Display screen data channel anomaly detection and repair method, system and display terminal

Abstract

The application provides a display screen data channel abnormality detection and repair method and system and a display terminal, and belongs to the technical field of display screen abnormality detection and repair. The method comprises the following steps: performing abnormality detection on a display screen, judging whether an abnormal data channel exists based on an improved pixel value comparison method; triggering a compensation mechanism when a compensation trigger condition is met; constructing a pre-trained pixel color compensation model, and compensating and repairing a faulty data channel based on the trained pixel color compensation model using a normal data channel when the compensation mechanism is triggered. The display screen data channel abnormality detection method is used to judge whether each primary color data channel is abnormal, and when it is confirmed that a primary color data channel is abnormal, a color compensation method based on the remaining normal data channels is used for color repair under the condition that the compensation trigger condition is met, so that accurate detection and timely compensation and repair of the display screen data channel are realized, and the user's viewing experience is improved.

Description

Technical Field

[0001] This invention relates to the technical field of display screen image processing, and more specifically, to a method, system, and display terminal for detecting and repairing abnormalities in display screen data channels. Background Technology

[0002] In display terminal devices, the MIPI interface is widely used to connect application processors and displays due to its high bandwidth, low power consumption, and anti-interference capabilities.

[0003] In practical applications, due to factors such as loose connectors, damage to the FPC (flexible printed circuit board), electrostatic discharge (ESD), or long-term aging, intermittent or permanent connection abnormalities may occur in a certain color data channel (e.g., the blue channel). These abnormalities may include increased contact resistance, short circuit to ground, short circuit to power supply, or broken signal lines. Such hardware failures can cause the pixel data corresponding to a specific color (e.g., blue) to fail to be transmitted correctly, resulting in flickering, distortion, or complete loss of that color area on the display, severely impacting user experience and product reliability.

[0004] Existing display screen color detection methods have the following problems: The main detection objects in the existing technology are usually the defects and dead pixels of the liquid crystal itself, while the data channel also includes the liquid crystal screen cable and the product circuit board terminals. For abnormalities in the data channel connection between the liquid crystal and the product circuit board, the existing technology often neglects to detect the fault. To address compensation methods for abnormal display screen data channels, existing technologies disclose an LCD display color calibration system and method. The method includes dividing the LCD display screen into several independent calibration zones and obtaining the display parameters of each calibration zone; constructing a calibration zone color difference compensation matrix based on the differences between the display parameters of each calibration zone and preset standard values, and fitting the linear compensation coefficients of the RGB channels using the least squares method; smoothing the compensation matrices of adjacent calibration zones using a bilinear interpolation algorithm to generate a global compensation mapping table; dynamically acquiring the temperature of each calibration zone through a temperature sensor network, correcting the compensation matrix based on the thermal expansion coefficient model of the liquid crystal material, and updating the global compensation mapping table; and inputting the updated global compensation mapping table into the storage unit of the display screen driver chip to complete the calibration.

[0005] Existing display color compensation methods have the following problems: In existing technologies, when there is an anomaly in the data channel, the color of the image pixel source is no longer correct, and the color after compensation by the global compensation mapping table is also incorrect. This solution uses real-time display data as a benchmark, which will not cause errors in the benchmark due to changes in the usage cycle, and can compensate and update the image data source in real time, thus solving the display anomalies caused by data channel anomalies. Currently, solutions to such problems mainly rely on hardware-level repair or replacement, lacking effective online fault detection and compensation mechanisms at the software level. Therefore, this paper proposes a display terminal for detecting and repairing abnormalities in the display screen's data channel, thereby improving the overall reliability of the product and the user experience. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method, system, and display terminal for detecting and repairing abnormalities in the data channel of a display screen. This solves the technical problem in existing technologies where hardware repair or replacement is required to resolve hardware malfunctions in a single color data channel of a MIPI display screen.

[0007] The present invention adopts the following technical solution.

[0008] According to a first aspect of the present invention, in one embodiment, a method for detecting and repairing abnormalities in a display screen data channel is provided, comprising the following steps: Anomaly detection is performed on the display screen, and an improved pixel value comparison method is used to determine whether there are abnormal data channels. Construct a compensation trigger mechanism to trigger the compensation mechanism when the compensation trigger conditions are met; A pre-trained pixel color compensation model is constructed. When the compensation mechanism is triggered, the faulty data channel is compensated and repaired based on the trained pixel color compensation model using the normal data channel.

[0009] Preferably, the anomaly detection of the display screen includes active detection and passive detection; Active detection includes: sending a test pattern to the display screen when the display screen is powered on, woken up, or periodically idle, and reading the actual display data packet of the status register inside the display screen through the reverse channel of the MIPI D-PHY data line to detect anomalies; Passive detection includes: sending RGB frame data while the display is running, and reading the actual display data packets from the status register inside the display via the reverse channel of the MIPI D-PHY data line to perform anomaly detection.

