Full-color module bad point dynamic replacement method and LED display screen

Abstract

The application relates to the technical field of full-color modules, and discloses a full-color module dead pixel dynamic replacement method and an LED display screen. The method comprises the following steps: in a blanking period triggered by a frame synchronization signal, PWM read-back instructions are sequentially sent to each pixel in a full-color module, and the first output duty cycles of R, G and B channels of each pixel are returned back; dead pixel type analysis is performed on each pixel to obtain a dead pixel mapping table; boundary checking and health screening are performed on the surrounding pixels of each dead pixel coordinate in the dead pixel mapping table to obtain a healthy neighborhood pixel set, and a neighborhood weight list is calculated; a brightness compensation gain is calculated, a compensation replacement value is injected into a pixel output register of the dead pixel coordinate, and dead pixel dynamic replacement is completed, thereby effectively compensating for the local area perceived brightness drop caused by the non-light emission of the dead pixel, guaranteeing the visual consistency of the area where the dead pixel is located and the surrounding picture brightness, and realizing the dead pixel dynamic replacement of the full-color module without stopping the machine and replacing physical devices.

Description

Technical Field

[0001] This application relates to the field of full-color module technology, and in particular to a method for dynamic replacement of defective pixels in a full-color module and an LED display screen. Background Technology

[0002] Full-color LED display modules consist of red, green, and blue sub-pixel arrays. However, as the module operates for longer periods, problems such as LED chip aging, loose solder joints, and partial failures of the driver IC inevitably lead to the continuous generation of dead pixels. The appearance of dead pixels directly affects the display quality and pixel uniformity.

[0003] In existing technologies, defective pixel handling in full-color modules relies on static calibration, where the coordinates of defective pixels are permanently written into a static defective pixel table in the driver controller. However, newly added defective pixels during actual operation of the full-color module cannot be detected, and the static defective pixel table is inconsistent with the actual state of the module for a long time. This results in a large number of defective pixels generated during operation being missed and unable to be processed in a timely manner. Existing technologies use fixed grayscale values ​​to fill the defective pixel positions, and the replacement values ​​do not change with the content of the current frame. This causes a fixed color difference and brightness discontinuity between the defective pixel area and the surrounding normal pixels. The visual defects are particularly prominent when playing dynamic video content, seriously affecting the overall display effect of the full-color module. Summary of the Invention

[0004] This application provides a method for dynamic replacement of defective pixels in a full-color module and an LED display screen. This application overcomes the passive detection method of traditional solutions that rely on manual inspection or static calibration at the factory. It realizes the frame-by-frame online active acquisition of the driving state of all pixels. The compensation replacement value changes in real time with the content of the current frame, rather than being filled with fixed grayscale, thus eliminating the color difference and layering between the replacement area and the surrounding screen.

[0005] The first aspect of this application provides a method for dynamic replacement of defective pixels in a full-color module, the method comprising: During the blanking period triggered by the frame synchronization signal, PWM readback commands are sent to each pixel in the full-color module one by one, and the first output duty cycle of the R, G, and B channels of each pixel is returned to generate the measured duty cycle matrix. Based on the measured duty cycle matrix, perform bad pixel type analysis on each pixel to obtain a bad pixel mapping table; Boundary verification and health screening are performed on the surrounding pixels of each bad pixel coordinate in the bad pixel mapping table to obtain a healthy neighbor pixel set, and a neighborhood weight list between the healthy neighbor pixel set and the bad pixel coordinate is calculated. The brightness compensation gain of the healthy neighborhood pixel set is calculated based on the neighborhood weight list. The compensation replacement value is injected into the pixel output register of the bad pixel coordinates. At the same time, the brightness compensation gain is applied to complete the dynamic replacement of bad pixels.

[0006] In conjunction with the first aspect, in the first implementation of the first aspect of the present invention, the step of sending PWM readback commands to each pixel in the full-color module sequentially during the blanking period triggered by the frame synchronization signal, and transmitting back the first output duty cycle of the R, G, and B channels of each pixel to generate a measured duty cycle matrix includes: The falling edge transition of the monitoring frame synchronization signal is activated at the falling edge moment, and the duration of the reading acquisition window is constrained within the blanking period. PWM reading instructions are sent to each pixel in the full-color module to obtain the first output duty cycle of the R, G, and B channels of each pixel. The first output duty cycle of each pixel's R, G, and B channels is transmitted back to the controller via the SPI serial bus and written into the on-chip memory area in pixel coordinate order to generate the measured duty cycle matrix.

[0007] In conjunction with the first aspect, in a second implementation of the first aspect of the present invention, the step of performing bad pixel type analysis on each pixel based on the measured duty cycle matrix to obtain a bad pixel mapping table includes: The R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel are calculated based on the measured duty cycle matrix. Based on the R channel deviation rate, the G channel deviation rate, and the B channel deviation rate, a bad pixel type analysis is performed on each pixel to obtain a bad pixel mapping table.

[0008] In conjunction with the first aspect, in a third implementation of the first aspect of the present invention, the step of calculating the R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel based on the measured duty cycle matrix includes: Read the measured duty cycle matrix and the duty cycle of the same pixel coordinate in the reference duty cycle lookup table in the row and column scanning order. Divide the measured duty cycle of the R channel of each pixel with the corresponding reference duty cycle of the R channel in the reference duty cycle lookup table and then divide by the reference duty cycle of the R channel to obtain the R channel deviation rate of each pixel. The G-channel deviation rate of each pixel is obtained by subtracting the measured duty cycle of the G-channel from the corresponding G-channel reference duty cycle in the reference duty cycle lookup table and then dividing by the G-channel reference duty cycle. The B-channel deviation rate of each pixel is obtained by subtracting the measured duty cycle of the B-channel from the corresponding B-channel reference duty cycle in the reference duty cycle lookup table and then dividing by the reference duty cycle of the B-channel.

[0009] In conjunction with the first aspect, in a fourth implementation of the first aspect of the present invention, the step of performing bad pixel type analysis on each pixel based on the R channel deviation rate, the G channel deviation rate, and the B channel deviation rate to obtain a bad pixel mapping table includes: The R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel are compared with the brightness deviation threshold channel by channel. When the deviation rate of any channel is not lower than the brightness deviation threshold, the corresponding pixel is determined to be a brightness defect. The deviation range value is calculated based on the R channel deviation rate, the G channel deviation rate, and the B channel deviation rate, and the deviation range value is compared with the chromaticity range threshold. When the deviation range value is not lower than the chromaticity range threshold, the bad pixel type of the corresponding pixel is determined to be a chromaticity offset bad pixel. When there is a pixel whose bad pixel type satisfies both luminance bad pixel and chromaticity offset bad pixel, the bad pixel type of the corresponding pixel is determined to be a mixed bad pixel. Write the defect type of each pixel along with the corresponding defect coordinates into the defect mapping table.

