ability to respond to environmental disturbances in real time, and the compensation coefficients are fixed in the hardware memory, which cannot be dynamically iterated according to the running state, resulting in rapid decay of the correction effect over time.

[0004] Therefore, the existing technology is limited by the computing power constraints of the local architecture and cannot meet the needs of high-precision and long-time display screen splicing seam elimination.

[0005] The above content is only used to assist in understanding the technical solutions of the present application and does not represent the acknowledgement of the above content as prior art. SUMMARY

[0006] The main purpose of the present application is to provide an LED splicing screen cloud correction method, device and equipment and storage medium, which aims to solve the technical problem that high-precision and long-time display screen splicing seam elimination cannot be met.

[0007] To achieve the above purpose, the present application provides an LED splicing screen cloud correction method, which comprises: Obtaining splicing screen data, and performing single-screen basic correction of screen modules based on the splicing screen data; After the single-screen basic correction is completed, whether there is a boundary mismatch is monitored by an edge computing node; the boundary mismatch includes deviation of brightness and / or chromaticity of the edge pixels of adjacent screen modules; If there is a boundary mismatch, the edge computing node uploads boundary data to the cloud; Based on the pre-trained boundary correction model in the cloud, boundary correction parameters are obtained, and the boundary correction parameters are received for cross-screen compensation.

[0008] In an embodiment, the step of obtaining the spliced screen data and performing single-screen basic correction of the screen module based on the spliced screen data comprises: obtaining a current brightness value, a current chrominance value, and a first working time of each screen module; uploading the current brightness value, the current chrominance value, and the first working time to the cloud; obtaining a brightness gain coefficient and / or a chrominance gain coefficient corresponding to the screen module through a pixel aging model pre-trained in the cloud; receiving the brightness gain coefficient and / or the chrominance gain coefficient to perform single-screen basic correction on the corresponding screen module, and storing the corrected data after the single-screen basic correction.

[0009] In an embodiment, the step of monitoring whether there is a boundary mismatch through the edge computing node after the single-screen basic correction is completed comprises: extracting a brightness correction value of each pixel in a target area of the spliced screen module; calculating an average corrected brightness of the target area based on the brightness correction value; calculating a brightness standard deviation of the target area through the average corrected brightness; comparing the brightness standard deviation with a preset brightness standard deviation threshold; if the brightness standard deviation is greater than the preset brightness standard deviation threshold, it is determined that there is a boundary mismatch, and the target area is marked as an abnormal area; otherwise, the brightness correction value is maintained.

[0010] In an embodiment, the step of monitoring whether there is a boundary mismatch through the edge computing node after the single-screen basic correction is completed further comprises: extracting a chrominance correction value of each pixel in a target area of the spliced screen module; calculating a color deviation based on the chrominance correction value and a pre-stored reference chrominance value; if the color deviation is greater than a preset color deviation threshold, the target area is marked as an abnormal area; otherwise, the chrominance correction value is maintained.

[0011] In an embodiment, the boundary correction model comprises a brightness compensation model and a chrominance compensation model; and the step of obtaining a boundary correction parameter based on the pre-trained boundary correction model in the cloud comprises: receiving the boundary data and pre-processing the boundary data; extracting an adjacent pixel brightness gradient and a screen module aging coefficient based on the pre-processed boundary data; inputting the adjacent pixel luminance gradient and the screen module aging coefficient into the luminance compensation model to generate a luminance compensation gradient field corresponding to the abnormal area; the luminance compensation gradient field is a luminance correction parameter of the boundary correction parameter.

[0012] In an embodiment, the boundary correction model comprises a luminance compensation model and a chrominance compensation model; the step of obtaining the boundary correction parameter based on the pre-trained boundary correction model in the cloud further comprises: receiving the boundary data, and performing trichromatic separation and feature extraction based on the boundary data; inputting the data obtained through feature extraction into the chrominance compensation model to calculate red / blue / green compensation coefficients respectively through channels; optimizing the red / blue / green compensation coefficients based on white balance constraints, color gamut constraints and visual smoothness constraints to obtain final gain adjustment coefficients; the final gain adjustment coefficients are chrominance correction parameters of the boundary correction parameter.

[0013] In an embodiment, the step of receiving the boundary correction parameter to perform cross-screen compensation further comprises: checking whether the luminance standard deviation and / or color deviation of the abnormal area after cross-screen compensation meet the standards; if the luminance standard deviation and / or color deviation do not meet the standards, marking the abnormal area that does not meet the standards as a secondary abnormal area, and performing cloud diagnosis or strategy iteration on the secondary abnormal area.

[0014] In addition, to achieve the above-mentioned purpose, the present application also provides an LED splicing screen cloud correction device, which comprises: a data acquisition and processing module, configured to acquire splicing screen data, and perform single-screen basic correction of a screen module based on the splicing screen data; a boundary monitoring module, configured to monitor whether there is boundary mismatch through an edge computing node after the single-screen basic correction is completed; the boundary mismatch includes deviation of luminance and / or chrominance of adjacent screen module edge pixels; a data communication module, configured to upload boundary data to the cloud through the edge computing node if there is boundary mismatch; a correction execution module, configured to obtain a boundary correction parameter based on a pre-trained boundary correction model in the cloud, and receive the boundary correction parameter to perform cross-screen compensation.

[0015] In addition, to achieve the above-mentioned purpose, the present application also provides an LED splicing screen cloud correction device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the LED splicing screen cloud correction method as described above.

[0016] In addition, to achieve the above object, the present application also provides a storage medium, which is a computer readable storage medium, and a computer program is stored on the storage medium, and the computer program realizes the steps of the LED splicing screen cloud correction method when executed by a processor.

[0017] In addition, to achieve the above object, the present application also provides a computer program product, which comprises a computer program, and the computer program realizes the steps of the LED splicing screen cloud correction method when executed by a processor.

