A method and system for detecting dead pixels of a liquid crystal display screen

Abstract

The present application discloses a method and system for detecting dead pixels of a liquid crystal display screen, which relates to the detection of liquid crystal screens and includes: obtaining the area of the dead pixel positions of the liquid crystal screen according to a preset test scheme; constructing a physical topology model of the liquid crystal screen according to the physical parameters of the liquid crystal screen; using an improved watershed algorithm to segment the area of the dead pixel positions to obtain a segmentation result; using a pre-constructed dead pixel feature knowledge base to perform feature analysis on the segmentation result to determine the dead pixel name and the cause of the dead pixel to obtain a detection result; using the physical topology model and adopting a flow rule based on the physical connection path to enable the watershed algorithm to diffuse only along the physical connection path to obtain a segmentation result; aiming at the problem in the prior art that the watershed algorithm is prone to over-segmentation, resulting in low dead pixel detection accuracy, the present application improves the watershed algorithm by establishing physical constraints by using the regular structure of the LCD pixel array, thereby improving the dead pixel detection accuracy.

Description

Technical Field

[0001] This application relates to the field of liquid crystal display screen detection, and particularly to a method and system for detecting dead pixels of a liquid crystal display screen. Background Art

[0002] In the field of liquid crystal display screen dead pixel detection technology, traditional methods mainly rely on manual visual inspection or simple image processing algorithms, such as threshold segmentation, edge detection, etc. With the development of computer vision technology, more complex image processing algorithms have been introduced into the dead pixel detection field. Among them, the watershed algorithm has been widely used in the division and recognition of dead pixel regions due to its good image segmentation ability.

[0003] The watershed algorithm is based on the concept of watershed in geography. The image is regarded as a topographic map, where regions with high gray values are like mountain peaks, and regions with low gray values are like valleys. The algorithm simulates the process of water gradually rising from low points. When the water in different valleys is about to connect, a watershed line is established, thereby realizing the regional segmentation of the image. This method can effectively segment different regions under ideal conditions and is suitable for determining the boundary between dead pixels and normal display regions.

[0004] However, the traditional watershed algorithm has serious over-segmentation defects when processing liquid crystal display screen dead pixel images. Since the algorithm is completely based on the change of image gray gradient and ignores the physical structure characteristics of the liquid crystal screen, it causes dead pixels that should be a single fault region to be wrongly segmented into multiple unconnected sub-regions. Especially when dead pixels spread along the driving line or present an irregular shape, the traditional watershed algorithm will overly establish watershed lines at the gradient change points, resulting in a large number of meaningless small regions, seriously affecting the subsequent dead pixel classification and cause analysis.

[0005] For example, the related patent document CN119068787A discloses a method, device, and storage medium for detecting dead pixels of a liquid crystal screen. The test equipment is controlled based on a preset test scheme to test the target liquid crystal screen, and the screen characteristic image of the target liquid crystal screen is obtained. The actual gradient value matrix of the target liquid crystal screen is constructed according to the screen characteristic image; the dead pixel position analysis of the target liquid crystal screen is carried out based on the actual gradient value matrix and the standard gradient value matrix of the target liquid crystal screen to obtain the dead pixel position region in the screen characteristic image; the screen characteristic image is segmented to obtain the sub-characteristic image corresponding to the dead pixel position region in the target liquid crystal screen; the feature analysis of each sub-characteristic image is carried out based on the dead pixel feature knowledge tree diagram concept network to obtain the dead pixel names and dead pixel causes of all dead pixel position regions in the target liquid crystal screen. However, this scheme uses the traditional watershed algorithm and may wrongly segment a single dead pixel region into multiple sub-regions. Therefore, the detection accuracy of this scheme needs to be further improved. Summary of the Invention

[0006] In view of the problem in the prior art that the watershed algorithm is prone to over-segmentation, resulting in low precision in detecting bad pixels, the present application provides a method and system for detecting bad pixels of a liquid crystal display screen. By utilizing the regular structure of the LCD pixel array, a physical constraint is established to improve the watershed algorithm, thereby improving the precision of bad pixel detection.

[0007] The objectives of the present application are achieved through the following technical solutions.

[0008] One aspect of the present application provides a method for detecting bad pixels of a liquid crystal display screen, including: S1, obtaining the bad pixel position area of the liquid crystal screen according to a preset test scheme; S2, constructing a physical topology model of the liquid crystal screen according to the physical parameters of the liquid crystal screen; S3, using an improved watershed algorithm to segment the bad pixel position area to obtain a segmentation result; S4, performing feature analysis on the segmentation result using a pre-constructed bad pixel feature knowledge base to determine the bad pixel name and the cause of the bad pixel, thereby obtaining a detection result; wherein, using the improved watershed algorithm to segment the bad pixel position area includes: utilizing the physical topology model and adopting a flow rule based on the physical connection path, so that the watershed algorithm only diffuses along the physical connection path to obtain a segmentation result;

[0009] Further, the physical parameters include the RGB pixel arrangement mode, the sub-pixel structure, and the driving circuit layout; the bad pixel feature knowledge base contains the feature parameters, cause analysis, and identification rules of various types of bad pixels;

[0010] Further, S1, obtaining the bad pixel position area of the liquid crystal screen according to a preset test scheme includes: determining the standard gradient value corresponding to each pixel point in the liquid crystal screen according to the preset test scheme to generate a standard gradient value matrix; the standard gradient value represents the gray gradient value corresponding to the pixel point under the preset normal working conditions; obtaining the test image of the liquid crystal screen under the test conditions; the test image represents the display image of the liquid crystal screen in the test mode; constructing an actual gradient value matrix according to the test image; and determining the bad pixel position area of the liquid crystal screen by comparing the standard gradient value matrix and the actual gradient value matrix.

[0011] Further, in S2, a physical topology model of the liquid crystal display screen is constructed, including: obtaining the RGB pixel arrangement of the liquid crystal display screen, determining the physical connection relationship between adjacent pixels, and the positional relationship between adjacent pixels; establishing a two-dimensional pixel coordinate system according to the physical connection relationship between adjacent pixels and the positional relationship between adjacent pixels; determining the physical connection path at the sub-pixel level according to the physical boundaries of the sub-pixels of the liquid crystal display screen; obtaining the layout data of the driving lines of the liquid crystal display screen, determining the connection relationship between the pixel points and the gate lines and data lines according to the physical routing structure of the driving lines, and identifying the pixel point groups sharing the same driving line by tracing the driving signal transmission path; constructing a pixel connection weight matrix according to the RGB pixel arrangement, sub-pixel and driving line layout data; generating a gradient weight matrix containing physical constraints according to the pixel connection weight matrix and the standard gradient value matrix; establishing a driving line fault model according to the driving line layout data through the transmission characteristics and attenuation law of the driving signal on the line; and constructing a physical topology model of the liquid crystal display screen according to the two-dimensional pixel coordinate system, the physical connection path at the sub-pixel level, the pixel connection weight matrix and the driving line fault model.