[0010] Preferably, the method of determining whether an abnormal data channel exists based on improved pixel value comparison specifically includes: The image color is verified to be abnormal by means of the image pixel packing rules. The actual display data packet is read from the reverse channel of the MIPI D-PHY data line into the status register of the display screen and CRC and ECC checks are performed. If both CRC and ECC checks pass, the data packet that has passed the check is obtained and enters the high-level verification stage. The high-level verification stage determines whether there is an abnormal data channel.

[0011] Preferably, the high-level verification step specifically includes: The data packets that pass the test are decompressed to obtain the values ​​of specific color components, and the deviation ratio between the values ​​of specific color components and the expected values ​​is calculated. If the deviation ratio between the value of a specific color component after unpacking and the expected value exceeds a preset threshold, it indicates that the data channel corresponding to that color component is abnormal. If the deviation ratio between the value of a specific color component after unpacking and the expected value does not exceed a preset threshold, it indicates that the data channel corresponding to that color component is normal.

[0012] Preferably, the deviation ratio Δ between the specific color component value and the expected value is calculated as follows: Δ=(∣Vcurrent Vref∣ / max(Vmax Vmin, ))×100% Where Vcurrent is the current actual pixel value of the LCD display, Vref is the actual pixel value to be displayed, and Vmax and Vmin are the maximum and minimum values ​​of the statistical pixels within the local window, respectively. It is a local minimum.

[0013] Preferably, the construction of the compensation triggering mechanism, which triggers the compensation mechanism when the compensation triggering condition is met, specifically includes: The display screen is divided into areas, and counters are set for different color data channels in each area for fault monitoring. When an abnormality is detected in a certain color data channel in a certain area, the counter corresponding to that color data channel in that area is incremented by 1. If the counter corresponding to a certain color data channel in a certain area counts more than a preset counting threshold within a preset time window, a compensation mechanism will be triggered for the abnormal data channel of that color in that area; otherwise, the compensation mechanism will not be triggered.

[0014] Preferably, constructing the pre-trained pixel color compensation model specifically includes: Construct a pixel color compensation model; Build the training dataset; Construct the loss function and optimization objective of the pixel color compensation model; The optimal coefficients of the pixel color compensation model are solved by combining the loss function and the optimization objective. The training dataset is used to train, validate, and adjust the training dataset to obtain the trained pixel color compensation model. The faulty data channel is compensated by the trained pixel color compensation model, thus completing the repair.

[0015] Preferably, for the i-th pixel of the display screen, the compensation value obtained by the pixel color compensation model is as follows: =a× +b× +c in, The RGB compensation value of the i-th pixel, obtained through the pixel color compensation model, represents the color to be compensated. , ...

[0016] Preferably, the loss function L(a,b,c) for constructing the pixel color compensation model is as follows: L(a,b,c)=Σ( - ) 2 in, Let be the RGB value of the color to be compensated for at the i-th pixel in the training data; The optimization objective is to minimize the loss function value L(a,b,c).

[0017] According to a second aspect of the present invention, in one embodiment, a display screen data channel anomaly detection and repair system is provided for implementing the display screen data channel anomaly detection and repair method as described in the first aspect of the present invention, comprising: The anomaly detection module is used to detect anomalies in the display screen by determining whether an abnormal data channel has occurred based on an improved pixel value comparison method. The compensation trigger module is used to construct a compensation trigger mechanism, which is triggered when the compensation trigger conditions are met. The model building module is used to build a pre-trained pixel color compensation model; The repair module is used to compensate and repair abnormal data channels that trigger the compensation mechanism based on the trained pixel color compensation model.

[0018] According to a third aspect of the present invention, in one embodiment, the present invention also provides a display terminal, comprising: a display screen data channel anomaly detection and repair system as described in the second aspect of the present invention.

[0019] The beneficial effects of this invention are that, compared with the prior art, this invention provides a display terminal for detecting and repairing abnormal data channels of a display screen. By constructing a display screen data channel abnormality detection method, it determines whether each primary color data channel is abnormal. After confirming that a certain primary color data channel is abnormal, it activates a color reconstruction algorithm based on the information of the remaining normal data channels when the compensation trigger condition is met, generates alternative data to compensate for the abnormal channel, and finally outputs the corrected display data. This method can achieve accurate detection and timely compensation and repair of display screen data channels, improving the user's viewing experience.

[0020] The beneficial effects of the present invention include at least the following: 1. Active and passive detection methods are used for the display screen data channel. Detection is performed in both the running and non-running states of the display screen. This can detect not only abnormalities in the display screen's LCD itself, but also abnormalities in the data channel connection between the LCD and the product circuit board. 2. By verifying the logical consistency of byte order and color component values ​​through pixel data packet decoding, the technical problem of traditional methods that only verify whether the data packet reception is correct without further verifying the correctness of pixel content is solved, filling the gap between the two detection stages and enabling precise location of hardware faults in specific color data channels. 3. This invention has high reliability. When the display hardware fails partially, the color compensation method proposed in this invention can maintain the basic display function of the display screen, transforming the hard failure into an acceptable soft degradation, so that the content on the display screen is still clear and recognizable, thus improving the user experience of the product. 4. The detection and repair method proposed in this invention is low in cost. It is implemented entirely through software or firmware, without increasing any hardware costs, and also avoids the after-sales repair costs for minor faults.