[0010] In conjunction with the first aspect, in the fifth implementation of the first aspect of the present invention, the step of performing boundary verification and health screening on the surrounding pixels of each bad pixel coordinate in the bad pixel mapping table to obtain a healthy neighboring pixel set, and calculating a neighborhood weight list between the healthy neighboring pixel set and the bad pixel coordinates, includes: Based on the coordinates of each bad point in the bad point mapping table, an 8-neighbor candidate pixel set is constructed, and boundary verification and health screening are performed on each candidate coordinate in the 8-neighbor candidate pixel set in sequence. The candidate coordinates that pass the boundary verification and health screening are added to the candidate neighbor pixel set. Count the number of valid pixels in the candidate neighbor pixel set and determine the healthy neighbor pixel set; Calculate the Euclidean distance between the target coordinates of each healthy neighbor pixel in the healthy neighbor pixel set and the coordinates of the bad pixel, and take the reciprocal of the Euclidean distance as the reciprocal distance weight of the corresponding healthy neighbor pixel; A neighborhood weight list is generated based on the target coordinates and the corresponding inverse distance weights.

[0011] In conjunction with the first aspect, in a sixth implementation of the first aspect of the present invention, the step of counting the number of valid pixels within the candidate neighboring pixel set and determining the healthy neighboring pixel set includes: The number of valid pixels in the candidate neighborhood pixel set is counted, and the number of valid pixels is compared with a threshold. When the number of effective pixels is not lower than the number threshold, the candidate neighborhood pixel set is directly determined as the healthy neighborhood pixel set; When the number of effective pixels is lower than the number threshold, the neighborhood range is expanded from 8 neighborhoods to 16 neighborhoods with the bad pixel coordinates as the center, and new candidate coordinates are extracted. Boundary verification and health screening are performed on the new candidate coordinates in sequence. The new candidate coordinates that pass the boundary verification and health screening are added to the candidate neighborhood pixel set to obtain the healthy neighborhood pixel set.

[0012] In conjunction with the first aspect, in the seventh implementation of the first aspect of the present invention, the step of calculating the brightness compensation gain of the healthy neighborhood pixel set based on the neighborhood weight list, injecting the compensation replacement value into the pixel output register of the bad pixel coordinates, and simultaneously applying the brightness compensation gain to complete the dynamic replacement of bad pixels includes: The compensation replacement value for the bad pixel coordinates is calculated based on the second output duty cycle of the R, G, and B channels of each healthy neighbor pixel in the neighborhood weight list and the measured duty cycle matrix, and the compensation replacement value is written into the pixel output register of the bad pixel coordinates. Calculate the average neighborhood distance from all healthy neighbor pixels in the healthy neighbor pixel set to the coordinates of the bad pixels, and calculate the brightness compensation gain for each healthy neighbor pixel based on the average neighborhood distance. The second output duty cycle of the R, G, and B channels of each healthy neighboring pixel is multiplied by the brightness compensation gain to obtain the target value. The minimum value of the target value and the upper limit of the output duty cycle is written back to the pixel output register of each healthy neighboring pixel to complete the dynamic replacement of bad pixels.

[0013] In conjunction with the first aspect, in the eighth implementation of the first aspect of the present invention, the step of calculating the average neighborhood distance from all healthy neighboring pixels in the healthy neighboring pixel set to the coordinates of the bad pixels, and calculating the brightness compensation gain of each healthy neighboring pixel based on the average neighborhood distance, includes: The average neighborhood distance is obtained by summing the Euclidean distances between each healthy neighbor pixel in the healthy neighbor pixel set and the coordinates of the bad pixel, and then dividing by the number of effective pixels. Based on the mean neighborhood distance and the Euclidean distance from each healthy neighboring pixel to the coordinates of the bad pixel, a brightness compensation coefficient inversely proportional to the distance is constructed, and a unit base value is superimposed on the brightness compensation coefficient to obtain the brightness compensation gain of each healthy neighboring pixel.

[0014] A second aspect of this application provides an LED display screen for performing a dynamic replacement method for defective pixels in a full-color module.

[0015] Compared with existing technologies, this application achieves frame-by-frame online active acquisition of the driving state of all pixels by performing pixel-by-pixel PWM readback during the blanking period triggered by the frame synchronization signal, actively transmitting the first output duty cycle of the R, G, and B channels of each pixel back to the controller, and generating a measured duty cycle matrix. This overcomes the passive detection method of traditional solutions that rely on manual inspection or factory static calibration, and realizes frame-by-frame online active acquisition of the driving state of all pixels. By calculating the pixel-by-pixel and channel-by-channel deviation rate using the measured duty cycle matrix and the pre-stored reference duty cycle lookup table, and classifying each pixel into luminance defective pixels, chromaticity offset defective pixels, and mixed defective pixels based on the luminance deviation threshold and chromaticity range threshold, the defective pixel mapping table is fully refreshed in each frame, realizing frame-by-frame online dynamic detection of defective pixels and continuous updating of the defective pixel table, completely solving the core defect of traditional static defective pixel tables that cannot detect newly added defective pixels during operation. In the defective pixel replacement process, a set of healthy neighboring pixels is constructed based on the coordinates of each defective pixel, and the inverse distance weight is calculated. The duty cycle of the three channels of the healthy neighboring pixels is then interpolated using an inverse distance weighting method. The resulting compensation replacement value changes in real time with the content of the current frame, rather than being a fixed grayscale fill, thus eliminating color difference banding between the replacement area and the surrounding image. Simultaneously, by applying a brightness compensation gain inversely proportional to the distance to the healthy neighboring pixels, the perceived brightness decrease in localized areas caused by the defective pixel not emitting light is effectively compensated, ensuring visual consistency between the defective pixel area and the surrounding image brightness. This achieves dynamic defective pixel replacement of the full-color module without stopping the system or replacing physical components. Attached Figure Description