[0012] Among them, a sub-pixel is the basic display unit that constitutes a single complete pixel in a liquid crystal display screen. It is usually composed of light-emitting units of three colors: red (R), green (G), and blue (B). Each sub-pixel can independently control its brightness level. The sub-pixel is the smallest physical unit for constructing the physical topology model. Its physical boundaries and arrangement structure determine the basic physical constraints of signal transmission and fault propagation, and are the basis for accurately analyzing the formation and diffusion mechanism of bad pixels.

[0013] The physical connection path refers to the signal transmission channel or mutual influence channel at the physical level between adjacent pixels or sub-pixels in the liquid crystal display screen, which characterizes the physical path through which the fault signal may propagate between display units. The physical connection path is determined based on the actual physical structure of the liquid crystal display screen, including wire connections in the TFT array layer, stress conduction between liquid crystal molecules, electric field coupling of shared electrodes, etc.; the physical connection path defines the legal flow channel of the "water flow" in the watershed algorithm, ensuring that the algorithm segmentation result conforms to the actual physical structure constraints and preventing the wrong segmentation of regions that should be physically connected.

[0014] The connection relationship between the pixel points and the gate lines and data lines describes the physical and electrical connection methods between the display units and the driving signal transmission lines, reflecting the addressing and driving mechanisms of the TFT liquid crystal display screen. The connection relationship between the pixel points and the driving lines reveals the physical channels of signal transmission in the liquid crystal display screen, which is the key basis for identifying the bad pixels of the driving line fault type and predicting the fault propagation path, and is also an important reference for constructing the pixel connection weight matrix.

[0015] A group of pixel points sharing the same driving line refers to a set of pixel points controlled by the same gate line or data line. These pixel points have a physical connection in the driving signal transmission path and may exhibit similar failure characteristics.

[0016] Among them, the gradient weight matrix containing physical constraints generated according to the pixel connection weight matrix and the standard gradient value matrix provides an accurate gradient value basis for the watershed algorithm, making the path selection in the immersion process more in line with the physical characteristics of the liquid crystal screen. This gradient control mechanism under physical constraints fundamentally ensures that the water flow only diffuses along the physical connection path, avoiding the wrong diffusion across physical isolation regions in traditional algorithms.

[0017] Furthermore, constructing the pixel connection weight matrix includes: obtaining the physical distance between pixel points and the sharing situation of driving lines; setting the connection weight of pixel points with a physical distance less than the threshold and sharing the driving line to N1; setting the connection weight of pixel points with a physical distance less than the threshold but not sharing the driving line to N2; setting the connection weight of pixel points with a physical distance exceeding the threshold to N3.

[0018] Furthermore, N1 is greater than N2, and N2 is greater than N3.

[0019] Furthermore, in S3, an improved watershed algorithm is used to segment the bad point position area to obtain the segmentation result, including: taking the bad point position area determined in S1 as the input area of the watershed algorithm; taking the physical connection path in the physical topology model constructed in S2 as the flow constraint condition of the watershed algorithm; setting the gradient value of the watershed algorithm according to the gradient weight matrix containing physical constraints generated in S2; setting the initial water level line and taking the pixel point with the lowest gradient value as the initial immersion point; during the immersion process, controlling the water flow diffusion path according to the pixel connection weight matrix, setting the diffusion priority of the path with connection weight N1 to P1, setting the diffusion priority of the path with connection weight N2 to P2, and setting the diffusion priority of the path with connection weight N3 to P3; predicting the potential bad point diffusion direction using the driving line fault model in S2 and adjusting the region growth strategy of the watershed algorithm; when the water level rises to the preset height, detecting the minimum pixel distance between adjacent sub-regions, establishing a watershed line according to the minimum pixel distance; according to the established watershed line, analyzing the corresponding continuity of the identified group of pixel points sharing the same driving line in the two-dimensional pixel coordinate system, and assigning the same region label to the continuously distributed pixel groups not separated by the watershed line; merging the pixel groups with the same region label to form multiple independent bad point sub-regions as the segmentation result.

[0020] Among them, according to the established watershed line, continuity analysis is performed on the pixel point groups sharing the same driving line, and the pixel groups that are continuously distributed and not separated by the watershed line are given the same area label, which technically ensures that each physically connected bad pixel area is correctly identified as a whole. The mechanism of merging pixel groups with the same area label to form independent bad pixel sub-areas avoids the problem of bad pixel fragmentation that may occur in traditional methods.

[0021] Further, P1 is greater than P2, and P2 is greater than P3;

[0022] Further, when the water level rises to a preset height, the minimum pixel distance between adjacent sub-areas is detected, and a watershed line is established according to the minimum pixel distance, including: calculating the minimum pixel distance between each sub-area formed during the immersion process; when the minimum pixel distance between adjacent sub-areas is less than the threshold, obtaining the position information of the pixel points of the adjacent sub-areas in the two-dimensional pixel coordinate system constructed in S2; according to the position information, obtaining the connection relationship between the pixel points in the adjacent sub-areas and the gate line and data line in the pixel point driving line layout data in S2; according to the position information and the connection relationship between the pixel points and the gate line and data line, obtaining the physical connection path of the pixel points in the adjacent sub-areas determined in S2, where the physical connection path is used to judge whether there is a direct physical path between adjacent sub-areas; when adjacent sub-areas share the same driving line and there is a direct physical path on the physical connection path, no watershed line is established; when adjacent sub-areas do not share the same driving line or there is no direct physical path on the physical connection path, a watershed line is established.

[0023] Among them, the minimum pixel distance refers to the shortest spatial distance between the boundary pixel points of two adjacent sub-areas formed during the immersion process of the watershed algorithm. The calculation method is the minimum Euclidean distance between the boundary pixel sets of the two sub-areas: , and are both boundary images; the distance can be calculated in the pixel coordinate space or converted into a physical distance (millimeter). The minimum pixel distance is used as a conditional parameter to trigger physical connection analysis. When the distance is less than the preset threshold, the algorithm will further analyze the physical connection characteristics of adjacent areas. This two-stage strategy significantly improves the algorithm efficiency.

