Micro-led dead pixel repair system and method

Abstract

The MicroLED bad point repair system and method disclosed by the application belong to the technical field of MicroLED display, and solve the problem of life attenuation caused by passive compensation and uneven load of existing MicroLED display bad point repair. The bad point repair system comprises a pixel risk detection module for real-time acquisition and calculation of a risk value; a permanent disconnection control module for isolating the corresponding pixel when the risk value exceeds a threshold value; a self-healing interconnection network for establishing a backup logical connection with the neighborhood pixels; a logical pixel migration module for scoring the healthy pixels in the neighborhood and migrating the bad point pixel display task to the pixel with the highest score; and a load balancing control module for dynamic rotation of the bearing pixels. The application realizes the transformation of the display system from post-failure repair to pre-failure prevention, ensures the display continuity and brightness consistency without a large number of backup pixels, and significantly improves the display reliability and service life.

Description

Technical Field

[0001] This invention belongs to the field of MicroLED (Micro Light Emitting Diode Display) display technology, specifically relating to a MicroLED dead pixel repair system. This invention also relates to a MicroLED dead pixel repair method. Background Technology

[0002] In recent years, MicroLED has been considered a key technology for next-generation near-eye displays due to its advantages such as high brightness, high contrast, low power consumption, fast response, and long lifespan. Especially in the field of AR (Augmented Reality) micro-displays, MicroLED has become an important solution for achieving high brightness and low power consumption. However, as the pixel size of MicroLED continues to shrink and the number of pixels continues to increase, its manufacturing difficulty and long-term reliability issues are becoming increasingly prominent. For MicroLED pixels with dimensions of only a few micrometers or even sub-micrometers, any tiny manufacturing defect can lead to pixel failure. For example, epitaxial defects, etching damage, bonding voids, metal fractures, electrostatic discharge (ESD) damage, and driver transistor failure can all result in dark spots, bright spots, flickering spots, or leakage pixels. In AR micro-displays, since the user's eyes are only a few centimeters away from the display chip, even a single bad pixel can significantly affect the visual experience, thus placing extremely high demands on display reliability.

[0003] Currently, MicroLED displays mainly employ three types of defective pixel repair methods. The first type involves repairing defective pixels during the manufacturing stage using laser repair or mass transfer. However, this method is only applicable to pre-shipment repair and cannot address newly added defective pixels during long-term use. The second type uses software interpolation algorithms to estimate the grayscale of the defective pixel based on surrounding pixels. However, the defective pixel itself still exists, only improving the visual effect and not restoring the actual light-emitting capacity. The third type uses a redundant pixel scheme, reserving a certain proportion of spare pixels to replace failed pixels. However, this method reduces the display aperture ratio, and the spare resources are limited, eventually leading to failure as the number of defective pixels continues to increase.

[0004] Meanwhile, most existing solutions are "post-failure repairs," meaning compensation is only performed when a pixel is completely damaged, without considering that pixel failure typically manifests as decreased brightness, abnormal current, increased leakage current, or excessively long cumulative lifespan. If high-risk pixels can be identified in advance and their workload proactively reduced, dead pixel formation can be effectively delayed, thus extending the lifespan of the entire display system.

[0005] Furthermore, in actual use of AR glasses, the central area displays menus, navigation information, and the UI (User Interface) for extended periods, while the peripheral areas remain static most of the time. This results in the central area pixels accumulating significantly more working time than the peripheral pixels, leading to noticeable localized aging. Traditional fixed-drive methods cannot achieve load balancing, limiting the lifespan of the entire display chip to a few high-load areas.

[0006] Therefore, there is an urgent need for a new MicroLED display architecture that can simultaneously possess pixel risk prediction, permanent isolation of bad pixels, self-healing interconnection, logic migration and dynamic load balancing capabilities, in order to achieve long-term reliable, high-quality and long-life display. Summary of the Invention

[0007] The purpose of this invention is to provide a MicroLED dead pixel repair system that solves the problem of lifespan degradation caused by passive compensation and uneven load in existing MicroLED display dead pixel repair.