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

[0017] The structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0018] Figure 1 This is a flowchart illustrating the dynamic replacement method for defective pixels in a full-color module provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the process of sending PWM readback commands to each pixel in the full-color module one by one in an embodiment of the present invention; Figure 3This is a schematic diagram illustrating the process of performing defect type analysis on each pixel in an embodiment of the present invention; Figure 4 This is a schematic diagram of the boundary verification and health screening process in an embodiment of the present invention; Figure 5 This is a schematic diagram of the process for dynamic replacement of bad pixels in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0021] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0022] It should also be further understood that the term "and / or" as used in this application specification and the appended claims refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes such combinations. See also Figure 1 One embodiment of the dynamic replacement method for defective pixels in a full-color module in this application includes: Step 100: During the blanking period triggered by the frame synchronization signal, send PWM readback commands to each pixel in the full-color module one by one, and send back the first output duty cycle of the R, G, and B channels of each pixel to generate the measured duty cycle matrix. Step 200: Based on the measured duty cycle matrix, perform bad pixel type analysis on each pixel to obtain a bad pixel mapping table; Step 300: Perform boundary verification and health screening on the surrounding pixels of each bad pixel coordinate in the bad pixel mapping table to obtain a set of healthy neighboring pixels, and calculate the neighborhood weight list between the set of healthy neighboring pixels and the bad pixel coordinates. Step 400: Calculate the brightness compensation gain of the healthy neighborhood pixel set based on the neighborhood weight list, inject the compensation replacement value into the pixel output register of the bad pixel coordinates, and apply the brightness compensation gain to complete the dynamic replacement of bad pixels.

[0023] In one specific embodiment, such as Figure 2 The process of executing step 100 may specifically include the following steps: Step 101: Monitor the falling edge transition of the frame synchronization signal, open the readback acquisition window at the falling edge moment, and constrain the duration of the readback acquisition window within the blanking period. Send PWM readback commands to each pixel in the full-color module to obtain the first output duty cycle of the R, G, and B channels of each pixel. Step 102: The first output duty cycle of each pixel's R, G, and B channels is transmitted back to the controller via the SPI serial bus and written into the on-chip memory area in pixel coordinate order to generate the measured duty cycle matrix.

[0024] Specifically, the FPGA main control logic continuously locks the edge state of the frame synchronization signal VSYNC, and registers the current moment as the start of readback acquisition at the instant VSYNC experiences a falling edge transition. At the same time, it opens a PWM readback acquisition window that is only for the current frame, and the duration of the readback acquisition window meets the following requirements. ,in, Indicates the duration of the readback acquisition window. This indicates the duration of the blanking period. Timing constraints ensure that the readback action falls entirely within the blanking period, avoiding screen flickering, grayscale disturbances, or line / field misalignment caused by occupying effective display time. For full-color modules with high resolution and a large number of pixels, the module is divided along the line direction into... Each row group is configured with an independent SPI serial bus, allowing multiple pixel acquisitions to be performed in parallel within the same readback window. The number of row groups meets the requirements. ,in, This indicates the total number of pixels in the module. This represents the total number of clock cycles required for a single pixel to complete a three-channel readback, with a value of 48 clock cycles. This value of 48 corresponds to the complete transmission overhead of the instruction frame header, row and column addresses, three-channel 10-bit data, and frame trailer. This indicates the SPI clock frequency, set to 25MHz. This 25MHz setting balances the stability of the driver interface with the data throughput within the blanking window. After entering the parallel acquisition phase, each SPI serial bus sends PWM readback commands to the pixels in its row group one by one in ascending column address order. Upon receiving the specified pixel coordinates, the driver IC encodes and transmits the actual output PWM duty cycle of the R, G, and B channels in a fixed channel order. Each channel is represented by a 10-bit binary value, with a duty cycle resolution of [missing value]. This resolution allows for the preservation of finer luminance quantization levels without significantly increasing the bus load. After receiving the three-channel readback results for each pixel, the controller writes them sequentially to the pre-allocated storage area of ​​the on-chip SRAM according to pixel coordinates, and organizes the first output duty cycles of the R, G, and B channels corresponding to the same pixel into triplets. ,in, , , They represent the first Line 1 The duty cycle of the first output of the three channels of each pixel. As all pixels complete sequential writing, the on-chip memory area forms a measured duty cycle matrix that corresponds one-to-one with the row and column coordinates of the module. In the matrix That is, the actual driving state of the corresponding coordinate pixel is read back during the blanking period of the current frame.

[0025] In one specific embodiment, such as Figure 3 The process of executing step 200 may specifically include the following steps: Step 201: Calculate the R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel based on the measured duty cycle matrix; Step 202: Analyze the bad pixel type of each pixel based on the R channel deviation rate, G channel deviation rate and B channel deviation rate to obtain a bad pixel mapping table.

[0026] Specifically, the controller is measured in the duty cycle matrix F map After the acquisition end flag is generated and set, the system switches to the reference comparison processing state and synchronously reads the measured duty cycle matrix F according to the fixed scanning order of pixel coordinates. map Baseline Duty Cycle Lookup Table (LUT) mapThe three-channel data at the same coordinate position establishes a one-to-one correspondence between the measured output of the current frame and the stable output of the reference for each pixel's R, G, and B channels. The degree of change of the current measured duty cycle of the three channels relative to the corresponding reference duty cycle is converted into a uniformly comparable deviation rate, thereby eliminating the direct comparison error caused by the inconsistency of the original brightness levels of different pixels and different channels. When a pixel experiences chip aging, unstable solder joint contact, or local driving abnormality, one or more of the three channels will continuously deviate from the reference value. This deviation is solidified into three judgment inputs: R channel deviation rate, G channel deviation rate, and B channel deviation rate, and the three results are temporarily stored in the pixel-by-pixel comparison buffer. The controller performs a bad pixel type analysis based on the three-channel deviation rate of each pixel. It screens for brightness anomalies in each of the three channels. If the absolute value of the deviation in any channel reaches the brightness bad pixel judgment boundary of 15%, the corresponding pixel is considered to have exceeded the normal operating fluctuation range. The controller also compares the consistency between the three-channel deviations. Because some pixels do not darken overall, but rather a single monochrome sub-pixel shows significant decay while the other two channels remain relatively normal, the visual result will appear as a significant color cast. Therefore, chromaticity shift identification is performed based on the dispersion of the three-channel deviations. When the deviation range reaches the chromaticity range judgment boundary of 20%, the pixel is marked as a chromaticity shift bad pixel. For pixels that simultaneously meet both brightness and chromaticity anomalies, the controller merges the bad pixel types and marks them as mixed bad pixels; positions that do not trigger either condition remain marked as healthy pixels. After all pixels have completed this round of coordinate-by-coordinate analysis, each pixel type flag and its corresponding coordinates are written into the bad pixel mapping table. Map It refreshes the current frame's bad pixel status in a full-coverage manner, enabling neighborhood filtering and dynamic replacement to be processed directly based on the latest bad pixel spatial distribution.