[0018] The one or more technical solutions provided by the present application have at least the following technical effects: The present application acquires splicing screen data, performs single-screen basic correction on a screen module based on the splicing screen data, monitors whether there is a boundary mismatch through an edge computing node after the single-screen basic correction is completed, the boundary mismatch includes that the brightness and / or chroma of the edge pixels of adjacent screen modules deviate, uploads boundary data to the cloud through the edge computing node if there is a boundary mismatch, obtains boundary correction parameters based on a pre-trained boundary correction model in the cloud, and receives the boundary correction parameters to perform cross-screen compensation. The present application eliminates the visible splicing joint of the splicing screen through cloud intelligent compensation, improves the display quality, dynamically compensates the influence of environmental factors, improves the adaptability of the splicing screen, improves the resource utilization efficiency by combining edge computing and cloud models, and reduces operation and maintenance costs. BRIEF DESCRIPTION OF DRAWINGS

[0019] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present application and, together with the specification, serve to explain the principles of the application.

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0021] Figure 1 A flowchart is provided for the LED splicing screen cloud correction method embodiment one of the present application; Figure 2 A flowchart is provided for the LED splicing screen cloud correction method embodiment two of the present application; Figure 3 A flowchart is provided for the LED splicing screen cloud correction method embodiment three of the present application; Figure 4A flowchart provided by the fourth embodiment of the LED splicing screen cloud correction method of the present application is shown in the figure. Figure 5 A flowchart provided by the fifth embodiment of the LED splicing screen cloud correction method of the present application is shown in the figure. Figure 6 A flowchart provided by the sixth embodiment of the LED splicing screen cloud correction method of the present application is shown in the figure. Figure 7 A module structure diagram of the LED splicing screen cloud correction device of the embodiment of the present application is shown in the figure. Figure 8 A device structure diagram of the hardware operating environment involved in the LED splicing screen cloud correction method of the embodiment of the present application is shown in the figure.

[0022] The object realization, functional features and advantages of the present application will be further explained with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0023] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and do not limit the present application.

[0024] In order to better understand the technical solutions of the present application, the specific embodiments will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] The prior art cannot meet the requirements of high-precision and long-time display screen splicing seam elimination.

[0026] The present application provides a solution, by acquiring splicing screen data, based on the splicing screen data, the single screen basic correction of the screen module is carried out; after the single screen basic correction is completed, whether there is a boundary mismatch is monitored by the edge computing node; the boundary mismatch includes the deviation of the brightness and / or chroma of the adjacent screen module edge pixels; if there is a boundary mismatch, the boundary data is uploaded to the cloud by the edge computing node; based on the pre-trained boundary correction model in the cloud, the boundary correction parameter is obtained, and the cross-screen compensation is carried out by receiving the boundary correction parameter. The present application eliminates the visible splicing seam of the splicing screen through cloud intelligent compensation, improves the display quality, dynamically compensates the influence of environmental factors, improves the adaptability of the splicing screen, and improves the resource utilization efficiency by combining edge computing with cloud model, reduces the operation and maintenance cost.

[0027] Based on this, the present application provides an LED splicing screen cloud correction method, which is described with reference to Figure 1 , Figure 1 A flowchart of the first embodiment of the LED splicing screen cloud correction method of the present application is shown in the figure.

[0028] In the present embodiment, the LED splicing screen cloud correction method comprises steps S10-S40: Step S10, acquiring splicing screen data, and performing single-screen basic correction on the screen module based on the splicing screen data; It should be noted that the splicing screen data refers to a collection of optical parameters (brightness, chrominance), device state parameters (cumulative working time, driving current) and environmental parameters (temperature, humidity) collected from each screen module of the LED splicing screen; the single-screen basic correction refers to a process of compensating for the pixel-level brightness and chrominance deviation within a single screen module to reach a preset standard value; the screen module refers to the smallest physical unit constituting the splicing screen, including an LED lamp bead array, a driving circuit and a control interface. Through real-time collection of original data by a local optical sensor (such as a spectrophotometer) and a state monitoring chip, combined with the gain coefficient issued by the cloud, display deviations caused by LED aging, driving circuit drift or environmental fluctuations within the single-screen module are eliminated, and the display uniformity of each module itself is ensured, laying a foundation for cross-screen collaborative correction.

[0029] In one possible implementation, the optical sensor collects the brightness / chrominance values of the center and four corners of the module at a sampling rate of 120 Hz, the infrared sensor synchronously records the surface temperature of the module, the ADC monitors the power voltage, the cumulative working time is read from the EEPROM of the receiving card, and the data is uploaded to the cloud through an encrypted link; the cloud pixel aging model can generate the gain coefficient based on a space-time convolution network (STCNN).

[0030] Step S20, after the single-screen basic correction is completed, whether there is a boundary mismatch is monitored by an edge computing node; the boundary mismatch includes a deviation of brightness and / or chrominance of the edge pixels of adjacent screen modules; It should be noted that the edge computing node refers to an embedded processor deployed locally on the splicing screen, which has the ability to analyze pixel data in real time. The boundary mismatch refers to a visual fault at the joint of adjacent screen modules caused by a jump in brightness and chrominance, and the quantitative indicators include the brightness standard deviation or the color deviation. The target area refers to a rectangular area extending a certain number of rows / columns (such as 2 columns) to the left and right of the splicing joint. The edge node extracts the corrected pixel data after the single-screen basic correction, detects the boundary anomaly through brightness standard deviation calculation and color deviation calculation, and dynamically sets the brightness standard deviation threshold and the color deviation threshold.

[0031] In a specific implementation, for the third row of the horizontal joint, the brightness values of 3 columns of pixels of the left and right modules are extracted, for example, a total of 6*1080 pixels; the brightness standard deviation σ=7.8% is calculated, which is greater than the preset brightness standard deviation threshold 5%, and the color deviation ΔE=4.2 is greater than the preset color deviation threshold 3; the joint area is marked as an abnormal area, and a JSON data packet containing the position coordinates and the deviation value is generated.