[0021] 5. This invention features intelligence and adaptability, possessing the ability to automatically detect, diagnose, and compensate without user intervention.

[0022] 6. In this invention, after passing CRC and ECC verification, a higher-level verification step is added to compare pixel values ​​one by one, avoiding misjudgment and missed detection of hidden faults, thereby providing an accurate basis for subsequent color compensation and filling the detection gap between correct data packet reception and correct pixel content. Attached Figure Description Figure 1 This is a flowchart of the display screen data channel anomaly detection and repair method in this invention; Figure 2 This is a structural diagram of the display screen data channel anomaly detection and repair system of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of this invention.

[0024] like Figure 1 As shown in the figure, this embodiment of the invention proposes a display terminal for detecting and repairing abnormalities in the display data channel. The abnormality detection and repair method specifically includes the following steps: Step 1: Perform anomaly detection on the display screen and determine whether there are abnormal data channels based on the improved pixel value comparison method; In this invention, the anomaly detection methods include an active diagnosis mode and a passive detection mode, as detailed below: Active diagnostic modes include: When the display is powered on, woken up, or periodically idle, a series of known solid color test patterns are sent to the display. The actual display data packets in the status register inside the display are read through the reverse channel of the MIPI D-PHY data line to determine whether the actual display data packets match the expectations. Among them, the solid color test pattern can be a full blue, full red, full green, or full white image. For example, when sending a full blue image, if the brightness of the blue pixels is detected to be much lower than expected or there is no response at all, while the red and green images are normal, then the blue data channel is determined to be abnormal.

[0025] Passive detection modes include: During display operation, RGB frame data is sent. The actual display data packet in the status register of the display is read through the reverse channel of the MIPI D-PHY data line for anomaly detection. The actual display data packet and the actual sent RGB frame data are analyzed and compared in real time to determine whether the display has an anomaly.

[0026] The display data is analyzed and compared with the actual RGB frame data sent in real time to determine whether there is an abnormality in the display screen. Specifically, if the blue component data (B value) undergoes an unexpected and drastic full-frame jump in multiple consecutive frames (for example, collectively changing from the normal value to 0 or 255), and this jump is unrelated to the image content, while the red and green component data change smoothly, it can be inferred that there is an intermittent poor contact in the blue channel (i.e., manifested as blue flicker).

[0027] More preferably, the method of determining whether an abnormal data channel exists based on improved pixel value comparison specifically includes: Verify image color anomalies using image pixel packing rules: The actual display data packet in the status register of the display screen is read from the reverse channel of the MIPI D-PHY data line, and CRC (Cyclic Redundancy Check) and ECC (Error Checking and Correcting) checks are performed. If both CRC and ECC checks pass, the actual display data packet that has passed the checks is obtained and enters the higher-level verification stage.

[0028] Furthermore, the high-level verification process is detailed below: If both the CRC check and the ECC check pass, the actual display data packet that passed the check is unpacked, and the deviation ratio between the unpacked value of a specific color component and the expected value is calculated. If the deviation ratio between the value of a specific color component after unpacking and the expected value exceeds a preset threshold, it indicates that the data channel corresponding to that color component is abnormal. If the deviation ratio between the value of a specific color component after unpacking and the expected value does not exceed a preset threshold, it indicates that the data channel corresponding to that color component is normal.

[0029] Specifically, calculating the deviation ratio Δ between a specific color component value and the expected value includes: Δ=(∣Vcurrent Vref∣ / max(Vmax Vmin, ))×100% Where Vcurrent is the current actual pixel value of the LCD display, Vref is the actual pixel value to be displayed, and Vmax and Vmin are the maximum and minimum values ​​of the statistical color pixels within the local window, respectively. By using the same color pixel values ​​at the same position in the current frame and several historical frames, the maximum and minimum values ​​of the statistical pixels are updated in real time, thereby determining the display range of a certain color at a certain position, which is more conducive to calculating drastic jumps. It is a local minimum.

[0030] Among them, the local window is a local display area, which can be preset by technicians.

[0031] The dynamic range is calculated based on the maximum and minimum values ​​of the statistical pixels, Vmax and Vmin. The dynamic range is Vmax. Vmin, when the dynamic range is greater than When the deviation ratio Δ is used, the denominator is taken as the dynamic range; otherwise, it is taken as... To avoid division by zero; minimum value It is not a fixed value. The value is the quantization step size of a certain color pixel value, which is obtained by using the same color pixel value at the same position in the current frame and several historical frames.

[0032] Furthermore, if the fault is determined to be local, only the faulty part needs to be compensated and repaired; if the fault is determined to be global, then global compensation and repair are required.

[0033] This invention verifies the logical consistency of byte order and color component values ​​by decoding pixel data packets. This solves the technical problem of traditional methods that only verify the correctness of data packet reception without further verifying the correctness of pixel content. It fills the gap in real-time updating statistics of the step size change of a pixel value at a certain position between two segments. For example, when the pixel value is rounded down and the quantization step size is 1... = 1 Detect blanks to accurately locate hardware faults in specific color data channels.