[0024] The physical connection path refers to the possible signal transmission channels or fault propagation channels between pixel points determined based on the actual physical structure of the liquid crystal display screen, reflecting the true connection relationship of the internal components of the liquid crystal display screen. It is jointly determined by the TFT array layer structure, the driving line layout, and the sub-pixel arrangement of the liquid crystal display screen. It includes multiple types: Electrical connection path: such as the connection between pixel points sharing the same gate line or data line; Molecular mechanics connection path: the connection formed through the stress transfer between liquid crystal molecules; Optical connection path: the influence propagation path formed by optical crosstalk between pixels; It has directionality and intensity attributes, reflecting the ease of physical influence propagation in different directions; It is usually represented by a weighted directed graph structure, with pixel points as nodes, physical connections as edges, and weights representing connection strength; It is pre-calculated and stored based on the physical boundaries of sub-pixels and the driving line layout data in stage S2. The physical connection path provides strict physical structure constraints, ensuring that the segmentation result of the watershed algorithm conforms to the actual physical connection characteristics of the liquid crystal display screen and avoiding incorrect segmentation of regions that should be physically connected.

[0025] The direct physical path refers to the unobstructed physical connection channel existing between adjacent sub-regions, allowing physical signals or fault effects related to bad points to directly propagate from one region to another, indicating that these sub-regions essentially belong to the same bad point region in terms of physics. The determination criteria include: The boundary pixel points in adjacent sub-regions share the same driving line (gate line or data line); There is a continuous physical connection path between sub-regions without a physical isolation structure blocking it; The weight value of the connection path is higher than the preset threshold, indicating that the connection strength is sufficient to support fault propagation; The verification method usually uses path tracing algorithms such as the A* or Dijkstra algorithm to find the shortest path between two regions on the physical connection path graph. The existence of the direct physical path is the key condition for determining whether to merge adjacent sub-regions, ensuring that the segmentation result of the watershed algorithm conforms to the physical formation mechanism and propagation law of bad points and preventing the over-segmentation problem.

[0026] The watershed line is the regional boundary line formed in the improved watershed algorithm, which is used to separate different bad pixel sub-regions. It is established based on the comprehensive evaluation of image gradient, physical distance, and physical connectivity, rather than relying solely on the image gradient. Differences from the traditional watershed line: The traditional watershed line is formed only based on the image gradient, which is prone to over-segmentation; the improved watershed line comprehensively considers the physical connection relationship and has physical significance; the minimum pixel distance between adjacent regions is less than the threshold (preliminary condition); adjacent regions do not share the same driving line or there is no direct physical path (decisive condition); the accurate positioning of the watershed line usually adopts the middle position of the boundary of adjacent sub-regions or is determined based on the local gradient maximum; each watershed line has a weight attribute, which reflects the confidence of the segmentation decision. The watershed line with a high weight has higher reliability. The watershed line is the final boundary for the segmentation of the bad pixel region, which determines the range and shape of each bad pixel sub-region in the detection result. Compared with the traditional method, the watershed line based on physical constraints more accurately reflects the true physical boundary of the bad pixels, significantly improving the accuracy of bad pixel segmentation and the reliability of subsequent analysis.

[0027] Another aspect of the present application also provides a liquid crystal display bad pixel detection system for implementing a liquid crystal display bad pixel detection method of the present application.

[0028] Compared with the prior art, the advantages of the present application are as follows:

[0029] (1) The traditional watershed algorithm has a serious over-segmentation defect when processing liquid crystal display bad pixel images. Since it only based on the change of image gray gradient and completely ignores the physical structure characteristics of the liquid crystal screen, it causes the bad pixels that should be a single fault region to be wrongly segmented into multiple unconnected sub-regions. The present application constructs a physical topology model, converts the RGB pixel arrangement, sub-pixel structure, and driving line layout of the liquid crystal screen into algorithm constraints, and makes the flow rule of the watershed algorithm conform to the physical connection characteristics of the liquid crystal screen, avoiding the over-segmentation phenomenon caused by the traditional watershed algorithm not considering the physical structure in terms of technical principle.

[0030] During the immersion process of the watershed algorithm, the water flow diffusion priority (P1, P2, P3) is controlled by the pixel connection weight matrix (N1, N2, N3), so that the pixel points sharing the driving line are preferentially merged, fundamentally preventing the bad pixels that should be classified into one category physically from being wrongly segmented.

[0031] (2) The traditional bad pixel boundary determination method has defects of boundary blur and boundary offset. It only relies on the pixel gray gradient and does not consider the physical separation structure of the LCD panel, and cannot accurately locate the bad pixel boundary in the region where the gradient change is not obvious. The sub-pixel level physical connection path established according to the physical parameters of the liquid crystal screen in the present application provides a strict physical basis for the establishment of the watershed line, making the segmentation boundary highly consistent with the actual physical structure of the liquid crystal screen.

[0032] When the water level rises to a preset height, by determining whether adjacent sub-regions share drive lines and whether there are physical paths, it intelligently decides whether to establish a watershed line, avoiding the problem of setting unnecessary dividing lines in physically connected regions.

[0033] (3) Traditional dead pixel detection methods have misclassification defects when dealing with dead pixels caused by drive line failures. They cannot recognize the physical correlation between dead pixels propagating along the same drive line, and often misjudge multiple dead pixels belonging to the same fault source as independent faults. This application uses a drive line fault model to predict the direction of dead pixel diffusion and adjusts the region growth strategy of the watershed algorithm, enabling the algorithm to identify dead pixels propagating along the drive line and correctly classify them into a single dead pixel region rather than multiple independent dead pixels. By reducing the gradient value weight between pixel points on the same drive line and increasing the diffusion priority along the drive line direction, it ensures from the technical principle that the segmentation result conforms to the physical fault propagation law of the liquid crystal display screen. Brief Description of the Drawings

[0034] This application will be further described in the form of exemplary embodiments, which will be described in detail through the drawings. These embodiments are not restrictive. In these embodiments, the same numbers represent the same structures, where:

[0035] Figure 1 is an exemplary flowchart of a method for detecting dead pixels of a liquid crystal display screen according to some embodiments of this application;

[0036] Figure 2 is an exemplary flowchart of obtaining the dead pixel position region according to some embodiments of this application;

[0037] Figure 3 is an exemplary flowchart of constructing a physical topology model of a liquid crystal display screen according to some embodiments of this application;

[0038] Figure 4 is an exemplary flowchart of obtaining a segmentation result according to some embodiments of this application. Detailed Description of the Embodiments

[0039] The methods and systems provided in the embodiments of this application will be described in detail below with reference to the drawings.

[0040] As Figure 1As shown, obtain the defective pixel position area of the liquid crystal display screen according to a preset test scheme; construct a physical topology model of the liquid crystal display screen according to the physical parameters of the liquid crystal display screen; use an improved watershed algorithm to segment the defective pixel position area to obtain a segmentation result; use a pre-constructed defective pixel feature knowledge base to perform feature analysis on the segmentation result to determine the defective pixel name and the cause of the defective pixel to obtain a detection result; among them, using the improved watershed algorithm to segment the defective pixel position area includes: using the physical topology model and adopting a flow rule based on the physical connection path, so that the watershed algorithm only diffuses along the physical connection path to obtain a segmentation result.