[0008] Another objective of this invention is to provide a method for repairing defective pixels in MicroLEDs.

[0009] The first technical solution adopted in this invention is: a MicroLED defective pixel repair system, wherein each pixel in the MicroLED is connected to an independent driving circuit in the CMOS (Complementary Metal Oxide Semiconductor) backplane, and further includes: The pixel risk detection module is used to collect the driving current, luminous brightness and operating temperature of each pixel driving circuit in real time and calculate the risk value. The permanent disconnect control module is used to isolate the corresponding pixel when the risk value exceeds the threshold. A self-healing interconnect network is used to establish backup logical connections with neighboring pixels; The logical pixel migration module is used to score healthy pixels in the neighborhood based on geometric distance, brightness consistency, current number of times it is handled, and remaining lifespan, and to migrate the display task of bad pixels to the pixel with the highest score. The load balancing control module is used to dynamically rotate the pixels that are responsible for a certain task.

[0010] The first technical solution of this invention is further characterized by: Risk value in pixel risk detection module The calculation method is as follows:

[0011] In the formula, For the first Each pixel drives the current. For the first The brightness of each pixel For the first Operating temperature of each pixel , , These are all set reference values. , , All are weighting coefficients.

[0012] The pixel risk detection module calculates the risk value. Pixels are divided into three levels: healthy state, warning state, and dangerous state. When the risk value is below the first threshold, it is in a healthy state and the pixel continues to work normally. When the risk value exceeds the first threshold but is below the second threshold, it is in a warning state and the effective driving time of the pixel in the display driving cycle is reduced. When the risk value exceeds the second threshold, it is in a dangerous state and the permanent disconnection control module is triggered.

[0013] When the pixel risk detection module calculates the risk value When the second threshold is exceeded, the permanent disconnection control module cuts off the driving circuit of the corresponding pixel via a MOS (Metal Oxide Semiconductor) switch, electronic fuse, one-time programmable fuse, or non-volatile memory control switch. Risk value Pixels exceeding the second threshold are isolated.

[0014] The self-healing interconnect network adopts a two-dimensional mesh structure, a crossbar interconnect structure, or a local neighborhood interconnect structure. Each pixel establishes a backup logical connection with at least four adjacent pixels above, below, left, and right. When a bad pixel is isolated, it searches for neighboring pixels until it finds the responsible pixel and re-establishes the display mapping relationship.

[0015] The scoring of healthy pixels (i.e., candidate undertaking pixels) within the neighborhood in the logical pixel migration module. Represented as:

[0016] In the formula, The geometric distance between candidate pixels and bad pixels. To ensure brightness consistency between candidate pixels and defective pixels, For the first The number of times a candidate pixel undertakes a task at the end of each load statistics cycle. For the first The predicted remaining lifetime of the candidate undertaking pixel at the end of each load statistics cycle. a , b , c , d All are weighting coefficients; The healthy pixel with the highest score is selected to handle the display task corresponding to the bad pixel.

[0017] Predicting remaining lifespan The calculation method is expressed as follows:

[0018] In the formula, The initial lifetime of the candidate pixel is determined. For the first The cumulative working time of the candidate pixels at the end of each load statistics cycle. α , β This represents the aging coefficient.

[0019] The load balancing control module is dynamically updated based on the number of times a pixel is handled. When the preset threshold is exceeded, the logical pixel migration module recalculates the candidate pixel score and selects the healthy pixel with the highest score to take over the display task corresponding to the bad pixel, thereby realizing dynamic rotation of display tasks.