[0034] As a preferred implementation, if the deviation ratio between the value of a specific color component after unpacking and the expected value exceeds a preset threshold, and the byte arrangement order is correct, then it is determined that the color data channel has a hidden fault with correct format but incorrect value; the hidden fault refers to the pixel data being able to pass the verification, but the actual color component has changed, that is, the data looks correct, but if the color value before sending and the actual color value are not further compared, it will be assumed that the actual display is correct. If the values ​​of specific color components consistently deviate from the expected values ​​and the deviation is stable after unpacking, it is determined to be a global data channel fault; if the values ​​of specific color components of different pixels show regular changes after unpacking, it is determined to be a local fault, and the MIPI data channel where the fault occurred can be further located.

[0035] As an example, the fault determination process is illustrated below: Assume the original data packet pixel payload (first 6 pixels, 18 bytes in total) is as shown in Table 1 below: Table 1: Original Data Packet Pixel Load Table

[0036] Verification method: Check if the first byte is the Blue component—in a full blue screen, Blue should be 255 (0xFF), pass.

[0037] Check if the second byte is the Green component—in a full blue screen, Green should be 0 (0x00), pass.

[0038] Check if the 3rd byte is the Red component—in a full blue screen, Red should be 0 (0x00), pass.

[0039] Check if the byte arrangement strictly follows the BGR cyclic pattern - pass.

[0040] Verification result: The byte order fully conforms to the RGB888 specification, and the packaging format verification is passed.

[0041] CRC / ECC check results: Perform CRC check on the data packet: The CRC generator polynomial is x¹ 6 +x¹²+x 5 +x 0 (Compliant with MIPI DSI protocol standard). The calculated CRC checksum matches the CRC checksum at the end of the packet, and the CRC check passes.

[0042] At this point, if the system relies solely on existing CRC / ECC verification techniques, it will report "data transmission is normal." However, in reality, all blue component values ​​are 255 (0xFF), which is as expected, and no anomaly is detected—because this example represents a normal situation. This application further adds a verification process based on the deviation ratio after CRC and ECC verification, unpacking the actual display data packets that pass the verification and calculating the deviation ratio between the unpacked specific color component values ​​and the expected values.

[0043] When the blue data channel is completely broken (e.g., the data line is short-circuited to ground), the blue component is pulled down to 0 during transmission. The original data packet pixel payload becomes as shown in Table 2 below: Table 2: Original Data Packet Pixel Load Table After Change

[0044] Observations: Blue component (byte 1): 0x00 (actual value 0 vs expected value 255) Green component (byte 2): 0x00 (expected value 0, matches) Red component (3rd byte): 0x00 (expected value 0, matches) Byte order verification: Verification results: The byte order is still correct (BGR loop mode remains unchanged), and the packaging format verification is successful.

[0045] Numerical verification: For ease of illustration, in this example, max(Vmax) Vmin, If the value is 255, then calculate the deviation ratio ΔB of the blue component: ΔB = |0 - 255| / 255 × 100% = 100% A change exceeding 50% is considered a drastic change. If more than 80% of the pixels in the entire display area show a consistent direction of blue change (either all increasing or all decreasing), then the change is confirmed to be frame-wide.

[0046] Overall judgment conclusion: The data packet was received completely at the link layer (CRC check passed) and the packet format was correct (byte order conforms to BGR specification), but the blue component value was systematically incorrect (always 0). Therefore, it was determined that there was a hardware fault in the blue data channel (data line breakage / short circuit to ground), and the color compensation mechanism was triggered.

[0047] When the blue component value of all pixels is 0 and the red and green components are normal, the system determines that there is a global data channel failure, that is, the failure occurs in all data channels of the MIPI physical layer (or at least affects the blue component of all pixels), and the compensation strategy is full-screen compensation.

[0048] If the fault only affects a specific area, for example, starting from a certain pixel position, the blue component suddenly becomes 0, while the blue component of previous pixels is normal, as shown in Table 3 below: Table 3: Correspondence between data byte sequence and blue component value for partial faults

[0049] At this point, the system determines that the fault is localized. It may be due to poor contact (intermittent) of a specific pin of the FPC connector, or intermittent failure of a lane in the MIPI data channel. The compensation strategy can be selected to compensate only for the abnormal area.

[0050] Step 2: Construct a compensation trigger mechanism to trigger the compensation mechanism when the compensation trigger conditions are met; To prevent false alarms, a confidence counter is installed inside the display terminal to monitor the data channels of different colors separately; More preferably, the display screen is pre-divided into several areas, and the data channels of different colors in each area are monitored and counted separately; When an anomaly is detected in the data channel of a certain color in a certain area, the counter corresponding to the data channel of that color in that area is incremented by 1; If the counter corresponding to a certain color data channel in a certain area counts more than a preset counting threshold within a preset time window, a compensation mechanism will be triggered for the abnormal data channel of that color in that area; otherwise, the compensation mechanism will not be triggered.