[0041] As Figure 2 shown, S1, obtaining the defective pixel position area of the liquid crystal display screen according to a preset test scheme includes: The preset test scheme includes a standard display pattern sequence designed for the characteristics of different types of liquid crystal display screens, usually including various test patterns such as pure white, pure black, pure red / green / blue, gray-scale gradient, and checkerboard. For each test pattern, the theoretical gray-scale gradient value that each pixel should have under each test pattern in the ideal state is determined in advance through theoretical calculation or through statistical analysis of a large number of qualified samples. The gray-scale gradient value represents the change rate of the gray-scale value between adjacent pixel points, and usually the Sobel operator, Prewitt operator, or Scharr operator is used to calculate the gradient values of each pixel point in the horizontal and vertical directions. For each test pattern, a corresponding standard gradient value matrix is generated, the dimension of the matrix is the same as the resolution of the liquid crystal display screen, and each element in the matrix represents the standard gradient value of the corresponding position pixel point.

[0042] Use a dedicated liquid crystal display screen test device to send the same test pattern sequence as the standard test scheme to the liquid crystal display screen to be tested through display interfaces such as HDMI, DVI, DisplayPort or directly driving the driving board of the liquid crystal display screen. After each test pattern is displayed stably (usually waiting for a stable time of 20 - 50 ms), use a high-resolution industrial camera or a dedicated image acquisition device to collect the actual display image of the liquid crystal display screen. The collected image needs to be preprocessed, including image calibration, perspective transformation correction, noise filtering, etc., to ensure that the collected image can accurately reflect the actual display state of the liquid crystal display screen. For large-size or high-resolution liquid crystal display screens, multiple sub-region acquisitions and stitching may be required to ensure that sufficient detailed information is captured.

[0043] For each acquired test image, use the same gradient calculation method as the standard gradient value matrix (such as the same Sobel operator) to calculate the actual gradient value of each pixel point in the image. If the test image is a color image, first decompose it into three channels: R, G, and B, calculate the gradient values of each channel respectively, and then synthesize the gradient values of the three channels into a comprehensive gradient value according to the sub-pixel structure characteristics of the liquid crystal screen. Considering the possible position offset during the camera acquisition process, use image registration technology to ensure that the pixel points of the test image strictly correspond to the corresponding positions in the standard gradient value matrix. The constructed actual gradient value matrix also has the same dimension as the resolution of the liquid crystal screen, and each element represents the actual gradient value of the pixel point at the corresponding position in the test image.

[0044] For each pixel point, calculate the difference between its actual gradient value and the standard gradient value to form a difference matrix. Set a reasonable threshold range. If the difference value of a certain pixel point exceeds the preset threshold, mark it as a potential bad pixel. To avoid false positive detections caused by ambient light interference and measurement errors, a method of comprehensive analysis of multiple test results is usually adopted, that is, a pixel point is determined as a bad pixel only when it shows abnormalities under multiple test patterns. Perform connectivity analysis on the marked potential bad pixel regions, combine adjacent bad pixel combinations into bad pixel regions, and each region has clear boundary coordinates and an internal pixel list. Finally, output the bad pixel position region information, including the position coordinates (upper left and lower right coordinates), area (number of pixels), shape features (such as rectangularity, circularity) of each bad pixel region, etc., as the input for subsequent analysis.

[0045] As Figure 3 shown, S2, construct a physical topology model of the liquid crystal screen according to the physical parameters of the liquid crystal screen; the physical parameters include the RGB pixel arrangement, sub-pixel structure, and driving circuit layout, including: first, it is necessary to obtain the RGB pixel arrangement of the liquid crystal screen to determine the physical connection relationship between adjacent pixels: obtain the RGB sub-pixel arrangement of the liquid crystal screen from the liquid crystal screen specification or through microscope measurement. Common arrangement methods include RGB stripe arrangement, RGB triangle arrangement, PenTile arrangement, and RGBW four-element arrangement, etc. For the RGB stripe arrangement, adjacent pixels usually have a direct physical connection relationship in the horizontal direction, and are connected through the electrodes of the TFT array layer in the vertical direction. For the PenTile arrangement, the sub-pixel sharing situation needs to be considered specially, and the physical connection relationship between adjacent pixels is more complex, usually showing an irregular grid structure. Measure and record the physical pixel pitch and sub-pixel pitch. These parameters directly affect the physical distance and connection strength between adjacent pixels. At the same time, determine the positional relationship between adjacent pixels, including 8-connected or 4-connected neighborhood relationships, as well as the physical distance and actual connection strength in each direction.

[0046] Based on the physical connection relationship and positional relationship between adjacent pixels, establish a two-dimensional pixel coordinate system: Take the pixel in the upper left corner of the liquid crystal screen as the origin (0, 0), establish a rectangular coordinate system, where the horizontal axis represents the column position of the pixel, and the vertical axis represents the row position of the pixel. For a liquid crystal screen with a conventional rectangular arrangement, the coordinate system is a regular grid; for a non-rectangular arrangement (such as a hexagonal arrangement), a transformation matrix needs to be established to map the actual physical positions to the two-dimensional coordinate system. Assign a unique coordinate identifier (x, y) to each pixel, and establish a pixel index table to facilitate the rapid positioning and access to any pixel. Record the list of adjacent pixels of each pixel in the coordinate system, including the directly adjacent first-order neighborhood and the indirectly adjacent higher-order neighborhood, and construct a pixel adjacency relationship graph.

[0047] Determine the physical connection path at the sub-pixel level according to the physical boundaries of the sub-pixels of the liquid crystal screen: Obtain the physical structure of the sub-pixels through a high-resolution microscope or design drawings, and measure the geometric dimensions and shapes of the sub-pixel boundaries. Analyze the physical isolation structures between sub-pixels, such as the black matrix and the edges of color filters, which will block or weaken the optical and electrical signal propagation between adjacent sub-pixels. Based on the physical boundaries of the sub-pixels, determine the connection path diagram at the sub-pixel level, and identify which sub-pixel boundaries allow signal transmission and which boundaries are completely isolated. For sub-pixels of different colors (R, G, B), consider their differences in physical structure and electrical characteristics, and establish connection path models respectively. Integrate the connection paths at the sub-pixel level to form a physical path diagram at the pixel level, which describes the possible propagation paths of signals and faults.