[0020] The second technical solution adopted in this invention is: a MicroLED defective pixel repair method, which collects the operating parameters of each pixel and calculates the risk value, isolates defective pixels according to the risk value, searches for healthy pixels in the neighborhood of the defective pixel and evaluates and selects the responsible pixel, then establishes a logical mapping to migrate the display task to the responsible pixel, and dynamically rotates the responsible pixel based on the cumulative number of times it has been responsible.

[0021] The second technical solution of this invention is further characterized by: a MicroLED defect repair method, comprising the following steps: Step 1: Collect the operating parameters of each MicroLED pixel in real time, including driving current, luminous intensity and operating temperature; Step 2: Calculate the risk value of each MicroLED pixel based on the collected operating parameters; Step 3: Compare the risk value of each MicroLED pixel with a preset threshold to determine the status of each MicroLED pixel; Step 4: Cut off the driving path of the MicroLED pixels that are determined to be in a dangerous state; Step 5: For bad pixels whose driving paths have been cut off, search for and select the responsible pixels from the healthy pixels in the neighborhood based on the geometric distance between them and the healthy pixels in the neighborhood, brightness consistency, current number of times they have been handled, and remaining lifetime score. Step 6: Transfer the display task corresponding to the bad pixel to the selected responsible pixel; Step 7: Record the cumulative number of times the assigned pixel is assigned. When the cumulative number of times the assigned pixel is assigned exceeds the preset threshold, jump to step 5 to realize the dynamic rotation of the assigned pixel.

[0022] The beneficial effects of this invention are as follows: The MicroLED defective pixel repair system and method of this invention assess the health status in real time through a pixel risk detection module and reduce the load of high-risk pixels in advance. Combined with a permanent disconnection control module, defective pixels are completely isolated to avoid leakage crosstalk. Then, a self-healing interconnection network and a logical pixel migration module are used to migrate the display function to the optimal healthy pixel. The load balancing control module dynamically rotates the pixels to ensure uniform aging. This realizes the transformation of the MicroLED display system from post-failure repair to pre-failure prevention. It ensures display continuity and brightness consistency without the need for a large number of spare pixels, and significantly improves display reliability and service life. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the MicroLED defect repair system of the present invention. Detailed Implementation

[0024] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0025] Example 1 This invention provides a MicroLED defect repair system and method. By establishing a pixel health status assessment model and combining a reconfigurable interconnect network and a dynamic load balancing mechanism, it realizes a novel display architecture for proactive prediction, permanent isolation, display function migration, and lifespan balancing of MicroLED displays. It is applicable to AR near-eye displays, VR (Virtual Reality) displays, automotive HUDs (Head-Up Displays), high-reliability industrial displays, and ultra-high resolution MicroLED micro-display systems. The system includes a MicroLED display array, a CMOS driving circuit, a pixel risk detection module, a permanent disconnection control module, a self-healing interconnect network, a logical pixel migration module, and a load balancing control module.

[0026] Through the above methods, this invention, for the first time, organically combines pixel risk prediction, self-healing interconnection, and load balancing to achieve proactive health management of MicroLED display systems. The risk model can identify potentially failing pixels in advance, shifting from "repair after failure" to "prevention before failure." A permanent disconnection mechanism can completely isolate bad pixels, avoiding leakage and crosstalk, thus improving display reliability. A reconfigurable interconnect network enables display function migration, eliminating the need for a large number of fixed spare pixels and improving pixel utilization. A pixel evaluation model is established, significantly improving the continuity and brightness consistency of the migrated image. A dynamic load balancing mechanism is proposed, effectively reducing the aging rate of local high-load areas and extending the lifespan of the display chip. It is compatible with GaN MicroLED, AlGaInP red MicroLED, and monolithic full-color MicroLED, and can be widely used in AR, VR, and high-reliability display systems.

[0027] Example 2 This invention provides a MicroLED defect repair system and method. Based on Example 1, the MicroLED display array consists of several RGB (Red, Green, Blue) pixels arranged in a matrix, with each pixel corresponding to an independent driving circuit. The pixel size is preferably 2–20 μm, and the pixel pitch is preferably 3–10 μm. GaN blue-green light and AlGaInP red light epitaxy can be used, or a monolithic full-color MicroLED structure can be used.