[0032] Step S30, if there is a boundary mismatch, uploading boundary data to the cloud through the edge computing node; It should be noted that the boundary data refers to the compressed data set of the abnormal area, including pixel coordinates, corrected brightness / chromaticity values, module ID, environmental parameters and diagnostic indicators; the encrypted data packet can be transmitted to the cloud message queue through the 5G / MQTT protocol.

[0033] In a possible implementation, the differential encoding is used to transmit the change amount and the Zstandard compression pixel data, and an encryption algorithm is used to transmit the data to the cloud. When the network is interrupted, the local cache is started, and the breakpoint resume transmission is performed after the link is restored.

[0034] In step S40, the boundary correction parameters are obtained based on the pre-trained boundary correction model in the cloud, and the boundary correction parameters are received for cross-screen compensation.

[0035] It should be noted that in the present application, the boundary correction model refers to a set of correction models deployed in the cloud, including a brightness compensation network and a chromaticity compensation network. Cross-screen compensation refers to synchronously issuing compensation coefficients to adjacent modules, adjusting PWM (pulse width modulation) driving parameters based on the boundary correction parameters combined by the brightness compensation gradient field and the final gain coefficient to realize the gradual fusion of the seam area brightness and chromaticity. The pre-trained boundary correction model in the cloud generates compensation parameters optimized by physical constraints, so that the brightness standard deviation and color deviation of the splicing seam are reduced to within the preset threshold.

[0036] Further, with reference to Figure 2 , the second embodiment of the data merging and scheduling method provides a flowchart. Based on the above Figure 2 , the step of "obtaining splicing screen data and performing single-screen basic correction of the screen module based on the splicing screen data" in step S10 is further refined, including steps A201-A204: Step A201, obtaining the current brightness value, the current chromaticity value, and the first working time of each screen module; Step A202, uploading the current brightness value, the current chromaticity value, and the first working time to the cloud; Step A203, obtaining the brightness gain coefficient and / or the chromaticity gain coefficient corresponding to the screen module through the pre-trained pixel aging model in the cloud; Step A204, receiving the brightness gain coefficient and / or the chromaticity gain coefficient to perform single-screen basic correction on the corresponding screen module, and storing the corrected data after single-screen basic correction.

[0037] It should be noted that in the present embodiment, the current luminance value refers to the screen module pixel-level luminous intensity data collected in real time by the optical sensor, which is used to quantify the actual output luminous flux of the LED. The current chrominance value can be represented as (x, y) coordinates in the CIE 1931 chrominance coordinate system and can be obtained by sampling the center and four corners of the module using a chroma meter to reflect the accuracy of color reproduction. The first working time refers to the cumulative running time (unit: hour) of the module from activation to the present, which is stored in the receiving card encrypted EEPROM and is used to quantify the aging degree of the LED chip. The pixel aging model refers to a machine learning model (such as a spatio-temporal convolution network) deployed on the cloud, which establishes a prediction function for luminance decay and chrominance drift by analyzing massive historical aging data. The luminance gain coefficient and the chrominance gain coefficient are scalar parameters in the range of 0.8-1.2, which are used to compensate for the performance degradation of the LED due to aging. Their application must satisfy the formula: output value = input value x gain coefficient x temperature compensation factor (T).

[0038] In the present embodiment, high-complexity aging modeling is integrated into a cloud GPU cluster, which significantly improves processing speed compared to the local scheme. Aggregating cross-module historical data to train the pixel aging model can improve compensation accuracy, and the gain coefficient issued by the cloud processing can dynamically adapt to real-time changes and online updates.

[0039] In one possible implementation, temperature data is obtained by a temperature sensor, and at the same time, the optical sensor collects the luminance value of the positioning point. The first working time is read from the encrypted storage area of the receiving card / storage chip. The edge gateway uploads Zstandard compressed data packets through the 5G network, and the cloud inputs the pixel aging model after analysis. Based on the historical pixel data stored in the cloud, the gain is predicted: the red channel gain is 1.08 to compensate for high-temperature red light attenuation, the green channel gain is 0.95 to suppress color deviation after long-term use, and the blue channel gain is 1.12 to resist rapid blue light decay. During compensation, the local receiving card triggers the buffer switching instruction, and the display refreshes seamlessly to the new parameters, making the compensated screen module luminance uniform. In addition, the correction data storage can adopt a partition block strategy: the basic gain coefficient is stored in the receiving card Flash memory, and the dynamic compensation parameter is temporarily stored in the FPGA cache area, supporting power failure protection and fast rollback.

[0040] Further, with reference to Figure 3 , the third embodiment of the data merging and scheduling method of the present application provides a flowchart. Based on the above Figure 3 , the step of "monitoring whether there is a boundary mismatch after the single-screen basic correction is completed" in step S20 is further refined, including steps A301-A305: Step A301, extracting the luminance correction value of each pixel in the target area of the spliced screen screen module; It should be noted that the target area refers to a rectangular area centered on the splicing seam and extending 2-3 columns of pixels to the left and right adjacent modules, which covers the optical jump area sensitive to the human eye; the brightness correction value is the pixel brightness data (unit: nit) after single-screen basic correction, which is collected by the edge computing node in real time through the optical sensor and stored in the local cache; the edge computing node reads data from the receiving card double-buffer memory during the vertical blanking period, ensuring that it does not affect the real-time display refresh rate, and the vertical blanking period refers to the interval time during which the electron beam returns from the bottom to the top of the screen after completing the scanning of a frame of picture; by extracting full-pixel data in the seam area instead of sampling data, missing detection caused by sampling is avoided, providing complete input for subsequent statistical calculation, and solving the problem of insufficient compensation accuracy caused by data islands.

[0041] Step A302, calculating the average correction brightness of the target area based on the brightness correction value; It should be noted that the average correction brightness is the arithmetic mean of the brightness correction values of all pixels in the target area, which is used to quantify the overall brightness level of the target area. The edge computing node can use a parallel accumulator hardware unit such as FPGA to achieve fast calculation. This average correction brightness value participates in the calculation of the brightness standard deviation and serves as the overall brightness reference for the region, eliminating single-point noise interference and significantly improving detection reliability.