[0051] When an abnormal event is detected, the counter increments, and the counter is counted separately for different abnormalities; If the counter count exceeds the threshold within a preset time window, a data channel failure is confirmed, a compensation mechanism is triggered, and the error is logged.

[0052] If the count does not exceed the threshold within the preset time window, it is determined that there is no fault. At this time, the count is reset to zero and the counting of the next time window restarts. If the compensation mechanism has been triggered, the count is also reset to zero and the counting of the next time window restarts.

[0053] As a preferred implementation, the time window is set to 1 second. The length of the 1-second time window is just enough to capture the flicker frequency that the human eye can perceive, while avoiding short-term noise interference. In a preferred embodiment, the number of active and passive detections performed within each time window is set by technicians according to the actual situation, and the sum of the number of active and passive detections performed within each time window is not less than 50, and the counting threshold is set to 5. If the count is greater than or equal to 5, the compensation mechanism is triggered.

[0054] Step 3: Construct a pre-trained pixel color compensation calculation model, and use the normal data channel to compensate for the faulty data channel based on the pixel color compensation calculation model; When an anomaly is confirmed in a certain channel and triggered, the module automatically switches to the compensation mode. The goal of the compensation is to estimate and predict the lost or incorrect color components based on the normal information of the two color channels.

[0055] For example, once a blue channel anomaly is confirmed, the module automatically switches to compensation mode. The goal of compensation is to estimate and predict the missing or erroneous blue (B) component based on the normal red (R) and green (G) channel information: when the blue channel (B) hardware fails, the compensated B value can be calculated using the still normally transmitted R and G values, combined with a pre-trained pixel color compensation calculation model, and the final display can be performed based on the compensated RGB value to complete the repair.

[0056] Specifically, based on the constraints of spectral continuity and color constancy, in most natural scenes, the RGB values ​​of pixels are not independent; they collectively constitute a finite natural image color space. By analyzing massive amounts of natural images, the intrinsic mathematical relationships between R, G, and B values ​​can be learned. Therefore, when the blue channel (B value) hardware fails, the still normally transmitted R and G values ​​can be used, combined with the pixel color compensation calculation model pre-built and trained in this application, to calculate the most probable and natural B value to compensate for the failed color data channel.

[0057] Step 3 specifically includes the following steps: Step 3-1, construct the pixel color compensation model as follows: B' = a * R + b * G + c Where B' is the RGB value of the color to be compensated, R and G are the RGB values ​​of the first normal color and the second normal color, respectively, and a, b, and c are the first coefficient, the second coefficient, and the third coefficient, respectively.

[0058] The pixel color compensation model is a multiple linear regression model whose goal is to find a set of optimal coefficients (a,b,c) such that for most natural images, the overall error between the B' value calculated by a*R + b*G + c and the true B value is minimized.

[0059] Step 3-2: Construct the training dataset; Data source: Collect a large-scale standard natural image library covering diverse UI display scenarios (device homepage, device individual control, intelligent control, intelligent definition, menu, settings page, etc.) to establish an image data source.

[0060] Preprocess the image data source: Convert all images to the target color space (e.g., sRGB). Extract (R, G, B) triples from all image pixels to form a large number of sample points. , , ), ( , , ) represents the pixel R value, G value, and B value of the i-th sample point.

[0061] Step 3-3: Construct the loss function and optimization objective of the pixel color compensation model; For the i-th sample point, the corresponding pixel color compensation model is: = a * + b * + c The constructed loss function L(a, b, c) is the mean squared error, which is obtained by summing over all training samples: L(a, b, c) = Σ ( - ) 2 The optimization objective is to find a set of (a, b, c) such that the loss function L(a, b, c) is minimized.

[0062] The loss function L(a, b, c) = Σ (B_i - B'_i)^2 is a quadratic convex function (shaped like a semicircular surface) with respect to a, b, and c. This surface has one and only one global minimum. At this point, the rate of change (i.e., the partial derivative) of the function with respect to each independent variable is zero.

[0063] Therefore, by taking the partial derivatives with respect to a, b, and c and setting them to zero, we obtain a set of equations (i.e., the normal equations obtained in step three). Solving the normal equations reveals the coordinates of the lowest point of this "semicircular" surface, which is the (a, b, c) point that minimizes the loss.

[0064] The detailed derivation of the loss function is as follows: (1) Expand the loss function: L(a, b, c) = Σ [ - 2 (a + b + c) + (a + b + c)^2] After simplification, we obtain the quadratic polynomial of L in terms of a, b, and c as follows: L(a,b,c)=a²Σ + b²Σ + c²N + 2abΣ + 2acΣ + 2bcΣ -2aΣ -2bΣ - 2cΣ +Σ

[0065] Where N is the total number of samples, and Σ represents the summation of the samples.