[0028] The CMOS driver circuit is located below the display array and is responsible for pixel driving, grayscale control, PWM (Pulse Width Modulation), current sampling, and data reading. Each driver unit can collect the driving current, luminance, and operating temperature of the corresponding pixel in real time and transmit the relevant data to the pixel risk detection module.

[0029] Example 3 This invention provides a MicroLED defect repair system and method, based on Example 1, such as... Figure 1 As shown, the pixel risk detection module is used to evaluate the health status of each MicroLED pixel in real time and establish a pixel risk value model. The risk value comprehensively considers multiple parameters such as driving current deviation, brightness deviation, and temperature deviation. Its calculation formula can be expressed as:

[0030] in: For the first Drive current per pixel; For the first The brightness of each pixel; For the first Operating temperature of each pixel; , , These are all set reference values; , , All are weighting coefficients, and their sum equals 1.

[0031] The higher the risk value, the closer the pixel is to failure. Risk levels can be categorized as healthy, warning, and dangerous. When the risk value exceeds a first threshold... R When the threshold is 1, the system automatically reduces the effective driving time and driving current of that pixel within the display driving cycle, thus reducing the load. When a pixel experiences a permanent open circuit, a permanent short circuit, or its risk value continuously exceeds the second threshold... R At step 2, the permanent disconnection control module controls the corresponding drive unit to cut off its drive path and starts the subsequent migration process.

[0032] Example 4 This invention provides a MicroLED defect repair system and method. Based on Embodiment 1, the permanent disconnection control module can be implemented by controlling the MOS switch through a MOS switch, a fuse, an electronic fuse, or non-volatile memory. The disconnection criterion can be expressed as:

[0033] In the formula, For pixel health status index, This indicates that the pixel is in a healthy state. This indicates that the pixel is in a warning state. This indicates that the pixel is in a dangerous state; This method can prevent failed pixels from continuing to leak electricity, emit abnormal light, or affect adjacent driving circuits, thereby improving the reliability of the entire display array.

[0034] Example 5 This invention provides a MicroLED defective pixel repair system and method. Based on Embodiment 1, a self-healing interconnect network is used to re-establish logical connections after a defective pixel is disconnected. This means that the image display signal that the defective pixel should have displayed is redirected at the data mapping level to the driving end of a neighboring healthy pixel, allowing the healthy pixel to logically replace the defective pixel in receiving and displaying image data, thus forming a virtual connection relationship. Preferably, a two-dimensional mesh network structure, a crossbar interconnect structure, or a local neighborhood interconnect structure is used. Each pixel establishes backup logical connections with at least four adjacent pixels (upper, lower, left, and right), and this can be extended to eight neighborhoods or even larger neighborhoods. After the defective pixel is isolated, the controller automatically searches for surrounding healthy pixels and re-establishes the display mapping relationship based on distance, brightness consistency, and load conditions, thereby ensuring continuous image display.

[0035] In the logical pixel migration module, the responsible pixel is automatically selected by the scoring function:

[0036] in: The geometric distance between candidate pixels and bad pixels; For brightness consistency, that is, the brightness difference between candidate pixels and bad pixels; For the first The number of times a candidate pixel has undertaken a task or the cumulative load value at the end of each load statistics period. For the first The predicted remaining lifetime of the candidate undertaking pixel at the end of each load statistics cycle. a , b , c , d All are weighting coefficients.

[0037] The controller prioritizes the nearest pixel with the closest brightness, the lowest current load, and the longest remaining lifespan to handle the task of displaying bad pixels, thereby reducing image distortion.