[0042] Step A303, calculating the brightness standard deviation of the target area based on the average correction brightness; It should be noted that the brightness standard deviation is used to quantify the dispersion of pixel brightness and directly reflects the visual uniformity of the splicing seam. The formula for the brightness standard deviation is: where N is the number of samples, represents the brightness correction value of the i-th pixel, μ represents the sample mean, and μ is the average correction brightness of the brightness correction values of all pixels in the target area.

[0043] Step A304, comparing the brightness standard deviation with a preset brightness standard deviation threshold value; It should be noted that the preset brightness standard deviation threshold value is a critical value set according to the visual sensitivity of the human eye, and when the detection value exceeds this threshold value, it is determined that there is a boundary mismatch. The edge computing node supports dynamic threshold adjustment: for example, the threshold value is set to 8% in high-brightness mode (greater than 5000 nit) and 5% in low-brightness mode (less than 1000 nit); the preset brightness standard deviation threshold value can also be adjusted according to the display scene. The comparison between the brightness standard deviation and the preset brightness standard deviation threshold value can be performed in real time by a hardware comparator.

[0044] Step A305, if the brightness standard deviation is greater than the preset brightness standard deviation threshold value, it is determined that there is a boundary mismatch, and the target area is marked as an abnormal area; otherwise, the brightness correction value is maintained.

[0045] It should be noted that when the luminance standard deviation is out of standard, the edge computing node triggers an exception mark and starts a data compression process; otherwise, the current correction parameter is maintained; at the same time, it is ensured that only the abnormal data is uploaded to the cloud to improve the resource utilization. The marked region coordinates will be used as the input for subsequent cloud compensation, forming a closed loop of detection-decision-execution.

[0046] In this embodiment, the luminance uniformity of the splicing seam area is quantified to objectively judge whether there is a boundary mismatch (i.e., the luminance deviation of the adjacent module edge pixels).

[0047] In a specific embodiment, the edge computing node extracts the luminance correction values of the 3 columns of pixels on the left and right of the 3rd row of the horizontal seam from the double-buffered memory; the average luminance μ = 4850 nit is calculated, and then the luminance standard deviation σ = 7.8% is calculated; the luminance standard deviation σ is compared with the preset luminance standard deviation threshold 5%, and the out-of-standard determination is triggered; the region "HS3" is marked as abnormal, and the compression mask data is generated; the edge computing node uploads the boundary data (including the luminance correction value, the average luminance, the standard deviation, etc.) of the abnormal region to the cloud, and the boundary correction model of the cloud generates the boundary correction parameter to perform cross-screen compensation to eliminate the luminance deviation of the splicing seam.

[0048] Further, with reference to Figure 4 , the fourth embodiment of the data merging and scheduling method provides a flowchart, which is based on the above Figure 4 embodiment, the step of "monitoring whether there is a boundary mismatch through the edge computing node after the single-screen basic correction is completed" in step S20 is further refined, and further includes steps A401-A403: Step A401, extracting the chrominance correction value of each pixel in the target region of the splicing screen module; It should be noted that the chrominance correction value is the actual chrominance data of each pixel point of the screen module after the single-screen basic correction, such as the (x, y) coordinates in the CIE 1931 chrominance coordinate system, which is collected and stored by the edge computing node through a chrominance meter.

[0049] Step A402, calculating the color deviation based on the chrominance correction value and the pre-stored reference chrominance value; It should be noted that the reference chrominance value refers to the preset standard chrominance value, which is stored in the memory of the edge computing node as a reference for judging the consistency of the chrominance. The color deviation is the difference between the chrominance correction value of the pixel in the target region and the reference chrominance value, and the commonly used quantitative indicators include the sum of the absolute values of CIE ΔE or Δx, Δy. The purpose of calculating the color deviation is to quantify the chrominance deviation and judge whether there is a boundary mismatch. For example, the CIE 2000 formula is used to calculate ΔE: wherein AL, AC, AH are the differences of lightness, chroma, hue respectively, AL*, AC*, AH* are the corresponding threshold values, and R_T is a rotation term.

[0050] In step A403, if the color deviation is greater than the preset color deviation threshold, the target region is marked as an abnormal region; otherwise, the chroma correction value is maintained.

[0051] It should be noted that the preset color deviation threshold is a critical value set according to the visual characteristics of the human eye and is stored in the memory of the edge computing node. The abnormal region also refers to the target region with a color deviation greater than the threshold value, which will be uploaded to the cloud for cross-screen compensation after being marked. If the color deviation is greater than the preset color deviation threshold, it means that the chroma deviation of the adjacent module edge has exceeded the range acceptable by the human eye; otherwise, the current chroma correction value is maintained, indicating that the chroma consistency is good. For example, when the calculated color deviation AE = 4 is greater than the preset color deviation threshold 3, the edge computing node marks the target region as an abnormal region, stores the coordinates, color deviation value and other information of the region, and uploads them to the cloud. The chroma compensation model of the cloud will generate chroma compensation parameters based on these data, and the edge computing node will perform cross-screen compensation after receiving the parameters, i.e., adjusting the chroma correction value of the pixels at the edge of the adjacent module to make the color deviation less than the preset color deviation threshold, thereby eliminating the chroma discontinuity at the splicing seam and ensuring that the chroma consistency is maintained for a long time. In addition, maintaining the chroma correction value does not mean that it is completely not adjusted. Instead, in the case where the color deviation is not over-standard, the current correction parameters are maintained to reduce the load of the cloud and improve resource utilization efficiency.

[0052] In one possible implementation, the selection of the target region can be adjusted according to the point spacing of the spliced screen; and the preset color deviation threshold can also be adjusted according to the display scene.