[0048] Obtain the layout data of the driving circuits of the liquid crystal screen and determine the connection relationship between the pixel points and the driving circuits: Extract the driving circuit layout information from the technical specification documents provided by the liquid crystal panel manufacturer, analyze the pin definitions and signal layout diagrams in the data sheet of the driving IC, and extract the number of gate lines, the number of data lines, the routing directions, and the corresponding pixel mapping relationships. Obtain the PCB design diagram of the LCD module or scan the PCB layout through an electron microscope to identify the connection points of the flexible printed circuit (FPC) where the driving IC is connected to the panel, and trace the connection positions and arrangement orders of the signal lines on the FPC to the panel edge. Use a high-resolution infrared camera to image the panel at a specific wavelength, use optical filtering technology to enhance the visibility of the TFT layer wiring, and identify the physical layout structures of the gate lines and data lines through image processing algorithms, construct the wiring vector diagrams of the gate lines and data lines, and record the starting points, ending points, and directions of each line.

[0049] Establish a two-dimensional mapping table from (row, column) to (gate_line, data_line). For a standard matrix driving structure, the mapping generally follows the rule relationship: gate_line = row_index + gate_offset, data_line = column_index + data_offset. Record mapping anomaly points and special regions, such as multiplexing regions and edge compensation regions. Design a sequential test pattern to activate single driving lines in sequence, and record the set of pixel coordinates that respond when each driving line is activated. For each pixel, establish its associated gate line ID and data line ID, and construct a two-dimensional connection relationship matrix M, where M[i, j] represents the driving lines connected to the pixel (i, j).

[0050] Based on the physical structure diagram of the TFT array layer of the liquid crystal screen, identify the positions of the TFT devices of each pixel, trace the physical connection path from the TFT gate to the gate line, and trace the physical connection path from the TFT source to the data line. For a common-source design, specifically mark the pixel groups sharing the data line. Select an appropriate driving structure model according to the design type of the liquid crystal screen (such as a-Si TFT, LTPS, Oxide TFT, etc.), and establish a calculation formula for the connection relationship reflecting specific technical parameters, considering the connection variations of special driving schemes (such as in-plane gate, double-gate design).

[0051] Based on the transmission line theory, establish a propagation model of the driving signal on the gate line and the data line, record the position of the signal source (output end of the driving IC) and the signal transmission direction, calculate the transmission delay and intensity attenuation of the signal reaching each pixel, and establish a signal propagation feature library for typical faults (such as open circuit, short circuit, high resistance). Use circuit simulation tools such as SPICE to construct an equivalent circuit model of the driving line, simulate the signal transmission characteristics on the gate line and the data line, analyze the signal propagation characteristics under different load conditions, and record the key nodes and signal integrity characteristics on the signal propagation path.

[0052] Record the row scanning order and column data loading order in the driving timing, analyze the set of pixels activated simultaneously within the same clock cycle, identify pixel groups related to timing, and these pixels may show similar responses to timing faults. Construct a timing dependency graph to describe the timing propagation sequence of the driving signal. Measure the signal integrity parameters at different positions on the gate line and the data line, record characteristics such as signal rise time, fall time, overshoot, and ringing, establish the correlation between signal integrity and pixel response, and identify signal degradation sensitive regions and potential high-fault regions.

[0053] For each gate line Li, define the pixel set G(Li) = {P(r, c) | P(r, c) is connected to the gate line Li}. Usually, G(Li) contains all the pixel points in the same row of the display screen. For special layouts (such as staggered gates), G(Li) may contain partial pixels in multiple rows. Assign the same gate line group identifier to the pixel points in G(Li), construct a gate line group list, and record the set of pixel coordinates included in each group.

[0054] For each data line Di, define the pixel set D(Di) = {P(r, c) | P(r, c) is connected to the data line Di}. Usually, D(Di) contains all the pixel points in the same column of the display screen. For multiplexed layouts, D(Di) may contain alternating column pixels or pixels in a specific pattern. Assign the same data line group identifier to the pixel points in D(Di), construct a data line group list, and record the set of pixel coordinates included in each group.

[0055] Based on the gate line groups and data line groups, construct more complex connection relation groups to identify the grouping of the output channels of the gate driver and data driver. For multiple gate lines sharing the same gate driver channel, construct a high-level gate group. For multiple data lines sharing the same data driver channel, construct a high-level data group. These multi-level groups are particularly important for identifying bad pixels of the drive IC failure type. Identify the pixel positions corresponding to the intersections of the gate lines and data lines, analyze the possible failure modes caused by the intersections of specific drive lines, construct an intersection connection matrix, and record the physical characteristics and potential risks of each intersection.

[0056] According to the RGB pixel arrangement, sub-pixel, and drive line layout data, construct a pixel connection weight matrix: Obtain the physical distance between pixel points, usually using the Euclidean distance or Manhattan distance, and select an appropriate distance metric in combination with the actual physical layout characteristics. Analyze the sharing situation of the drive lines between pixel points to determine whether two pixel points share the same gate line or data line. Set a physical distance threshold, usually selected as 1.5 times the pixel pitch, as a reference for judging the physical connection strength between pixels. According to the physical distance and drive line sharing situation, assign weights to the connections between each pair of pixels: For pixel points with a physical distance less than the threshold and sharing a drive line, set a high connection weight N1 (for example, with a value of 9.0); for pixel points with a physical distance less than the threshold but not sharing a drive line, set a medium connection weight N2 (for example, with a value of 3.0); for pixel points with a physical distance exceeding the threshold, set a low connection weight N3 (for example, with a value of 0.5); the weight values satisfy the relationship N1 > N2 > N3. Usually, N1 is about 3 times N2, and N2 is about 6 times N3. This setting ensures that pixel points sharing a drive line have a significantly higher weight advantage than other connections. Organize the connection weights between all pairs of pixel points in matrix form to form a pixel connection weight matrix.

[0057] Generate a gradient weight matrix with physical constraints based on the pixel connection weight matrix and the standard gradient value matrix: fuse the standard gradient value matrix generated in S1 with the pixel connection weight matrix to obtain the gradient value considering physical connection characteristics. For pixel pairs with high connection weights (such as adjacent pixels sharing a driving line), reduce their gradient values so that they are less likely to form a segmentation boundary in the watershed algorithm. For pixel pairs with low connection weights (such as physically isolated pixels), maintain or increase their gradient values so that they are more likely to form a segmentation boundary in the watershed algorithm. The specific fusion method uses the gradient value adjustment formula: adjusted gradient value = original gradient value × (1 - α × connection weight), where α is an adjustment factor, usually taking values between 0.1 and 0.3. The generated physical constraint gradient weight matrix maintains the same size as the original image, but each element value has incorporated the physical connection constraint information.

[0058] Establish a driving line fault model based on the driving line layout data: analyze the electrical characteristics and optical manifestations of different types of driving line faults, such as gate line open circuit, data line short circuit, etc. Study the transmission characteristics of driving signals on the lines, including signal delay, attenuation, and crosstalk. Establish a mathematical model to describe how the fault signal propagates along the driving line, including propagation speed, attenuation coefficient, and influence range. For common fault types, such as open circuit, short circuit, high impedance, leakage, etc., establish prediction models respectively to infer the fault source and possible diffusion directions. Use historical fault data to verify and optimize the model parameters to improve the accuracy of fault propagation prediction.