[0038] To avoid the need for pixels to operate continuously for extended periods, this invention establishes a load balancing control module:

[0039] in: The load statistics period is numbered and can be divided according to the system's preset refresh count, display time, or working duration, rather than a single frame refresh period. For the first The cumulative load value after the end of each load statistics cycle, where 1 indicates that the pixel has undertaken at least one logical display task within the current load statistics cycle.

[0040] Number of times Reaching the preset threshold Then, the controller automatically recalculates the responsible pixels to achieve the rotation of the responsible objects.

[0041] In addition, lifespan prediction models can be established:

[0042] in: The initial lifetime of the candidate pixel is determined. For the first The cumulative working time of the candidate pixels at the end of each load statistics cycle. α , β All are aging coefficients.

[0043] The controller prioritizes healthy pixels with longer lifespans to undertake display tasks, making the aging of the entire display array more uniform.

[0044] Example 6 The MicroLED dead pixel repair system mainly includes a MicroLED display array, a driver backplane, a pixel risk detection module, a permanent disconnection control module, a self-healing interconnect network, a logical pixel migration module, and a load balancing control module.

[0045] The MicroLED display array preferably employs GaN blue-green MicroLEDs, AlGaInP red MicroLEDs, or monolithic full-color MicroLED structures, with each MicroLED pixel connected to an independent CMOS driving circuit in the driving backplane. The pixel size is preferably 2–20 μm, the pixel pitch is preferably 3–10 μm, and the display resolution can be selected from 640×480, 1280×720, 1920×1080, or higher resolutions depending on the application, to meet the requirements of AR near-eye display, high-brightness display, and high pixel density display.

[0046] The driver backplane is responsible not only for driving the current and controlling the grayscale of each MicroLED pixel, but also integrates current detection and operating temperature statistics functions. During system operation, each driver unit continuously collects the driving current, luminous intensity, and operating temperature of the corresponding pixel, and transmits these parameters in real time to the pixel risk detection module, forming a health status database for the entire display array. To ensure sufficiently high real-time performance of risk detection, the detection cycle is preferably 10–100 ms. In high-speed AR display mode, it can also be updated synchronously according to the refresh cycle of each frame, enabling the risk assessment to reflect the pixel health status in real time.

[0047] The pixel risk detection module establishes a pixel risk assessment model based on the collected operating parameters. For any pixel in the display array, the system calculates the corresponding risk value by comprehensively considering factors such as drive current deviation, brightness attenuation, and operating temperature deviation. The risk value can be expressed as:

[0048] Among them, the weighting coefficient , , The preferred values ​​are 0.30–0.40, 0.20–0.35, and 0.20–0.30, respectively, with the sum of each weight being 1. Through the above model, the system can simultaneously reflect the instantaneous working state and long-term aging trend of a pixel, rather than relying solely on a single parameter to determine whether a pixel has failed.

[0049] Based on the calculated risk value, the system classifies all pixels into three levels: healthy state, warning state, and dangerous state. When the risk value is below the first threshold, the pixel continues to operate normally; when the risk value is between the first and second thresholds, the controller records that the pixel has entered the warning state and appropriately reduces its effective driving time or peak current within the display driving cycle to delay further aging; when the risk value continues to be above the second threshold, or when an unrecoverable failure such as permanent open circuit, permanent short circuit, or abnormal leakage is detected in the pixel, the permanent disconnection control module is immediately activated to perform a permanent isolation operation on the pixel.

[0050] The permanent disconnection control module is preferably implemented using a MOS switch, electronic fuse (eFuse), one-time programmable fuse (OTP Fuse), or non-volatile memory control switch on the CMOS backplane. When the controller confirms that a pixel has entered an unrecoverable state, the drive path is permanently shut down, preventing it from participating in subsequent display and thus avoiding abnormal leakage, continuous illumination, or crosstalk affecting surrounding healthy pixels. The corresponding disconnection criterion can be expressed as:

[0051] in, This indicates that the corresponding pixel has been permanently removed from the display system.