[0053] Further, with reference to Figure 5 , the fifth embodiment of the data merging and scheduling method of the present application provides a flowchart, which is based on the above-mentioned Figure 5 embodiment, the boundary correction model includes a luminance compensation model and a chroma compensation model; the step of "obtaining boundary correction parameters based on the pre-trained boundary correction model in the cloud" in step S40 is further refined, including steps A501-A503: Step A501, receiving the boundary data and pre-processing the boundary data; It should be noted that the boundary data refers to the abnormal area related data uploaded by the edge computing node, including the brightness correction value of the pixels in the target area, the pixel position coordinates, and the first working time information of the screen module. These data are used to judge the boundary mismatch and are also the input of the compensation parameter generated by the cloud. The preprocessing is a process of cleaning and standardizing the boundary data, aiming to remove noise in the data and unify the data format, such as normalizing the brightness value to the range of 0-1, to ensure the accuracy of feature extraction. For example, a Gaussian filter is used to remove high-frequency noise in the brightness correction value, and min-max normalization is used to map the brightness value to the 0-1 interval, to avoid inaccurate feature extraction due to data noise.

[0054] Step A502, extracting adjacent pixel brightness gradient and screen module aging coefficient based on preprocessed boundary data; It should be noted that the adjacent pixel brightness gradient is the change rate of the brightness value of the adjacent pixels in the target area, reflecting the degree of brightness jump at the splicing joint. The screen module aging coefficient is a decay coefficient calculated based on the first working time of the screen module, reflecting the degree of brightness attenuation of the LED lamp beads due to long-term operation.

[0055] Step A503, inputting the adjacent pixel brightness gradient and the screen module aging coefficient into the brightness compensation model to generate a brightness compensation gradient field corresponding to the abnormal area; the brightness compensation gradient field is the brightness correction parameter of the boundary correction parameter.

[0056] It should be noted that the brightness compensation model is a machine learning model (such as convolutional neural network CNN or visual Transformer) pre-trained by the cloud, and its input is the adjacent pixel brightness gradient and the screen module aging coefficient, and its output is the brightness compensation gradient field of the abnormal area; the model learns a large number of historical splicing joint correction data to establish a collaborative compensation of brightness jump and aging attenuation, solving the problem that the local algorithm cannot run complex models due to limited computing power. The brightness compensation gradient field is the distribution of the brightness compensation parameter of each pixel in the abnormal area, and its essence is to adjust the brightness driving current of each pixel to reduce the brightness gradient at the splicing joint to a range that cannot be perceived by the human body.

[0057] In this embodiment, the boundary data is cleaned by preprocessing, the adjacent pixels reflecting the brightness jump and the screen module aging coefficient reflecting the aging attenuation are extracted, and both are input into the brightness compensation model pre-trained by the cloud to generate an accurate brightness compensation gradient field, solving the data island problem, and using the cloud computing power to run a complex boundary correction model, avoiding the constraints of local computing power.

[0058] In a possible implementation, the preprocessing includes removing random noise in the brightness correction value by using a Gaussian filter, and then using min-max normalization to map the brightness value to the range of 0-1, ensuring consistent data scale. The adjacent pixel brightness gradient can be calculated by using a Sobel operator to calculate the horizontal and vertical gradients of each pixel to obtain the gradient size; the screen module aging coefficient is calculated by using an exponential decay model based on the first working time. The training steps of the brightness compensation model include using a 3-layer convolutional neural network (CNN), setting the input layer to 2 channels, fusing the adjacent pixel brightness gradient and the screen module aging coefficient to form a fused feature map; the middle layer is set to 2-3 convolutional layers, each convolutional layer contains 32 3*3 convolutional kernels, and the activation function uses ReLU, which is used to extract spatial features, wherein, a batch normalization layer is added after the convolutional layer to speed up the training convergence, and a residual connection is added to avoid the gradient vanishing problem of the deep network; the output layer is set to 1 convolutional layer, which is a 1*1 convolutional kernel, and the activation function uses Sigmoid, and outputs the brightness compensation gradient field, such as 1080*4*1, each pixel corresponds to a compensation coefficient of 0-1, and the brightness compensation coefficient of 0.9-1.1 is obtained by inverse normalization. To ensure that the compensation gradient field predicted by the model is accurate (reduces the brightness standard deviation) and smooth (avoids new jumps), the loss function uses a combination of MSE (Mean Squared Error) + gradient loss: the MSE loss measures the difference between the predicted compensation gradient field and the true label; the gradient loss is added to add visual smoothing constraints to make the compensation coefficient change more smoothly. In addition, after the model is deployed, it can be continuously optimized through online learning: such as collecting compensation effect data fed back by edge computing nodes, regularly fine-tuning the model with new data, and improving the timeliness of the model.

[0059] The training process of the brightness compensation model realizes accurate mapping from boundary data to brightness compensation gradient field by using large-scale labeled data, CNN processing spatial features, and considering accuracy and smoothness guided by the loss function. Through cloud computing power and deep learning models, it can quickly adapt to different specifications of spliced screens, has strong universality, and realizes high-precision, long-time splicing seam brightness compensation of local spliced screens.

[0060] In a specific implementation, it is assumed that the training data contains 100,000 abnormal area data, and the input features of one data are: adjacent pixel brightness gradient 0.2, aging coefficient 0.3679, and the true label is compensation coefficient 1.1. After the model is trained, the predicted compensation coefficient is 1.08, the brightness standard deviation of the abnormal area is reduced from 8% to 2% after application, which is lower than the brightness standard deviation threshold of 5%, and the effect of "visual seamless" is achieved.