[0059] Finally, integrate the foregoing components to construct a complete physical topology model of the liquid crystal display: integrate the two-dimensional pixel coordinate system, sub-pixel level physical connection paths, pixel connection weight matrix, and driving line fault model to form a multi-level physical topology model. The model is represented in a graph structure, with nodes being pixel points and edges being physical connection relationships, and the weights of the edges representing the connection strength. Attach a driving line layer to the model to record the subordinate relationship between each pixel point and the driving line, as well as the cross-connections between driving lines. Establish a mapping relationship from sub-pixels to pixels to enable model analysis and operations at different precision levels. The finally formed physical topology model can accurately express the physical structure characteristics of the liquid crystal display and provide an important basis for the subsequent watershed algorithm based on physical constraints.

[0060] Such as Figure 4As shown in the figure, in step S3, an improved watershed algorithm is used to segment the bad pixel position area to obtain the segmentation result. First, the information obtained in the previous steps is used as the input of the watershed algorithm: the bad pixel position area determined in S1 is used as the input area of the watershed algorithm. These areas are usually sets of pixels with obvious gray-scale anomalies and have been preliminarily determined as areas where bad pixels may exist. The input area is usually represented as a binary mask image, where the pixel value of the bad pixel area is 1 and the pixel value of the non-bad pixel area is 0. To improve the calculation efficiency, usually each independent bad pixel position area is cut into sub-regions so that the algorithm can process multiple independent regions in parallel. Each sub-region will be appropriately expanded before processing to include a small number of normal pixel points around it to ensure that the segmentation boundary can be accurately determined.

[0061] The physical connection paths in the physical topology model constructed in S2 are used as the flow constraint conditions of the watershed algorithm: The traditional watershed algorithm allows water to diffuse to any adjacent pixel, while the improved algorithm strictly restricts the water flow to only flow along the physical connection paths. In specific implementation, a list of allowed flow neighborhoods is established for each pixel point, and this list is determined by the physical connection paths determined in S2, rather than simply using the geometric 8-neighborhood or 4-neighborhood. For different types of liquid crystal displays, the physical connection paths have different characteristics: In traditional TFT-LCDs, the physical connections mainly run along the gate lines and data lines; OLED displays may have a more complex connection topology. In the algorithm implementation, the flow probability between adjacent pixels without physical connections is set to zero to ensure that the water flow does not cross the physically isolated areas.

[0062] The gradient value of the watershed algorithm is set according to the gradient weight matrix containing physical constraints generated in S2: The traditional watershed algorithm uses the image gray-scale gradient as the terrain height, while the improved algorithm uses the gradient weight matrix containing physical constraints as the terrain height map. This gradient weight matrix has already incorporated the physical structure information of the liquid crystal display, making the gradient values of the areas that should be physically connected lower, and the gradient values of the areas that should be physically separated higher. When specifically used, the gradient weight matrix is converted into a terrain height map, where the areas with low gradient values correspond to the terrain lows (reservoirs), and the areas with high gradient values correspond to the terrain highs (ridges). In actual implementation, usually the gradient weight matrix is normalized to ensure that the value range is within [0, 1] or [0, 255] for subsequent processing.

[0063] Set the initial water level line and use the pixel point with the lowest gradient value as the initial immersion point: Within the area of each bad pixel location to be processed, scan the gradient weight matrix containing physical constraints to find the set of pixel points with the lowest gradient values. These lowest points are used as the initial immersion points, and usually a unique area label is assigned to each initial immersion point as the seed point for the watershed algorithm. If multiple significantly separated gradient lowest points are detected within a single bad pixel area, they may correspond to multiple bad pixel sources, each serving as an independent initial immersion point. Initialize the priority queue data structure, sort all the initial immersion points by their gradient values and enqueue them to prepare for the subsequent immersion process.

[0064] During the immersion process, control the water flow diffusion path according to the pixel connection weight matrix: For a path with a connection weight of N1 (physical distance less than the threshold and sharing a driving line), set the diffusion priority to P1 (for example, with a value of 9); for a path with a connection weight of N2 (physical distance less than the threshold but not sharing a driving line), set the diffusion priority to P2 (for example, with a value of 3); for a path with a connection weight of N3 (physical distance exceeding the threshold), set the diffusion priority to P3 (for example, with a value of 1); satisfying the relationship P1 > P2 > P3 to ensure that the algorithm preferentially diffuses along the physical connection path sharing the driving line. In implementation, modify the diffusion rule of the traditional watershed algorithm, sort the candidate diffusion pixels by the diffusion priority, and process the diffusion pixels of the high-priority path first; the specific diffusion process uses the priority queue data structure, and the priority of the elements in the queue is jointly determined by the pixel gradient value and the path diffusion priority. The calculation formula for the priority is: Diffusion priority = Pixel gradient value / Path priority coefficient, so that the diffusion direction with a high path priority corresponds to a lower comprehensive priority value and will be processed first.

[0065] Predict the potential spread direction of bad pixels using the driving line fault model constructed in S2, and adjust the region growing strategy of the watershed algorithm: Analyze the abnormal pixel distribution pattern within the bad pixel location area and identify pixel anomalies distributed along specific driving lines. Specifically, when implementing, count the number of pixels overlapping with each driving line in the bad pixel area. When the proportion of abnormal pixels on a certain driving line exceeds a preset threshold (usually 40%), it is determined as an anomaly distributed along that driving line. When pixel anomalies distributed along a specific driving line are detected, in the gradient weight matrix with physical constraints, reduce the gradient value weights between other pixel points on that driving line. Usually, the reduction coefficient is 30% - 50% of the original value. This will reduce the possibility of forming a watershed line between pixel points on that driving line. According to the physical routing structure of the driving line, calculate the diffusion priority coefficient in the driving line direction. The specific method is to define a vector along the driving line direction and calculate the angle between the current diffusion direction and the driving line direction. The smaller the angle, the higher the priority coefficient. Apply the calculated diffusion priority coefficient to the immersion process of the watershed algorithm, multiplying it by the original path priority value to further increase the diffusion priority in the driving line direction. This dual adjustment of gradient value weights and diffusion priority makes the watershed algorithm tend to classify abnormal pixels on the same driving line into the same region when forming the region boundary, avoiding over-segmentation in terms of technical principles. Apply the adjusted region growing strategy to the entire immersion process of the watershed algorithm to continuously optimize the finally formed segmentation result.