[0066] (2) Take the partial derivative with respect to the first coefficient a and set its value to zero: L / a = Σ [ -2 ( - a - b - c) ] = 0 → Σ = a Σ + b Σ )+ c Σ

[0067] (3) Take the partial derivative with respect to the second coefficient b and set its value to zero: L / b = Σ [ -2 ( - a - b - c) ] = 0 → Σ = a Σ + b Σ + c Σ

[0068] (4) Take the partial derivative with respect to the third coefficient c and set its value to zero: L / c = Σ [ -2( - a - b - c) ] = 0 → Σ = a Σ + b Σ + cN; Steps 3-4: Solve for the optimal coefficients of the pixel color compensation model based on the least squares method, combined with the loss function and the optimization objective; Based on the principle and derivation of step 3-3, the calculation equations involved in steps (2), (3), and (4) of the loss function derivation process are written in matrix form, normal equations are constructed, and the following normal equations are solved:

[0069] The above normal equation can be denoted as: X T Xβ = X T Y Where X is an N x 3 design matrix, and each row of the matrix is ​​[ 1]; Y is an N x 1 column vector, consisting of all composition; β is the 3×1 coefficient vector [abc] to be solved; The optimal solution for obtaining the coefficient vector β is: β = (X T X) -1 X T Y The globally optimal coefficient vector β can be solved in one step using standard matrix inversion or more stable numerical algorithms (such as Cholesky decomposition and QR decomposition). Based on the coefficient vector β, the optimal values ​​of the first coefficient a, the second coefficient b, and the third coefficient c in the pixel color compensation model can be obtained.

[0070] Steps 3-5: Train, validate, and adjust the training dataset based on the training dataset to obtain the trained pixel color compensation model; The training data is divided into a training set and a validation set according to a certain ratio (usually 8:2).

[0071] Calculate the coefficients (a, b, c) using the training set, and then calculate the average error between the estimated B' and the true B on the validation set to verify whether the coefficients meet the requirements.

[0072] It can be fine-tuned using a dedicated dataset for specific application scenarios (such as air conditioner UI interfaces, washing machine UI interfaces, etc.) to obtain more suitable coefficients.

[0073] Steps 3-6 involve using the trained pixel color compensation model to compensate for the faulty data channels, thus completing the repair.

[0074] Furthermore, the following are examples of how the trained pixel color compensation model can be used: By collecting natural image samples for training, a set of optimal coefficients was obtained (taking 8-bit RGB as an example, with values ​​ranging from 0 to 255): a = 0.20, b = 0.65, c = 12.0; The physical meaning of these coefficients is that in natural images, the blue component is positively correlated with the red and green components, with green contributing the most to blue (coefficient 0.65), followed by red (0.20), plus a constant bias of about 12 (to compensate for dark details).

[0075] Fault Scenario Setting A hardware failure occurred in the device's blue data channel (such as a broken data cable or a short circuit to ground), causing the blue component received by the display screen to consistently show as 0. The red and green channels are functioning normally.

[0076] For example, the true color of a pixel in the original image is (R=120, G=180, B=200) (light blue). Due to a blue channel malfunction, the actual color received by the display screen is (120, 180, 0), which appears as a dark yellow and is severely distorted.

[0077] Compensation calculation process Upon detecting an anomaly in the blue channel, the system automatically activates the compensation model. For each pixel, using the normal red value R and green value G, the replacement blue value B' is calculated according to the formula: B' = a × R + b × G + c Substituting the values ​​of a, b, and c into the above values, we get the following: B' = 0.20 × 120 + 0.65 × 180 + 12.0= 24 + 117 + 12= 153 The compensation module replaces the original pixel (120, 180, 0) with (120, 180, 153) and then sends it to the display screen.

[0078] Effect Comparison The original true color pixels are (120, 180, 200), and the blue component is light blue (normal). The pixel after the fault is (120, 180, 0), at which point the blue component appears dark yellow (severe distortion). The compensated pixels are (120, 180, 153), and the blue component displays a bluish-cyan color (close to the original, acceptable). Visual differences: The original light blue (120,180,200) leans towards sky blue; after compensation (120,180,153), it leans towards blue-cyan, with slightly lower saturation, but retains the basic blue tone, and the content of the image is still recognizable.

[0079] As can be seen, the repair method proposed in this invention can maintain the basic color tone of the actual pattern to be displayed, ensuring that the content is clearly visible and improving the user experience of the product.

[0080] Batch processing example of multiple pixels Assuming there are multiple pixels in a frame of image, the compensation module calculates the values ​​pixel by pixel as shown in Table 4 below.

[0081] Table 4: Schematic table of multi-pixel compensation for a single frame of image

[0082] Boundary case handling Cropping out of range: The calculated B' may be less than 0 or greater than 255. For example, for an extremely bright pixel (R=255, G=255): B' = 0.20×255+0.65×255+12 = 51+165.75+12 = 228.75, which is still within the range of 0-255. If a pixel is calculated to have B'=270, it will be cropped to 255; if B'=-5, it will be cropped to 0.

[0083] Integerization: The final output must be an integer, which can be rounded to the nearest integer.

[0084] like Figure 2As shown, another embodiment of the present invention also proposes a display screen data channel anomaly detection and repair system to implement the above-mentioned display screen data channel anomaly detection and repair method. In the display driving link, the present invention introduces a data channel detection and repair system located after the frame buffer. By analyzing the RGB data output to the MIPI host, the present invention realizes the anomaly detection and repair of the display screen data channel. In the present invention, the MIPI host refers to the DSI Host (Display Serial Interface Host).