[0027] In one specific embodiment, the process of performing step 201 may specifically include the following steps: Read the measured duty cycle matrix and the duty cycle of the same pixel coordinate in the reference duty cycle lookup table in the row and column scanning order. Divide the measured duty cycle of the R channel of each pixel with the corresponding reference duty cycle of the R channel in the reference duty cycle lookup table and then divide by the reference duty cycle of the R channel to obtain the R channel deviation rate of each pixel. The G-channel deviation rate of each pixel is obtained by subtracting the measured duty cycle of the G-channel from the reference duty cycle in the lookup table and dividing by the reference duty cycle of the G-channel. The deviation rate of the B channel for each pixel is obtained by subtracting the measured duty cycle of the B channel from the corresponding reference duty cycle in the reference duty cycle lookup table and then dividing by the reference duty cycle of the B channel.

[0028] Specifically, the controller in the measured duty cycle matrix After writing is complete, the system enters a pixel-by-pixel reference comparison state. The address generation unit then issues coordinate access requests sequentially in row-major and column-incrementing order, so that the measured duty cycle matrix matches the reference duty cycle lookup table. Output the same pixel coordinates under the same clock beat The corresponding three-channel data forms co-located paired R-channel, G-channel, and B-channel data. A reference duty cycle lookup table can be pre-loaded into the on-chip Flash after module calibration. The reference data originates from the thermal stability readback results under full-white frame excitation, where the full-scale duty cycle is 100%, and the continuous stable emission time is no less than 30 minutes. Full-scale excitation fully exposes the three-channel output capabilities, and 30 minutes of continuous emission helps the LED junction temperature enter a stable range, reducing the reference deviation caused by thermal drift during the cold start phase. Therefore, the deviation rate obtained by frame-by-frame comparison is closer to the actual degradation state of the module. After completing the co-located readout, the calculation unit performs normalization processing on each pixel's three channels by subtracting the reference value from the measured value and then dividing by the reference value, uniformly converting the original brightness level differences of different pixels into directly comparable relative changes. The R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate are expressed as follows:

[0029]

[0030]

[0031] in, , , They represent the first Line 1 The measured duty cycles of the R, G, and B channels of the column pixels, obtained from the current frame readback, are expressed in % (%). , , These represent the three-channel reference duty cycles of the same coordinate in the reference duty cycle lookup table, in percentages (%). , , The value is a dimensionless signed number; a value greater than 0 indicates that the current measured output is higher than the baseline, and a value less than 0 indicates that the current measured output is lower than the baseline. To ensure that this calculation chain meets the requirements of frame-by-frame online detection, the three-channel subtraction operation, division operation, and result write-back can be organized into a pipelined processing structure. That is, while reading the three-channel data of the current pixel, the data that was subtracted in the previous clock cycle is divided, and the data that was divided in the clock cycle before that is written to the deviation rate buffer, thus making the scanning reading, deviation calculation, and result buffering continuous. After the three-channel deviation rate calculation is completed for each coordinate, the controller writes the three results into the deviation rate register area or the on-chip RAM temporary area in the original pixel coordinate order, forming a deviation rate data surface with the same dimension as the measured duty cycle matrix.

[0032] Before sequentially reading the duty cycle of the measured duty cycle matrix and the reference duty cycle lookup table for the same pixel coordinate in the row and column scanning order, the process includes generating the reference duty cycle lookup table: During the module factory calibration stage, a full-white frame excitation is applied to the full-color module, causing the R, G, and B channels to be continuously driven simultaneously at their rated full-scale duty cycles. The frame-by-frame readback values ​​of the output duty cycle of each pixel's three channels are monitored in real time, and the absolute value of the difference between adjacent frame readback values ​​is calculated. When the readback difference between adjacent frames of all pixels' three channels is lower than the stability threshold, the module is deemed to have reached a thermally stable state. After reaching thermal stability, a complete pixel-by-pixel PWM readback is performed, and the output duty cycles of the R, G, and B channels of each pixel are recorded as the R-channel reference duty cycle, G-channel reference duty cycle, and B-channel reference duty cycle, respectively. The R-channel reference duty cycle, G-channel reference duty cycle, and B-channel reference duty cycle of each pixel, along with the corresponding pixel coordinates, are written into the on-chip Flash in row and column order to generate a reference duty cycle lookup table. The reference duty cycle lookup table serves as a fixed reference for defect detection throughout the entire life cycle of the module and is only overwritten and updated during module recalibration.

[0033] In one specific embodiment, the process of performing step 202 may specifically include the following steps: The R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel are compared with the brightness deviation threshold channel by channel. When the deviation rate of any channel is not lower than the brightness deviation threshold, the corresponding pixel is determined to be a brightness defect. The deviation range value is calculated based on the deviation rate of the R channel, the deviation rate of the G channel, and the deviation rate of the B channel, and then compared with the chromaticity range threshold. When the deviation range value is not lower than the chromaticity range threshold, the bad pixel type of the corresponding pixel is determined to be a chromaticity offset bad pixel. When there is a pixel whose bad pixel type satisfies both luminance bad pixel and chromaticity offset bad pixel, the bad pixel type of the corresponding pixel is determined to be a mixed bad pixel. Write the defect type of each pixel along with the corresponding defect coordinates into the defect mapping table.