[0066] When the water level rises to the preset height, detect the minimum pixel distance between adjacent sub-regions and decide whether to establish a watershed line based on the minimum pixel distance and physical connection relationship: Calculate the minimum pixel distance between each sub-region formed during the immersion process. Usually, use the Euclidean distance between the two closest pixel points in the two regions as the metric. When the minimum pixel distance between adjacent sub-regions is less than the preset threshold (usually 3 - 5 pixels), further analyze their physical connection relationship to prevent over-segmentation. Obtain the position information of the pixel points of adjacent sub-regions in the two-dimensional pixel coordinate system to determine their accurate positions in physical space. According to the position information, query the driving line layout data to obtain the connection relationship between the pixel points in adjacent sub-regions and the gate lines and data lines, and judge whether they share the same driving line. At the same time, according to the position information and driving line connection relationship, query the physical connection path data to judge whether there is a direct physical path between adjacent sub-regions. When adjacent sub-regions share the same driving line and there is a direct physical path in the physical connection path, do not establish a watershed line, but regard them as different parts of the same bad pixel area. This is the key mechanism to prevent over-segmentation. When adjacent sub-regions do not share the same driving line or there is no direct physical path in the physical connection path, establish a watershed line to divide them into different bad pixel areas.

[0067] According to the established watershed line, perform continuity analysis on the pixel point groups and merge them to form the final segmentation result: For the identified pixel point groups sharing the same driving line, analyze their spatial continuity in the two-dimensional pixel coordinate system. Pixel groups that are continuously distributed and not interrupted by the watershed line are assigned the same region label, indicating that they belong to the same defective pixel region. Continuity analysis usually uses the connected component labeling algorithm, but the constraints of the physical connection path need to be considered. Only pixels that are physically connected are considered continuous. Merge the pixel groups with the same region label to form multiple independent defective pixel sub-regions, and each sub-region corresponds to a possible defective pixel or a group of related defective pixels. For each formed independent defective pixel sub-region, record its boundary coordinates, area, shape characteristics, and the driving line information involved, providing a basis for subsequent defective pixel classification and cause analysis. The finally output segmentation result is presented in the form of a region label map, where different region label values represent different defective pixel sub-regions, and the region with a value of 0 represents the non-defective pixel region.

[0068] S4. Use the pre-constructed defective pixel feature knowledge base to perform feature analysis on the segmentation result, determine the defective pixel name and the cause of the defective pixel, and obtain the detection result, including: The system receives the segmented result image with region labels output by the S3 watershed algorithm, where each independent defective pixel sub-region has a unique label value. Perform preprocessing on the received segmentation result, including noise filtering and boundary smoothing. Usually, small regions with an area smaller than a threshold (such as 3 - 5 pixels) are filtered out to eliminate possible false detections. Establish a data structure for each independent defective pixel sub-region to record its basic information such as the label value, the list of included pixel coordinates, and the list of boundary pixel coordinates, etc., for subsequent analysis. Establish an index table for the defective pixel sub-regions to facilitate quick positioning and retrieval of information for specific regions during subsequent analysis.

[0069] Calculate the geometric feature parameters for each region based on the independent defective pixel sub-regions: Region area: Calculate the total number of pixels included in each defective pixel sub-region. For high-resolution displays, the number of pixels is usually converted to physical area (square millimeters). Perimeter: Calculate the number of boundary pixels of the defective pixel sub-region. A boundary pixel is defined as a pixel point where at least one adjacent pixel does not belong to this region. For a more accurate perimeter calculation, usually, a boundary tracing algorithm based on 8-connectivity is used. Aspect ratio: Determine the minimum bounding rectangle of the defective pixel sub-region and calculate the ratio of its major axis to its minor axis. The aspect ratio is an important parameter for distinguishing dot-like and line-like defective pixels. Usually, the aspect ratio of line-like defective pixels > 3.

[0070] Shape complexity: Calculate the shape complexity index. Common methods include: Circularity: 4π×area / perimeter², a perfect circle is 1, and the more irregular the shape, the smaller the value; Rectangularity: Region area / minimum bounding rectangle area, a perfect rectangle is 1; Fractal dimension: Calculate through the relationship between the boundary length and the measurement scale change.

[0071] Distribution density: The distribution density of pixel points within the analysis area, including: Pixel density: The number of defective pixels within the area / the area of the minimum circumscribed rectangle; Density gradient: The change rate of pixel density from the center to the edge of the area; Void ratio: The number of non-defective pixels within the area / the total area of the area.

[0072] According to the two-dimensional pixel coordinate system, analyze the position characteristics of each defective sub-region on the liquid crystal screen: Calculate the centroid coordinates of the defective sub-region, that is, the average value of all pixel coordinates within the region, as the representative point of the region position. Calculate the distances of the defective sub-region from the four sides of the screen to determine whether it belongs to edge defective pixels (usually the distance is less than 5% of the screen size). Edge defective pixels are usually related to problems such as frame pressure and poor soldering. Analyze the distance and direction of the defective pixels relative to the position of the driving IC, which is particularly important for judging defective pixels caused by driving IC failures.

[0073] Calculate the relative position relationship between the current defective sub-region and other defective regions, including: Nearest defective pixel distance: The minimum distance between the current region and the nearest other defective region; Spatial distribution pattern: Whether multiple defective regions exhibit a specific geometric arrangement (such as linear arrangement, grid arrangement); Direction consistency: Whether multiple defective pixels are arranged in the same direction; Record the distribution of defective pixels in different functional regions of the display screen (such as menu area, signal line area, main display area), which is of great significance for the judgment of specific types of defective pixels.

[0074] According to the driving line layout data, analyze the correlation between the defective sub-region and the driving lines: Check whether the defective sub-region overlaps with specific gate lines or data lines, and calculate the overlap degree (the number of overlapping pixels / the total number of pixels in the region). When the overlap degree exceeds the preset threshold (usually 60%), it is determined that the defective pixel is highly correlated with the driving line. Analyze the ductility of the defective region in the direction of the driving line, and calculate the proportion of the projection length of the defective region in the direction of the driving line to the total length of the line. Detect whether the defective region spans multiple parallel driving lines. If so, it may be related to a failure at the intersection of the driving lines. Analyze the distribution pattern of multiple separated defective regions on the same driving line to judge whether there are regular failures caused by signal transmission attenuation of the driving signal. According to the transmission direction of the driving signal, analyze the distribution characteristics of the defective region in the signal flow direction, which is particularly important for judging signal transmission failures. Record the identification, type, and physical characteristics of the driving lines related to the defective pixels to provide a basis for subsequent cause analysis.

[0075] Classify the defective pixels based on geometric feature parameters, position features, and the relevance of drive lines: Criteria for judging dot-like defective pixels: The defective pixel area consists of N adjacent pixels (usually N < 10); the aspect ratio is close to 1 (usually < 1.5); the shape complexity index is close to a circle or a square; the pixel distribution within the area is uniform without obvious directionality; typical dot-like defective pixels include single bright / dark dots, stuck pixels (fixed-color dots), etc.