[0061] Further, with reference to Figure 6 , the sixth embodiment of the data merging and scheduling method of the present application provides a flowchart, which is based on the aboveFigure 6 In the illustrated embodiment, the boundary correction model includes a luminance compensation model and a chrominance compensation model; the step of "obtaining boundary correction parameters based on the pre-trained boundary correction model in the cloud" in step S40 is further refined, and further includes steps A601-A603: Step A601, receiving the boundary data, performing trichromatic separation and feature extraction based on the boundary data; It should be noted that the boundary data refers to the abnormal region related data uploaded by the edge computing node, the HIA includes the chrominance correction value of the pixels in the target region, as well as the pixel position coordinates, and the first working time of the screen module; trichromatic separation is to decompose the chrominance data into the chrominance values of the red (R), green (G), and blue (B) channels, for example, the chrominance coordinates of the R, G, and B components of each pixel, because the color of the LED display is composed of three color light beads, and because the decay rates of RGB three-color LEDs are different, processing by channel can accurately solve the single-channel chrominance deviation; feature extraction is to extract features related to chrominance deviation from each channel, including: 1) chrominance deviation (the difference between the current chrominance value and the pre-stored reference chrominance value, such as the red channel x=0.65, the reference x=0.62, and the deviation is 0.03); 2) chrominance gradient (the rate of change of adjacent pixel chrominance values, reflecting the chrominance jump degree at the splicing seam, such as the red channel x deviation of 0.02 / pixel for horizontally adjacent pixels); 3) channel aging coefficient (the decay coefficient of each channel calculated based on the first working time, such as the blue channel aging rate is faster, and the aging coefficient is 0.85). The extraction of these features provides accurate input for subsequent channel compensation, solving the chrominance fault problem caused by the inability of basic correction to integrate cross-screen three-color features.

[0062] Step A602, inputting the feature extraction data into the chrominance compensation model to calculate red / blue / green compensation coefficients respectively; It should be noted that the chrominance compensation model is a machine learning model pre-trained in the cloud, such as a convolutional neural network CNN, which inputs the features of each channel (chrominance deviation, gradient, and aging coefficient) and outputs the compensation coefficients of the corresponding channel. Channel-by-channel calculation is a precise compensation for the chrominance deviation of each channel, which makes the overall color return to the reference.

[0063] Step A603, based on the white balance constraint, the color gamut constraint, and the visual smoothing constraint, the red / blue / green compensation coefficients are optimized cooperatively to obtain the final gain adjustment coefficient; the final gain adjustment coefficient is the chrominance correction parameter of the boundary correction parameter.

[0064] It should be noted that the white balance constraint refers to the combination of the compensated RGB three colors needing to meet the pre-stored reference white balance, avoiding the overall color deviation after compensation, such as excessive compensation of the red channel leading to a red picture; the color gamut constraint refers to the chroma value of each channel after compensation cannot exceed the color gamut range of the LED module, preventing color distortion; the visual smoothness constraint refers to the compensation coefficient change rate of adjacent pixels cannot exceed the pre-set threshold, avoiding the appearance of chroma fault after compensation, for example, the blue channel compensation coefficient of adjacent pixels changes from 0.98 to 1.05, causing the boundary to be perceived by the human eye, avoiding the formation of a "patchy feeling" due to excessive compensation of a single point. Collaborative optimization is to find the optimal final gain adjustment coefficient under these constraint conditions, solve the chroma deviation of each channel, and ensure normal color and no fault through nonlinear programming (such as sequential quadratic programming) or genetic algorithm.

[0065] In one possible implementation, the trichromatic separation adopts a way of directly extracting trichromatic values from sensor data; the feature extraction is through calculating the chroma deviation of each channel, the chroma gradient of adjacent pixels (such as the x deviation of horizontally adjacent pixels divided by the pixel spacing), and the channel aging coefficient (based on an exponential decay model of the first working time); the chroma compensation model can adopt a 3-layer CNN, the input is 3 features (deviation, gradient, aging coefficient) of each channel, and the output is 1 chroma compensation coefficient; the collaborative optimization adopts a nonlinear programming algorithm, the objective function is to minimize the sum of squares of the chroma deviation of each channel, and the constraint conditions include: 1) white balance constraint; 2) color gamut constraint; 3) visual smoothness constraint.

[0066] Further, the step of receiving the boundary correction parameter for cross-screen compensation includes: checking whether the luminance standard deviation and / or color deviation of the abnormal area after cross-screen compensation meets the standard; If the luminance standard deviation and / or color deviation does not meet the standard, the abnormal area that does not meet the standard is marked as a secondary abnormal area, and the secondary abnormal area is diagnosed or iterated in the cloud.

[0067] It should be noted that the cross-screen compensation is an operation of adjusting the luminance value and / or chrominance value of the edge pixels of the adjacent screen modules by the edge computing node receiving the boundary correction parameters (such as the luminance compensation gradient field and the final gain adjustment coefficient) generated by the cloud, for eliminating the luminance jump and chrominance fault at the splicing joint and realizing the "visual seamless" display. The verification can be completed by the edge computing node, and whether the cross-screen compensation effectively solves the boundary mismatch problem is verified through the verification. If the luminance standard deviation and / or color deviation still exceeds the preset threshold, it indicates that the compensation is not complete and needs to be further processed; and the display quality of the spliced screen is ensured to meet the user's requirement of "no splicing joint". The secondary abnormal area refers to the abnormal area whose luminance standard deviation and / or color deviation still does not meet the standard after the cross-screen compensation, and is the target object for subsequent cloud diagnosis and strategy iteration. The cloud diagnosis is to find out the reasons for not meeting the standard by analyzing the related data of the secondary abnormal area, such as the boundary data integrity, the model input feature accuracy and the compensation parameter rationality, such as the calculation error of the aging coefficient caused by the lack of "first working time" of the boundary data, or the compensation parameter exceeding the reasonable range caused by the model overfitting; and the strategy iteration is to adjust the parameters or algorithm of the boundary correction model according to the diagnosis result, such as retraining the model and optimizing the collaborative optimization constraint condition, so as to improve the adaptability of the model to the complex scene.

[0068] The application also provides an LED spliced screen cloud correction device, please refer to Figure 7 , the LED spliced screen cloud correction device comprises: A data acquisition and processing module 10 is configured to acquire spliced screen data and perform single-screen basic correction on the screen module based on the spliced screen data. A boundary monitoring module 20 is configured to monitor whether there is boundary mismatch through the edge computing node after the single-screen basic correction is completed; the boundary mismatch includes deviation of the luminance and / or chrominance of the edge pixels of the adjacent screen modules. A data communication module 30 is configured to upload the boundary data to the cloud through the edge computing node if there is boundary mismatch. A correction execution module 40 is configured to obtain boundary correction parameters based on the pre-trained boundary correction model in the cloud, and perform cross-screen compensation by receiving the boundary correction parameters.