[0034] Specifically, after the three-channel deviation rate buffer is formed, the controller starts the pixel-by-pixel determination pipeline and continuously maintains the same row and column scanning order as the previous calculation stage, so that the first... Line 1 The R-channel, G-channel, and B-channel deviation rates of the column pixels are synchronously sent to the comparison unit within the same judgment period. For brightness anomaly identification, the comparison unit compares the absolute values ​​of the three channel deviation rates channel by channel with a brightness deviation threshold of 15%. The 15% threshold is used to distinguish between normal aging fluctuations and abnormal intervals requiring compensation. Slight brightness fluctuations fall within the warning range, while reaching 15% exceeds the acceptable attenuation boundary in daily operation, easily creating a visually identifiable difference in brightness. Therefore, as long as any channel meets the threshold... The current location is marked as a luminance defect, and the defect type flag is preset to 0x01. After completing the luminance anomaly identification, the controller continues to perform chrominance anomaly identification on the same pixel, calculating the deviation range value of the three-channel deviation rate:

[0035] in, Indicates the first Line 1 The value indicates the dispersion of the three-channel deviation of the column pixels. The larger the value, the worse the attenuation consistency of the three channels and the more obvious the pixel color cast. After comparison with the chromaticity range threshold of 20%, it meets the requirements. When a pixel is identified as a chromaticity shift defective pixel, the controller determines that the current position is a defective pixel and sets the defective pixel type flag to 0x02. A 20% value corresponds to the engineering boundary where monochrome sub-pixel imbalance can cause significant color shift. The determination logic merges the two types of results. If the same pixel simultaneously meets both the luminance defective pixel condition and the chromaticity shift defective pixel condition, the controller corrects the defective pixel type to a mixed defective pixel and writes the flag to 0x03. If neither condition is triggered, the current position is maintained as a healthy pixel, and the flag is written to 0x00. After all pixels have been determined, the pixel coordinates are written to the control unit according to the original coordinate order. Write the corresponding bad pixel type flag into the bad pixel mapping table Bad. Map The write buffer is filled, and after all coordinate determinations for the current frame are completed, the entire frame result is overwritten to the read storage area, thus causing Bad... Map Maintain an M-row, N-column two-dimensional matrix structure, with matrix elements taking values ​​only within the range of {0x00, 0x01, 0x02, 0x03}.

[0036] The process of writing the defect type of each pixel along with the corresponding defect coordinates into the defect mapping table also includes: within each VSYNC trigger cycle, writing the defect type determination results of all pixels in the current frame into the write buffer of the defect mapping table in coordinate order; after all pixel determination results have been written, overwriting the contents of the write buffer into the read storage area of ​​the defect mapping table to complete the full overwrite refresh of the defect mapping table for the current frame; after the full overwrite refresh, comparing the number of non-zero coordinates of defect types in the current frame's defect mapping table with the number of defective pixels in the previous frame; when the number of newly added defect coordinates in the current frame exceeds the threshold for the number of newly added defective pixels, writing a defect number surge alarm flag and the corresponding frame number to the system log register; during the pixel data stream output stage of the next frame, the defect replacement module directly reads the refreshed defect mapping table in the storage area to perform the replacement operation, ensuring that each frame replacement operation is based on the latest defect distribution state of the current frame and does not reference any historical frame defect mapping results.

[0037] In one specific embodiment, such as Figure 4 The process of executing step 300 may specifically include the following steps: Step 301: Based on the coordinates of each bad point in the bad point mapping table, construct an 8-neighborhood candidate pixel set, and perform boundary verification and health screening on each candidate coordinate in the 8-neighborhood candidate pixel set in turn. Add the candidate coordinates that pass the boundary verification and health screening to the candidate neighborhood pixel set. Step 302: Count the number of valid pixels in the candidate neighboring pixel set and determine the healthy neighboring pixel set; Step 303: Calculate the Euclidean distance between the target coordinates and the bad point coordinates of each healthy neighbor pixel in the healthy neighbor pixel set, and take the reciprocal of the Euclidean distance as the reciprocal distance weight of the corresponding healthy neighbor pixel. Step 304: Generate a neighborhood weight list based on the target coordinates and the corresponding inverse distance weights.

[0038] Specifically, the controller maps the bad pixels to the bad pixel mapping table. Map After the refresh is complete, the coordinates of bad pixels with non-zero bad pixel type flags are detected one by one in the row and column scanning order, and the current bad pixel coordinates are used. A processing link is constructed around the center to enter the neighborhood. An 8-neighborhood candidate pixel set is established around the coordinates of the defective pixel. This yields eight candidate locations surrounding the distribution of defective pixel coordinates. For each candidate coordinate, boundary checks and health screening are performed sequentially. Boundary checks determine whether the candidate coordinates still fall within the valid row and column range of the module. For example, using... and As a pass condition, health screening is used to determine whether the type flag of the candidate position in the bad pixel mapping table is 0x00. Only candidate coordinates that simultaneously meet the conditions of valid boundary and healthy pixel are incorporated into the candidate neighbor pixel set, thereby avoiding the erroneous introduction of out-of-bounds coordinates, brightness bad pixel coordinates, chroma offset bad pixel coordinates, or mixed bad pixel coordinates into the interpolation data source. The controller counts the number of valid pixels in the candidate neighbor pixel set and compares the statistical result with a number threshold. When the number of valid pixels is not lower than the number threshold, the candidate neighbor pixel set can be directly determined as a healthy neighbor pixel set; when the number of valid pixels is lower than the number threshold, the controller then enters the external expansion and supplementation process to ensure that the healthy neighbor pixel set can still provide sufficient input for the compensation calculation. After the healthy neighbor pixel set is confirmed, the target coordinates of each healthy neighbor pixel in the set are... coordinates of bad pixels Calculate the Euclidean distance using the following relationship: ,in, This represents the spatial distance between healthy neighboring pixels and bad pixels. Within an 8-neighborhood, only the orthogonal distance of 1 and the diagonal distance typically occur. Two types of values ​​are used, measured by Euclidean distance, which can simultaneously reflect the true spatial proximity in the up, down, left, right, and diagonal directions. Following the principle that the closer the distance, the greater the contribution to compensation, the reciprocal of the Euclidean distance is defined as the reciprocal distance weight, and the calculation relationship is as follows:

[0039] This results in healthy pixels adjacent to bad pixels receiving higher weights in the interpolation link, while the contribution of relatively distant healthy pixels is automatically reduced. The controller writes the target coordinates, corresponding Euclidean distance, and corresponding inverse distance weight of each selected healthy neighboring pixel into the neighborhood weight list in coordinate order. .

[0040] In one specific embodiment, the process of performing step 302 may specifically include the following steps: The number of valid pixels within the candidate neighborhood pixel set is counted, and the number of valid pixels is compared with the number threshold. When the number of effective pixels is not less than the number threshold, the candidate neighborhood pixel set is directly determined as the healthy neighborhood pixel set; When the number of effective pixels is lower than the number threshold, the neighborhood range is expanded from 8 neighborhoods to 16 neighborhoods with the bad pixel coordinates as the center, and new candidate coordinates are extracted. Boundary verification and health screening are performed on the new candidate coordinates in sequence. The new candidate coordinates that pass the boundary verification and health screening are added to the candidate neighborhood pixel set to obtain the healthy neighborhood pixel set.