[0085] Specifically, the system includes: The anomaly detection module is used to detect anomalies in the display screen by determining whether an abnormal data channel has occurred based on an improved pixel value comparison method. The anomaly detection methods in this invention include an active diagnosis mode and a passive detection mode, as detailed below: Active diagnostic modes include: When the display is powered on, woken up, or periodically idle, a series of known solid color test patterns are sent to the display. The actual display data packets in the status register inside the display are read through the reverse channel of the MIPI D-PHY data line to determine whether the actual display data packets match the expectations. Among them, the solid color test pattern can be a full blue, full red, full green, or full white image. For example, when sending a full blue image, if the brightness of the blue pixels is detected to be much lower than expected or there is no response at all, while the red and green images are normal, then the blue data channel is determined to be abnormal.

[0086] Passive detection modes include: During display operation, RGB frame data is sent. The actual display data packet in the status register of the display is read through the reverse channel of the MIPI D-PHY data line for anomaly detection. The actual display data packet and the actual sent RGB frame data are analyzed and compared in real time to determine whether the display has an anomaly.

[0087] Furthermore, the method of determining whether an abnormal data channel exists based on improved pixel value comparison specifically includes: The image color is verified to be abnormal by means of the image pixel packing rules. The actual display data packet is read from the reverse channel of the MIPI D-PHY data line into the status register of the display screen and CRC and ECC checks are performed. If both CRC and ECC checks pass, the data packet that has passed the check is obtained and enters the high-level verification stage. The high-level verification stage determines whether there is an abnormal data channel.

[0088] The high-level verification process specifically includes: The data packets that pass the test are decompressed to obtain the values ​​of specific color components, and the deviation ratio between the values ​​of specific color components and the expected values ​​is calculated. If the deviation ratio between the value of a specific color component after unpacking and the expected value exceeds a preset threshold, it indicates that the data channel corresponding to that color component is abnormal. If the deviation ratio between the value of a specific color component after unpacking and the expected value does not exceed a preset threshold, it indicates that the data channel corresponding to that color component is normal.

[0089] As a preferred embodiment, the deviation ratio Δ between the value of a specific color component and the expected value is calculated as follows: Δ=(∣Vcurrent Vref∣ / max(Vmax Vmin, ))×100% Where Vcurrent is the current actual pixel value of the LCD display, Vref is the actual pixel value to be displayed, and Vmax and Vmin are the maximum and minimum values ​​of the statistical pixels within the local window, respectively. It is a local minimum.

[0090] The compensation trigger module is used to construct a compensation trigger mechanism, which is triggered when the compensation trigger conditions are met. Specifically, a compensation trigger mechanism is constructed, which is triggered when the compensation trigger conditions are met. This includes: The display screen is divided into areas, and counters are set for different color data channels in each area for fault monitoring. When an abnormality is detected in a certain color data channel in a certain area, the counter corresponding to that color data channel in that area is incremented by 1. If the counter corresponding to a certain color data channel in a certain area counts more than a preset counting threshold within a preset time window, a compensation mechanism will be triggered for the abnormal data channel of that color in that area; otherwise, the compensation mechanism will not be triggered.

[0091] The model building and training module is used to build a pre-trained pixel color compensation model; building the pre-trained pixel color compensation model specifically includes: Construct a pixel color compensation model; Build the training dataset; Construct the loss function and optimization objective of the pixel color compensation model; The optimal coefficients of the pixel color compensation model are solved by combining the loss function and the optimization objective. The training dataset is used to train, validate, and adjust the training dataset to obtain the trained pixel color compensation model.

[0092] For the i-th pixel on the display screen, the compensation value obtained through the pixel color compensation model is as follows: =a× +b× +c in, The RGB compensation value of the i-th pixel, obtained through the pixel color compensation model, represents the color to be compensated. , ...

[0093] The loss function L(a,b,c) for constructing the pixel color compensation model is as follows: L(a,b,c)=Σ( - ) 2 in, Let be the RGB value of the color to be compensated for at the i-th pixel in the training data; The optimization objective is to minimize the loss function value L(a,b,c).

[0094] The repair module is used to compensate and repair abnormal data channels that trigger the compensation mechanism based on the trained pixel color compensation model.

[0095] This invention also provides a display terminal, which includes: a display screen data channel anomaly detection and repair system as proposed in the above embodiments.

[0096] The beneficial effect of the present invention is that, compared with the prior art, the method proposed in the present invention determines whether each primary color data channel is abnormal by constructing a display screen data channel anomaly detection method, and after confirming that a certain primary color data channel is abnormal, it activates a color compensation method based on the other normal data channels to perform color repair when the compensation trigger condition is met, thereby achieving accurate detection and timely compensation repair of the display screen data channels and improving the user's viewing experience.