[0069] The LED spliced screen cloud correction device provided by the application adopts the LED spliced screen cloud correction method in the above embodiment, and can solve the technical problem that the display screen splicing joint elimination requirement of high precision and long time efficiency cannot be met. Compared with the prior art, the beneficial effects of the LED spliced screen cloud correction device provided by the application are the same as those of the LED spliced screen cloud correction method provided by the above embodiment, and the other technical features in the LED spliced screen cloud correction device are the same as those disclosed in the above embodiment method, which will not be repeated here.

[0070] The application provides an LED splicing screen cloud correction device, which comprises at least one processor and a memory connected with the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the LED splicing screen cloud correction method in the above embodiment one.

[0071] Reference will be made to the following description Figure 8 which shows a structural schematic diagram of the LED splicing screen cloud correction device suitable for being used to implement the embodiments of the application. The LED splicing screen cloud correction device in the embodiments of the application can include but is not limited to mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (Personal Digital Assistant), PADs (Portable Application Description), PMPs (Portable Media Player), vehicle-mounted terminals (for example, vehicle-mounted navigation terminals) and the like, and fixed terminals such as digital TVs, desktop computers and the like. Figure 8 The shown LED splicing screen cloud correction device is only an example, and should not bring any limitation to the functions and use range of the embodiments of the application.

[0072] As Figure 8As shown, the LED splicing screen cloud correction device can include a processing apparatus 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to programs stored in a read-only memory 1002 or loaded from a storage apparatus 1003 into a random access memory 1004. Various programs and data required for the operation of the LED splicing screen cloud correction device are also stored in the random access memory 1004. The processing apparatus 1001, the read-only memory 1002, and the random access memory 1004 are connected to each other through a bus 1005. An input / output interface 1006 is also connected to the bus. Generally, the following systems can be connected to the input / output interface 1006: input apparatuses 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output apparatuses 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; the storage apparatus 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication apparatus 1009. The communication apparatus 1009 can allow the LED splicing screen cloud correction device to communicate with other devices wirelessly or by wire to exchange data. Although the LED splicing screen cloud correction device with various systems is shown in the figure, it should be understood that all the systems shown are not required to be implemented or possessed. More or fewer systems can be alternatively implemented or possessed.

[0073] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program codes for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through the communication apparatus, or installed from the storage apparatus 1003, or installed from the read-only memory 1002. When the computer program is executed by the processing apparatus 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are performed.

[0074] The LED splicing screen cloud correction device provided by the present application adopts the LED splicing screen cloud correction method in the above-mentioned embodiments, and can solve the technical problem that the display screen splicing seam elimination requirement of high precision and long time efficiency cannot be met. Compared with the prior art, the LED splicing screen cloud correction device provided by the present application has the same beneficial effects as the LED splicing screen cloud correction method provided by the above-mentioned embodiments, and other technical features in the LED splicing screen cloud correction device are the same as the features disclosed in the previous embodiment method, which will not be described here.

[0075] It should be understood that various aspects of the disclosure can be implemented in hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any appropriate manner in any one or more embodiments or examples.

[0076] The above description is merely illustrative of the application and is not intended to limit the scope of the application. Any variations and modifications that can be made by any person skilled in the art within the spirit and scope of the application are intended to be encompassed by the application. The scope of the application is defined by the appended claims.

[0077] The application provides a computer readable storage medium having computer readable program instructions (i.e. computer programs) stored thereon, the computer readable program instructions being used to perform the LED splicing screen cloud correction method in the above embodiments.

[0078] The computer readable storage medium provided by the application may, for example, be a U disk, but is not limited to an electric, magnetic, optical, electromagnetic, infrared, or semiconductor system or device, or any combination thereof. More specific examples of the computer readable storage medium can include, but are not limited to, an electric connection having one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any appropriate combination thereof. In the present embodiment, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer readable storage medium can be transmitted by any appropriate medium, including but not limited to an electric wire, an optical cable, an RF (Radio Frequency), etc., or any appropriate combination thereof.

[0079] The above computer readable storage medium can be contained in the LED splicing screen cloud correction device; or can exist separately and not be assembled into the LED splicing screen cloud correction device.

[0080] The computer readable storage medium described above carries one or more programs, when the one or more programs are executed by the LED splicing screen cloud correction device, the LED splicing screen cloud correction device is caused to: perform single-screen basic correction of a screen module based on splicing screen data obtained by the LED splicing screen cloud correction device; after the single-screen basic correction is completed, whether there is a boundary mismatch is monitored by an edge computing node; the boundary mismatch includes that the brightness and / or chroma of the edge pixels of adjacent screen modules deviate; if there is a boundary mismatch, boundary data is uploaded to the cloud by the edge computing node; based on a pre-trained boundary correction model in the cloud, boundary correction parameters are obtained, and the boundary correction parameters are received to perform cross-screen compensation. Through cloud intelligent compensation, the splicing seam visible to the naked eye is eliminated, the display quality is improved, the influence of environmental factors is dynamically compensated, the adaptability of the splicing screen is improved, the resource utilization efficiency is improved by the combination of edge computing and cloud model, and the operation and maintenance cost is reduced.

[0081] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0082] The flow and block diagrams in the drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flow and block diagrams can represent a module, a segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustrations, and combinations thereof, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or combinations of hardware and software.

[0083] The modules involved in the embodiments of the present application can be implemented in the form of software or in the form of hardware. In some cases, the name of the module does not constitute a limitation on the unit itself.

[0084] The readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e. computer program) for executing the LED splicing screen cloud correction method described above, and can solve the technical problem that the high-precision and long-time display screen splicing seam elimination requirement cannot be met. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the LED splicing screen cloud correction method provided by the above-mentioned embodiments, and will not be repeated here.