[0041] Specifically, the system accumulates all candidate coordinates that have passed boundary verification and health screening within the set to obtain the number of valid pixels corresponding to the current bad pixel coordinates. This number of valid pixels is then compared to a threshold. A threshold of 4 is recommended because inverse distance weighted interpolation requires basic coverage in the four directions (up, down, left, and right) in planar space to maintain good directional balance in the compensation result. When the number of samples is less than 4, the pixel distribution involved in the calculation tends to be biased towards one side or local corners, easily causing significant bias in the interpolation result and weakening the spatial continuity after bad pixel replacement. Based on this principle, when the number of valid pixels is not less than 4, the controller can directly determine the current candidate neighbor pixel set as the healthy neighbor pixel set and write the set size, coordinates within the set, and corresponding bad pixel coordinates into the working register for direct use by the distance calculation and weight generation modules.

[0042] When the number of valid pixels falls below a threshold, the controller performs a neighborhood expansion centered on the current bad pixel coordinates, extending the original 8-neighborhood search range to 16 neighborhoods, and extracting new candidate coordinates from the expanded spatial range. The expanded candidate set can be represented as follows: This corresponds to a total of 24 candidate positions. Because when a defective pixel is near the module edge or corner, or when multiple defective pixels already exist nearby, relying solely on a ring of adjacent pixels is insufficient to form a stable interpolation input. Expanding the radius to two pixels increases the probability of obtaining usable healthy pixels without making the sampling range too large and weakening the targeted nature of local compensation. After the new candidate coordinates are extracted, the controller uses the same processing rules as the initial screening, performing boundary checks and health checks on each new coordinate sequentially, i.e., first determining whether the new coordinate falls within the valid row and column range. and Within this range, it is then determined whether the corresponding position is in a healthy state according to the bad pixel mapping table. Only newly added coordinates that meet both conditions are added to the original candidate neighbor pixel set, thus forming an expanded healthy neighbor pixel set.

[0043] After the above supplementation, the controller obtains a new set of available neighboring pixels for the current defective pixel. For most edge defects or locally sparse defects, this expansion process can increase the number of effective samples to a usable range, ensuring that defective pixel replacement remains local and continuous. However, if the number of effective pixels is still less than 4 after expanding to 16 neighbors, the current defective pixel can be further identified as an isolated dense defective pixel, and the interpolation compensation link can be stopped to avoid generating distorted replacement results when healthy samples are significantly insufficient.

[0044] In one specific embodiment, such as Figure 5 The process of executing step 400 may specifically include the following steps: Step 401: Calculate the compensation replacement value of the bad pixel coordinates based on the second output duty cycle of the R, G, and B channels of each healthy neighboring pixel in the neighborhood weight list and the measured duty cycle matrix, and write the compensation replacement value into the pixel output register of the bad pixel coordinates. Step 402: Calculate the average neighborhood distance from all healthy neighbor pixels in the healthy neighbor pixel set to the coordinates of the bad pixels, and calculate the brightness compensation gain for each healthy neighbor pixel based on the average neighborhood distance. Step 403: Multiply the second output duty cycle of the R, G, and B channels of each healthy neighboring pixel by the brightness compensation gain to obtain the target value. Write the minimum value of the target value and the upper limit of the output duty cycle back to the pixel output register of each healthy neighboring pixel to complete the dynamic replacement of bad pixels.

[0045] Specifically, the processing link is deployed inside the FPGA bad pixel replacement module in the current frame pixel data stream output stage, and the coordinate matching unit continuously monitors whether the current output pixel coordinates are consistent with the bad pixel coordinates in the bad pixel mapping table; when the first bad pixel is detected... Line 1 When the column output coordinates match the bad pixel coordinates, the bad pixel replacement module calls the neighborhood weight list that was generated and cached in the working register in the previous stage. Meanwhile, from the measured duty cycle matrix Read each healthy neighbor pixel in the neighborhood weight list sequentially. The corresponding R-channel second output duty cycle, G-channel second output duty cycle, and B-channel second output duty cycle are used as interpolation source data for compensation calculations, based on the current frame measured outputs of these healthy neighboring pixels. Distance-inverse weighted summation is performed on the R, G, and B channels respectively, and normalization is completed within each channel. The R-channel, G-channel, and B-channel compensation replacement values ​​for the bad pixel coordinates are as follows: ; ; ; in, , , These represent the coordinates of the bad pixels. Compensation replacement values ​​on the three channels, expressed as a duty cycle percentage; , , Representing healthy neighboring pixels respectively The duty cycle of the second output of the three channels in the current frame; This represents the inverse distance weight of the corresponding healthy neighboring pixels. After obtaining the three-channel compensation replacement values, the controller combines them into a triplet of compensation replacement values ​​for the bad pixel coordinates and directly writes it to the pixel output register corresponding to the bad pixel coordinates, overwriting the abnormal output at the original bad pixel position, thereby completing the real-time replacement of the bad pixel itself.

[0046] After replacing the defective pixel, local brightness compensation parameters are prepared around the set of healthy neighboring pixels. The controller first summarizes the distances from all healthy neighboring pixels to the coordinates of the defective pixel and calculates the average neighborhood distance. Then, based on the relationship between the average neighborhood distance and the distance from each healthy neighboring pixel to the defective pixel, the brightness compensation gain of each healthy neighboring pixel is calculated and temporarily stored in the gain register of the corresponding neighboring pixel. By compensating for the local perceived brightness gap caused by the extinguishing or abnormal attenuation of the defective pixel, the healthy pixels around the defective pixel appropriately bear the visual energy compensation. The healthy neighboring pixels closer to the defective pixel bear a relatively higher compensation magnitude, while the healthy neighboring pixels further away bear a relatively lower compensation magnitude, thereby maintaining the continuity of the local transition. The controller applies the calculated brightness compensation gain to the second output duty cycle of the R, G, and B channels of each healthy neighboring pixel (m,n) within the healthy neighboring pixel set. The current output value of each channel is multiplied by the corresponding gain channel by channel to obtain the compensated target value. The target value is then compared with the upper limit of the output duty cycle, and the smaller value is written back to the pixel output register of the corresponding healthy neighboring pixel. The upper limit of the output duty cycle can be 100%, which corresponds to the full-scale duty cycle boundary of the driver IC. Using this upper limit can prevent the compensated value from exceeding the allowable range of the driver, thus preventing new over-brightness distortion or register overflow.