[0097] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0098] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0099] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0100] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for detecting and repairing abnormalities in the data channel of a display screen, characterized in that, Includes the following steps: Anomaly detection is performed on the display screen, and an improved pixel value comparison method is used to determine whether there are abnormal data channels. Construct a compensation trigger mechanism to trigger the compensation mechanism when the compensation trigger conditions are met; A pre-trained pixel color compensation model is constructed. When the compensation mechanism is triggered, the abnormal data channel triggered by the compensation mechanism is compensated and repaired based on the normal data channel using the pre-trained pixel color compensation model.

2. The method for detecting and repairing abnormal data channels of a display screen according to claim 1, characterized in that, The abnormality detection of the display screen includes active detection and passive detection; Active detection includes: sending a test pattern to the display screen when the display screen is powered on, woken up, or periodically idle, and reading the actual display data packet of the status register inside the display screen through the reverse channel of the MIPI D-PHY data line to detect anomalies; Passive detection includes: sending RGB frame data while the display is running, and reading the actual display data packets from the status register inside the display via the reverse channel of the MIPI D-PHY data line to perform anomaly detection.

3. The method for detecting and repairing abnormal data channels of a display screen according to claim 2, characterized in that, The method for determining whether an abnormal data channel exists based on improved pixel value comparison specifically includes: The image color is verified to be abnormal by means of the image pixel packing rules. The actual display data packet is read from the reverse channel of the MIPI D-PHY data line into the status register of the display screen and CRC and ECC checks are performed. If both CRC and ECC checks pass, the data packet that has passed the check is obtained and enters the high-level verification stage. The high-level verification stage determines whether there is an abnormal data channel.

4. The method for detecting and repairing abnormal data channels of a display screen according to claim 3, characterized in that, The high-level verification process specifically includes: The data packets that pass the test are decompressed to obtain the values ​​of specific color components, and the deviation ratio between the values ​​of specific color components and the expected values ​​is calculated. If the deviation ratio between the value of a specific color component after unpacking and the expected value exceeds a preset threshold, it indicates that the data channel corresponding to that color component is abnormal. If the deviation ratio between the value of a specific color component after unpacking and the expected value does not exceed a preset threshold, it indicates that the data channel corresponding to that color component is normal.

5. The method for detecting and repairing abnormal data channels of a display screen according to claim 4, characterized in that, The deviation ratio Δ between the specific color component value and the expected value is calculated as follows: Δ=(∣Vcurrent Vref∣ / max(Vmax Vmin, ))×100% Where Vcurrent is the current actual pixel value of the LCD display, Vref is the actual pixel value to be displayed, and Vmax and Vmin are the maximum and minimum values ​​of the statistical pixels within the local window, respectively. It is a local minimum.

6. The method for detecting and repairing abnormal data channels of a display screen according to claim 1, characterized in that, The aforementioned compensation triggering mechanism, which triggers the compensation mechanism when the compensation triggering conditions are met, specifically includes: The display screen is divided into areas, and counters are set for different color data channels in each area for fault monitoring. When an abnormality is detected in a certain color data channel in a certain area, the counter corresponding to that color data channel in that area is incremented by 1. If the counter corresponding to a certain color data channel in a certain area counts more than a preset counting threshold within a preset time window, a compensation mechanism will be triggered for the abnormal data channel of that color in that area; otherwise, the compensation mechanism will not be triggered.

7. The method for detecting and repairing abnormal data channels of a display screen according to claim 1, characterized in that, The construction of the pre-trained pixel color compensation model specifically includes: Construct a pixel color compensation model; Build the training dataset; Construct the loss function and optimization objective of the pixel color compensation model; The optimal coefficients of the pixel color compensation model are solved by combining the loss function and the optimization objective. The training dataset is used to train, validate, and adjust the training dataset to obtain the trained pixel color compensation model. The faulty data channel is compensated by the trained pixel color compensation model, thus completing the repair.

8. The method for detecting and repairing abnormal data channels of a display screen according to claim 7, characterized in that, For the i-th pixel on the display screen, the compensation value obtained through the pixel color compensation model is as follows: =a× +b× +c in, The RGB compensation value of the i-th pixel, obtained through the pixel color compensation model, represents the color to be compensated. , ...

9. The method for detecting and repairing abnormal data channels of a display screen according to claim 8, characterized in that, The loss function L(a,b,c) for constructing the pixel color compensation model is as follows: L(a,b,c)=Σ( - ) 2 in, Let be the RGB value of the color to be compensated for at the i-th pixel in the training data; The optimization objective is to minimize the loss function value L(a,b,c).

10. A display screen data channel anomaly detection and repair system, used to implement the display screen data channel anomaly detection and repair method according to any one of claims 1-9, characterized in that, include: The anomaly detection module is used to detect anomalies in the display screen by determining whether an abnormal data channel has occurred based on an improved pixel value comparison method. The compensation trigger module is used to construct a compensation trigger mechanism, which is triggered when the compensation trigger conditions are met. The model building module is used to build a pre-trained pixel color compensation model; The repair module is used to compensate and repair abnormal data channels that trigger the compensation mechanism based on the trained pixel color compensation model.

11. A display terminal, characterized in that, include: The display screen data channel anomaly detection and repair system as described in claim 10.

Images