[0052] When a defective pixel is permanently disconnected, the self-healing interconnect network begins to re-establish the display mapping relationship. This invention preferably employs a two-dimensional mesh interconnect network, where each pixel establishes backup connections with at least four adjacent pixels (top, bottom, left, and right). In another preferred embodiment, an eight-neighborhood interconnect structure can be further established, ensuring that each pixel maintains a logical connection with eight surrounding healthy pixels. When a defective pixel appears, the system first searches its first-order neighborhood; if no healthy pixel meets the conditions within the neighborhood, it continues to expand to the second-order neighborhood or even a larger range until a suitable replacement pixel is found. For larger display arrays, Crossbar interconnects or globally reconfigurable networks can also be used to improve defective pixel repair capabilities.

[0053] To ensure display quality after migration, this invention further establishes a pixel evaluation model. The controller comprehensively considers multiple factors such as the geometric distance between candidate pixels and bad pixels, brightness consistency, current load, and remaining lifetime to calculate a comprehensive score. Its evaluation function can be expressed as:

[0054] Among them, the weighting coefficient a , b , c and dThe preferred values ​​are 0.35, 0.30, 0.20, and 0.15, respectively. The controller ultimately selects the healthy pixel with the highest overall score to handle the display task corresponding to the defective pixel, thereby maximizing the continuity of the displayed image and the uniformity of brightness.

[0055] Since prolonged operation of a single pixel may create new high-load areas, this invention further introduces a load balancing control module to continuously record the cumulative number of times each pixel has been loaded and to establish dynamic updates.

[0056] When the number of tasks exceeds a preset threshold, the controller recalculates the task assignment, causing the display task to periodically rotate among multiple healthy pixels. Simultaneously, the system establishes a remaining lifetime prediction model.

[0057] By simultaneously considering working time and load, the controller always prioritizes the healthiest pixels with the longest lifespan and lowest load to participate in the migration, thereby making the aging of the entire display array more uniform and effectively avoiding premature failure in local areas.

[0058] During actual operation, when the display array is running, the CMOS driver backplane continuously collects the operating parameters of each pixel and calculates the risk value in real time. If a pixel enters a dangerous state, the controller immediately cuts off the driving path of that pixel, the self-healing interconnect network automatically searches for the optimal replacement pixel and re-establishes the logical mapping, and then the new replacement pixel continues to complete the display task corresponding to the original bad pixel. The entire migration process requires no user intervention and can be completed in milliseconds without affecting normal image refresh. After the system has been running for a long time, the load balancing module continues to dynamically optimize the load distribution, ensuring that the entire display array always maintains a uniform aging state, thereby significantly improving the long-term reliability, display uniformity, and lifespan of the MicroLED display system.

Claims

1. A MicroLED dead pixel repair system, characterized in that, Each pixel in a MicroLED is connected to an independent driving circuit in the CMOS backplane, and also includes: The pixel risk detection module is used to collect the driving current, luminous brightness and operating temperature of each pixel driving circuit in real time and calculate the risk value. The permanent disconnect control module is used to isolate the corresponding pixel when the risk value exceeds the threshold. A self-healing interconnect network is used to establish backup logical connections with neighboring pixels; The logical pixel migration module scores healthy pixels in the neighborhood based on geometric distance, brightness consistency, current number of assignments, and remaining lifetime, and migrates the display task of bad pixels to the assigned pixel with the highest score. The assigned pixel in the logical pixel migration module is automatically selected by the scoring function. in: The geometric distance between candidate pixels and bad pixels; For brightness consistency, that is, the brightness difference between candidate pixels and bad pixels; For the first The number of times a candidate pixel has undertaken a task or the cumulative load value at the end of each load statistics period. For the first The predicted remaining lifetime of the candidate undertaking pixel at the end of each load statistics cycle. a , b , c , d All are weighting coefficients; the load balancing control module is used to dynamically rotate the pixels it handles; establishing the load balancing control module: in: The load statistics period is numbered and is divided according to the system's preset refresh count, display time, or working duration, rather than a single frame refresh period. For the first The cumulative load value after the end of each load statistics cycle, where 1 indicates that the pixel has undertaken at least one logical display task within the current load statistics cycle; Establish a lifespan prediction model: in: The initial lifetime of the candidate pixel. For the first The cumulative working time of the candidate pixels at the end of each load statistics cycle. α , β All are aging coefficients.