[0047] The compensation replacement value for bad pixel coordinates is calculated based on the second output duty cycles of the R, G, and B channels of each healthy neighboring pixel in the neighborhood weight list and the measured duty cycle matrix. This includes: sequentially reading the second output duty cycle of the R channel corresponding to the coordinates of each healthy neighboring pixel in the neighborhood weight list from the measured duty cycle matrix; multiplying the second output duty cycle of the R channel of each healthy neighboring pixel by its corresponding inverse distance weight and summing them one by one to obtain the weighted accumulated value of the R channel; dividing the weighted accumulated value of the R channel by the sum of all inverse distance weights in the neighborhood weight list to obtain the compensation replacement value of the R channel; sequentially reading the second output duty cycle of the G channel corresponding to the coordinates of each healthy neighboring pixel in the neighborhood weight list from the measured duty cycle matrix; multiplying the second output duty cycle of the R channel of each healthy neighboring pixel by its corresponding inverse distance weight and summing them one by one to obtain the weighted accumulated value of the R channel; and dividing the weighted accumulated value of the R channel by the sum of all inverse distance weights in the neighborhood weight list to obtain the compensation replacement value of the R channel; and sequentially reading the second output duty cycle of the G channel of each healthy neighboring pixel in the neighborhood weight list from the measured duty cycle matrix... The duty cycle is multiplied by the corresponding inverse distance weight and then summed one by one to obtain the weighted cumulative value of the G channel. The weighted cumulative value of the G channel is divided by the sum of all inverse distance weights in the neighborhood weight list to obtain the compensation replacement value of the G channel. The second output duty cycle of the B channel corresponding to the coordinates of each healthy neighbor pixel in the neighborhood weight list is read sequentially from the measured duty cycle matrix. The second output duty cycle of the B channel of each healthy neighbor pixel is multiplied by the corresponding inverse distance weight and then summed one by one to obtain the weighted cumulative value of the B channel. The weighted cumulative value of the B channel is divided by the sum of all inverse distance weights in the neighborhood weight list to obtain the compensation replacement value of the B channel. The compensation replacement value of the R channel, the compensation replacement value of the G channel, and the compensation replacement value of the B channel are combined to obtain the compensation replacement value of the bad pixel coordinates.

[0048] In one specific embodiment, the process of performing step 402 may specifically include the following steps: The mean neighborhood distance is obtained by summing the Euclidean distances between each healthy neighbor pixel and the coordinates of the bad pixel in the healthy neighbor pixel set and dividing by the number of valid pixels. Based on the mean neighborhood distance and the Euclidean distance from each healthy neighboring pixel to the coordinates of the bad pixel, a brightness compensation coefficient that is inversely proportional to the distance is constructed. Then, a unit base value is superimposed on the brightness compensation coefficient to obtain the brightness compensation gain of each healthy neighboring pixel.

[0049] Specifically, it calls each healthy neighboring pixel within the set. coordinates of bad pixels The Euclidean distance between them has already been calculated. The accumulation unit reads the distance values ​​one by one along the coordinate order of the healthy neighboring pixel set, accumulates all the distance values, and then divides them by the number of effective pixels. The mean neighborhood distance is obtained. The computational relationship is expressed as follows:

[0050] in, This represents the average spatial distance from all healthy neighboring pixels to the current bad pixel. Indicates the number of valid pixels. Represents healthy neighboring pixels The Euclidean distance to the coordinates of the defective pixel. When the neighboring pixels are densely distributed, the average distance is relatively small, and the compensation amplitude converges accordingly. When the neighboring pixels are sparsely distributed, the average distance is relatively large, and the compensation amplitude can be appropriately increased, thus maintaining the consistency of the compensation strategy at different defective pixel locations. The controller calculates the brightness compensation gain for each healthy neighboring pixel in the healthy neighboring pixel set. The compensation gain formula is:

[0051] in, Represents healthy neighboring pixels The brightness compensation gain. The "1" in the formula is used to maintain the original brightness output as the base component. Used to add extra compensation to the location of adjacent bad pixels, so when When the value is smaller, healthy neighboring pixels are closer to bad pixels, the denominator decreases accordingly, and the compensation gain increases accordingly; when... When the distance is large, healthy neighboring pixels are relatively farther from the defective pixel, and the compensation gain naturally decreases. This method of distributing compensation in reverse according to distance allows positions closer to the defective pixel to undertake more local brightness correction tasks. After the controller completes the gain calculation, it can write the coordinates, Euclidean distance, mean neighborhood distance, and brightness compensation gain of each healthy neighboring pixel into the gain register area, forming a set of local brightness compensation parameters around the coordinates of the defective pixel.

[0052] The above describes the method for dynamic replacement of dead pixels in a full-color module in the embodiments of this application. The following describes the LED display screen in the embodiments of this application. One embodiment of the LED display screen in this application is used to perform the method for dynamic replacement of dead pixels in a full-color module.

[0053] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0054] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0055] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for dynamic replacement of defective pixels in a full-color module, characterized in that, include: During the blanking period triggered by the frame synchronization signal, PWM readback commands are sent to each pixel in the full-color module one by one, and the first output duty cycle of the R, G, and B channels of each pixel is returned to generate the measured duty cycle matrix. Based on the measured duty cycle matrix, a bad pixel type analysis is performed on each pixel to obtain a bad pixel mapping table. Boundary verification and health screening are performed on the surrounding pixels of each bad pixel coordinate in the bad pixel mapping table to obtain a healthy neighbor pixel set, and a neighborhood weight list between the healthy neighbor pixel set and the bad pixel coordinate is calculated. The brightness compensation gain of the healthy neighborhood pixel set is calculated based on the neighborhood weight list. The compensation replacement value is injected into the pixel output register of the bad pixel coordinates. At the same time, the brightness compensation gain is applied to complete the dynamic replacement of bad pixels.

2. The method for dynamic replacement of defective pixels in a full-color module according to claim 1, characterized in that, During the blanking period triggered by the frame synchronization signal, PWM readback commands are sent sequentially to each pixel in the full-color module, and the first output duty cycle of the R, G, and B channels of each pixel is returned, generating a measured duty cycle matrix, including: The falling edge transition of the monitoring frame synchronization signal is activated at the falling edge moment, and the duration of the reading acquisition window is constrained within the blanking period. PWM reading instructions are sent to each pixel in the full-color module to obtain the first output duty cycle of the R, G, and B channels of each pixel. The first output duty cycle of each pixel's R, G, and B channels is transmitted back to the controller via the SPI serial bus and written into the on-chip memory area in pixel coordinate order to generate the measured duty cycle matrix.