[0085] The present application also provides a computer program product, comprising a computer program, which is executed by a processor to implement the steps of the LED splicing screen cloud correction method as described above.

[0086] The computer program product provided by the present application can solve the technical problem that the high-precision and long-time display screen splicing seam elimination requirement cannot be met. Compared with the prior art, the computer program product provided by the present application has the same beneficial effects as the LED splicing screen cloud correction method provided by the above-mentioned embodiments, and will not be repeated here.

[0087] The above only describes some embodiments of the present application, and does not limit the patent scope of the present application, and any equivalent structural transformation made by using the content of the present application specification and drawings, or direct / indirect application in other related technical fields is included in the patent protection scope of the present application.

Claims

1. A cloud-based calibration method for LED splicing screens, characterized in that: The cloud-based correction method for the LED splicing screen includes: Acquire splicing screen data, and perform single-screen basic calibration of the screen module based on the splicing screen data; After the single-screen basic calibration is completed, edge computing nodes are used to monitor whether there is a boundary mismatch; the boundary mismatch includes deviations in brightness and / or chromaticity of edge pixels of adjacent screen modules; If there is a boundary mismatch, the boundary data is uploaded to the cloud through the edge computing node; Based on a pre-trained boundary correction model in the cloud, boundary correction parameters are obtained, and the boundary correction parameters are received to perform cross-screen compensation.

2. The cloud-based calibration method for LED splicing screens according to claim 1, wherein: The step of obtaining the splicing screen data and performing single-screen basic calibration of the screen module based on the splicing screen data includes: Obtain the current brightness value, current chromaticity value, and first working time of each screen module; Uploading the current brightness value, current chromaticity value and first working time to the cloud; Obtain the brightness gain coefficient and / or chromaticity gain coefficient of the corresponding screen module through the pre-trained pixel aging model in the cloud; The brightness gain coefficient and / or the chroma gain coefficient are received to perform single-screen basic correction on the corresponding screen module, and correction data after the single-screen basic correction is stored.

3. The cloud-based calibration method for LED splicing screens according to claim 2, wherein: After the single-screen basic calibration is completed, the step of monitoring whether there is a boundary mismatch through the edge computing node includes: Extract the brightness correction value of each pixel in the target area of ​​the spliced ​​screen module; Calculating an average corrected brightness of the target area based on the brightness correction value; Obtaining a brightness standard deviation of the target area by calculating the average corrected brightness; Comparing the brightness standard deviation with a preset brightness standard deviation threshold; If the brightness standard deviation is greater than the preset brightness standard deviation threshold, it is determined that there is a boundary mismatch and the target area is marked as an abnormal area; otherwise, the brightness correction value is maintained.

4. The cloud-based calibration method for LED splicing screens according to claim 2, wherein: After the single-screen basic calibration is completed, the step of monitoring whether there is a boundary mismatch through the edge computing node further includes: Extract the chromaticity correction value of each pixel in the target area of ​​the spliced ​​screen module; Calculating a color deviation based on the chromaticity correction value and a pre-stored reference chromaticity value; If the color deviation is greater than a preset color deviation threshold, the target area is marked as an abnormal area; otherwise, the chromaticity correction value is maintained.

5. The cloud-based calibration method for LED splicing screens according to claim 4, wherein: The boundary correction model includes a brightness compensation model and a chromaticity compensation model; the step of obtaining boundary correction parameters based on the boundary correction model pre-trained in the cloud includes: receiving the boundary data and preprocessing the boundary data; Extract adjacent pixel brightness gradient and screen module aging coefficient based on preprocessed boundary data; The adjacent pixel brightness gradient and the screen module aging coefficient are input into the brightness compensation model to generate a brightness compensation gradient field corresponding to the abnormal area; the brightness compensation gradient field is a brightness correction parameter of the boundary correction parameter.

6. The cloud-based calibration method for LED splicing screens according to claim 4, wherein: The boundary correction model includes a brightness compensation model and a chromaticity compensation model; the step of obtaining boundary correction parameters based on the boundary correction model pre-trained in the cloud further includes: receiving the boundary data, and performing three-color separation and feature extraction based on the boundary data; The data obtained by feature extraction is input into the chromaticity compensation model, and red / blue / green compensation coefficients are obtained by channel calculation respectively; The red / blue / green compensation coefficients are collaboratively optimized based on white balance constraints, color gamut constraints, and visual smoothness constraints to obtain a final gain adjustment coefficient; the final gain adjustment coefficient is a chromaticity correction parameter of the boundary correction parameter.

7. The cloud-based calibration method for LED splicing screens according to claim 6, wherein: After the step of receiving the boundary correction parameter and performing cross-screen compensation, the following steps are included: Verify whether the brightness standard deviation and / or color deviation of the abnormal area after cross-screen compensation meets the standards; If the brightness standard deviation and / or color deviation do not meet the standards, the abnormal area that does not meet the standards is marked as a secondary abnormal area, and cloud diagnosis or strategy iteration is performed on the secondary abnormal area.

8. A cloud-based calibration device for LED splicing screens, characterized in that: The LED splicing screen cloud correction device includes: A data acquisition and processing module is used to obtain splicing screen data and perform single-screen basic calibration of the screen module based on the splicing screen data; A boundary monitoring module is used to monitor whether there is a boundary mismatch through an edge computing node after the single-screen basic calibration is completed; the boundary mismatch includes a deviation in brightness and / or chromaticity of edge pixels of adjacent screen modules; A data communication module, configured to upload boundary data to the cloud via the edge computing node if there is a boundary mismatch; The correction execution module is used to obtain boundary correction parameters based on a boundary correction model pre-trained in the cloud, and receive the boundary correction parameters to perform cross-screen compensation.

9. A cloud-based calibration device for LED splicing screens, characterized in that: The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the cloud-based correction method for an LED splicing screen according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the cloud-based correction method for the LED splicing screen according to any one of claims 1 to 7 are implemented.

Images