2. The MicroLED dead pixel repair system as described in claim 1, characterized in that, Risk value in the pixel risk detection module The calculation method is as follows: In the formula, For the first Each pixel drives the current. For the first The brightness of each pixel For the first Operating temperature of each pixel , , These are all set reference values. , , All are weighting coefficients.

3. The MicroLED defect repair system as described in claim 2, characterized in that, The pixel risk detection module calculates the risk value. Pixels are divided into three levels: healthy state, warning state, and dangerous state. When the risk value is below the first threshold, it is in a healthy state and the pixel continues to work normally. When the risk value exceeds the first threshold but is below the second threshold, it is in a warning state and the effective driving time of the pixel in the display driving cycle is reduced. When the risk value exceeds the second threshold, it is in a dangerous state and the permanent disconnection control module is triggered.

4. The MicroLED defective pixel repair system as described in claim 3, characterized in that, When the risk value calculated by the pixel risk detection module When the second threshold is exceeded, the permanent disconnection control module cuts off the driving circuit of the corresponding pixel via a MOS switch, electronic fuse, one-time programmable fuse, or non-volatile memory control switch. Risk value Pixels exceeding the second threshold are isolated.

5. The MicroLED dead pixel repair system as described in claim 1, characterized in that, The self-healing interconnection network adopts a two-dimensional Mesh structure, a Crossbar interconnection structure, or a local neighborhood interconnection structure. Each pixel establishes a backup logical connection with at least four adjacent pixels above, below, left, and right. When a bad pixel is isolated, it searches for neighboring pixels until it finds the responsible pixel and re-establishes the display mapping relationship.

6. The MicroLED defect repair system as described in claim 1, characterized in that, The load balancing control module establishes dynamic updates based on the number of times a pixel is handled. When the preset threshold is exceeded, the logical pixel migration module recalculates the candidate pixel score and selects the healthy pixel with the highest score to take over the display task corresponding to the bad pixel, thereby realizing dynamic rotation of display tasks.

7. A method for repairing MicroLED dead pixels, using the MicroLED dead pixel repair system as described in claim 1, characterized in that, Collect the operating parameters of each pixel and calculate the risk value. Isolate bad pixels based on the risk value, then search for healthy pixels in the neighborhood of the bad pixels and evaluate and select the responsible pixel. After that, establish a logical mapping to migrate the display task to the responsible pixel, and dynamically rotate the responsible pixel based on the cumulative number of times it has been undertaken. This includes the following steps: Step 1: Collect the operating parameters of each MicroLED pixel in real time, including driving current, luminous brightness and operating temperature; Step 2: Calculate the risk value of each MicroLED pixel based on the collected operating parameters; Step 3: Compare the risk value of each MicroLED pixel with a preset threshold to determine the status of each MicroLED pixel; Step 4: Cut off the driving path of the MicroLED pixels that are determined to be in a dangerous state; Step 5: For bad pixels whose driving paths have been cut off, search for and select the responsible pixels from the healthy pixels in the neighborhood based on the geometric distance between them and the healthy pixels in the neighborhood, brightness consistency, current number of times they have been handled, and remaining lifetime score. Step 6: Transfer the display task corresponding to the bad pixel to the selected responsible pixel; Step 7: Record the cumulative number of times the assigned pixel is assigned. When the cumulative number of times the assigned pixel is assigned exceeds the preset threshold, jump to step 5 to realize the dynamic rotation of the assigned pixel.

Images