3. The method for dynamic replacement of defective pixels in a full-color module according to claim 2, characterized in that, The step of performing a bad pixel type analysis on each pixel based on the measured duty cycle matrix to obtain a bad pixel mapping table includes: The R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel are calculated based on the measured duty cycle matrix. Based on the R channel deviation rate, the G channel deviation rate, and the B channel deviation rate, a bad pixel type analysis is performed on each pixel to obtain a bad pixel mapping table.

4. The method for dynamic replacement of defective pixels in a full-color module according to claim 3, characterized in that, The calculation of the R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel based on the measured duty cycle matrix includes: Read the measured duty cycle matrix and the duty cycle of the same pixel coordinate in the reference duty cycle lookup table in the row and column scanning order. Divide the measured duty cycle of the R channel of each pixel with the corresponding reference duty cycle of the R channel in the reference duty cycle lookup table and then divide by the reference duty cycle of the R channel to obtain the R channel deviation rate of each pixel. The G-channel deviation rate of each pixel is obtained by subtracting the measured duty cycle of the G-channel from the corresponding G-channel reference duty cycle in the reference duty cycle lookup table and then dividing by the G-channel reference duty cycle. The B-channel deviation rate of each pixel is obtained by subtracting the measured duty cycle of the B-channel from the corresponding B-channel reference duty cycle in the reference duty cycle lookup table and then dividing by the reference duty cycle of the B-channel.

5. The method for dynamic replacement of defective pixels in a full-color module according to claim 4, characterized in that, The step of performing bad pixel type analysis on each pixel based on the R channel deviation rate, the G channel deviation rate, and the B channel deviation rate to obtain a bad pixel mapping table includes: The R-channel deviation rate, G-channel deviation rate, and B-channel deviation rate of each pixel are compared with the brightness deviation threshold channel by channel. When the deviation rate of any channel is not lower than the brightness deviation threshold, the corresponding pixel is determined to be a brightness defect. The deviation range value is calculated based on the R channel deviation rate, the G channel deviation rate, and the B channel deviation rate, and the deviation range value is compared with the chromaticity range threshold. When the deviation range value is not lower than the chromaticity range threshold, the bad pixel type of the corresponding pixel is determined to be a chromaticity offset bad pixel. When there is a pixel whose bad pixel type satisfies both luminance bad pixel and chromaticity offset bad pixel, the bad pixel type of the corresponding pixel is determined to be a mixed bad pixel. Write the defect type of each pixel along with the corresponding defect coordinates into the defect mapping table.

6. The method for dynamic replacement of defective pixels in a full-color module according to claim 5, characterized in that, The step of performing boundary verification and health screening on the surrounding pixels of each bad pixel coordinate in the bad pixel mapping table to obtain a healthy neighbor pixel set, and calculating a neighborhood weight list between the healthy neighbor pixel set and the bad pixel coordinates, includes: Based on the coordinates of each bad point in the bad point mapping table, an 8-neighbor candidate pixel set is constructed, and boundary verification and health screening are performed on each candidate coordinate in the 8-neighbor candidate pixel set in sequence. The candidate coordinates that pass the boundary verification and health screening are added to the candidate neighbor pixel set. Count the number of valid pixels in the candidate neighbor pixel set and determine the healthy neighbor pixel set; Calculate the Euclidean distance between the target coordinates of each healthy neighbor pixel in the healthy neighbor pixel set and the coordinates of the bad pixel, and take the reciprocal of the Euclidean distance as the reciprocal distance weight of the corresponding healthy neighbor pixel; A neighborhood weight list is generated based on the target coordinates and the corresponding inverse distance weights.

7. The method for dynamic replacement of defective pixels in a full-color module according to claim 6, characterized in that, The step of counting the number of valid pixels in the candidate neighboring pixel set and determining the healthy neighboring pixel set includes: The number of valid pixels in the candidate neighborhood pixel set is counted, and the number of valid pixels is compared with a threshold. When the number of effective pixels is not lower than the number threshold, the candidate neighborhood pixel set is directly determined as the healthy neighborhood pixel set; When the number of effective pixels is lower than the number threshold, the neighborhood range is expanded from 8 neighborhoods to 16 neighborhoods with the bad pixel coordinates as the center, and new candidate coordinates are extracted. Boundary verification and health screening are performed on the new candidate coordinates in sequence. The new candidate coordinates that pass the boundary verification and health screening are added to the candidate neighborhood pixel set to obtain the healthy neighborhood pixel set.

8. The method for dynamic replacement of defective pixels in a full-color module according to claim 7, characterized in that, The process of calculating the brightness compensation gain of the healthy neighborhood pixel set based on the neighborhood weight list, injecting the compensation replacement value into the pixel output register of the bad pixel coordinates, and simultaneously applying the brightness compensation gain to complete the dynamic replacement of bad pixels includes: The compensation replacement value for the bad pixel coordinates is calculated based on the second output duty cycle of the R, G, and B channels of each healthy neighbor pixel in the neighborhood weight list and the measured duty cycle matrix, and the compensation replacement value is written into the pixel output register of the bad pixel coordinates. Calculate the average neighborhood distance from all healthy neighbor pixels in the healthy neighbor pixel set to the coordinates of the bad pixels, and calculate the brightness compensation gain for each healthy neighbor pixel based on the average neighborhood distance. The second output duty cycle of the R, G, and B channels of each healthy neighboring pixel is multiplied by the brightness compensation gain to obtain the target value. The minimum value of the target value and the upper limit of the output duty cycle is written back to the pixel output register of each healthy neighboring pixel to complete the dynamic replacement of bad pixels.

9. The method for dynamic replacement of defective pixels in a full-color module according to claim 8, characterized in that, The calculation of the average neighborhood distance from all healthy neighboring pixels in the healthy neighboring pixel set to the coordinates of the bad pixels, and the calculation of the brightness compensation gain for each healthy neighboring pixel based on the average neighborhood distance, includes: The average neighborhood distance is obtained by summing the Euclidean distances between each healthy neighbor pixel in the healthy neighbor pixel set and the coordinates of the bad pixel, and then dividing by the number of effective pixels. Based on the mean neighborhood distance and the Euclidean distance from each healthy neighboring pixel to the coordinates of the bad pixel, a brightness compensation coefficient inversely proportional to the distance is constructed, and a unit base value is superimposed on the brightness compensation coefficient to obtain the brightness compensation gain of each healthy neighboring pixel.

10. An LED display screen, characterized in that, Used to perform the dynamic replacement method for bad pixels in a full-color module as described in any one of claims 1-9.

Images