[0076] Criteria for judging line-like defective pixels: The defective pixel area extends significantly along the drive line direction; the aspect ratio is large (usually > 3); the overlap with a specific drive line is high (> 60%); the extensibility of the area in one direction is much greater than that in other directions; typical line-like defective pixels include gate line open circuits, data line short circuits, abnormal drive line signal transmission, etc.

[0077] Criteria for judging block-like defective pixels: The defective pixel area is a set of pixels distributed in a rectangular or irregular area; the area is large (usually > 100 pixels); it may affect multiple drive lines simultaneously.

[0078] Further classify according to the shape: Rectangular block: High rectangularity (> 0.8), regular boundary; Irregular block: Complex shape, irregular boundary; Typical block-like defective pixels include TFT array defects, backlight leakage, poor liquid crystal injection, etc.; For composite defective pixels (defective pixels with multiple characteristics), use comprehensive multi-characteristic judgment to determine their main type and secondary type.

[0079] Match the classification results with the characteristic parameters in the defective pixel feature knowledge base: The defective pixel feature knowledge base is a pre-constructed database containing characteristic parameters, cause analysis, and identification rules for various types of defective pixels. This knowledge base is usually constructed based on a large number of historical defective pixel cases and expert experience.

[0080] For the combination of characteristic parameters of each defective pixel sub-region, perform multi-dimensional feature matching in the knowledge base. Commonly used matching algorithms include: Nearest neighbor matching based on distance: Calculate the Euclidean distance between the current defective pixel feature vector and the feature vectors of various types of defective pixels in the knowledge base, and select the type with the smallest distance; Fuzzy matching based on similarity: Use metrics such as cosine similarity to calculate the similarity degree of feature vectors; Expert system matching based on rules: Use predefined IF-THEN rules for judgment; The matching process usually adopts a weighted method, assigning different weights to different characteristic parameters to reflect their importance in defective pixel type judgment. Evaluate the confidence level of the matching result. When the confidence level is lower than the threshold, manual confirmation or further analysis may be required. Based on the matching result, determine the specific type and name of the defective pixel, such as "single-pixel dark dot", "gate line open circuit", "liquid crystal leakage block", etc.

[0081] According to the identified types and distribution characteristics of bad pixels, combined with the cause analysis rules in the knowledge base, infer the possible inducing factors for the formation of bad pixels: For each type of bad pixel, a list of possible inducing factors and their associated characteristic patterns are predefined in the knowledge base. By comparing the characteristic pattern of the current bad pixel with the characteristic pattern associated with the inducing factor, calculate the probability scores of each possible inducing factor. Multiple factors are considered during the inducing factor analysis, including: the type and location characteristics of the bad pixel; the relevance between the bad pixel and the driving circuit; the spatial relationship between multiple bad pixels; the change pattern of pixel values within the bad pixel area; the main inducing factors of similar bad pixels in historical statistical data.

[0082] Common inducing factors for bad pixels include: manufacturing process defects, such as TFT process defects, mask alignment errors, lithography defects, etc.; material problems, such as liquid crystal material impurities, polarizer defects, color filter defects, etc.; electrical failures, such as damaged driving ICs, open or short circuits in gate lines / data lines, etc.; mechanical damages, such as external force extrusion, panel bending, thermal stress, etc.; aging degradation, such as backlight attenuation, liquid crystal molecule orientation failure, etc.; Generate an inducing factor analysis report, list the possible inducing factors and their probability rankings, and provide a reference for improving the production process.

Claims

1. A method for detecting dead pixels of a liquid crystal display screen, characterized in that, Including: S1. Obtain the defective pixel position area of the liquid crystal display screen according to a preset test scheme; S2. Construct a physical topology model of the liquid crystal display screen according to the physical parameters of the liquid crystal display screen, including: obtaining the RGB pixel arrangement mode of the liquid crystal display screen, determining the physical connection relationship between adjacent pixels, and the positional relationship between adjacent pixels; establishing a two-dimensional pixel coordinate system according to the physical connection relationship between adjacent pixels and the positional relationship between adjacent pixels; determining the physical connection path at the sub-pixel level according to the physical boundaries of the sub-pixels of the liquid crystal display screen; obtaining the driving circuit layout data of the liquid crystal display screen, determining the connection relationship between pixel points and gate lines and data lines according to the physical routing structure of the driving circuit, and identifying pixel point groups sharing the same driving circuit by tracing the driving signal transmission path; constructing a pixel connection weight matrix according to the RGB pixel arrangement mode, sub-pixels and driving circuit layout data; generating a gradient weight matrix containing physical constraints according to the pixel connection weight matrix and the standard gradient value matrix; establishing a driving line fault model according to the driving circuit layout data through the transmission characteristics and attenuation rules of driving signals on the lines; constructing a physical topology model of the liquid crystal display screen according to the two-dimensional pixel coordinate system, the physical connection path at the sub-pixel level, the pixel connection weight matrix and the driving line fault model; S3. Use an improved watershed algorithm to segment the defective pixel position area to obtain a segmentation result, including: using the defective pixel position area determined in S1 as the input area of the watershed algorithm; using the physical connection path in the physical topology model constructed in S2 as the flow constraint condition of the watershed algorithm; setting the gradient value of the watershed algorithm according to the gradient weight matrix containing physical constraints generated in S2; setting an initial water level line and using the pixel point with the lowest gradient value as the initial immersion point; controlling the water flow diffusion path according to the pixel connection weight matrix during the immersion process; predicting the potential defective pixel diffusion direction using the driving line fault model in S2 and adjusting the region growth strategy of the watershed algorithm; when the water level rises to a preset height, detecting the minimum pixel distance between adjacent sub-regions, and establishing a watershed line according to the minimum pixel distance; analyzing the corresponding continuity in the two-dimensional pixel coordinate system for the identified pixel point groups sharing the same driving circuit according to the established watershed line, and assigning the same region label to the pixel groups that are continuously distributed and not separated by the watershed line; merging the pixel groups with the same region label to form multiple independent defective pixel sub-regions as the segmentation result; S4. Perform feature analysis on the segmentation result using a pre-constructed defective pixel feature knowledge base to determine the defective pixel name and the cause of the defective pixel to obtain a detection result.

2. The method for detecting defective pixels of a liquid crystal display screen according to claim 1, wherein: The defective pixel feature knowledge base includes the feature parameters, cause analysis and identification rules of various types of defective pixels.

3. The method for detecting defective pixels of a liquid crystal display screen according to claim 2, wherein: S1. Obtain the defective pixel position area of the liquid crystal display screen according to a preset test scheme, including: Determine the standard gradient values corresponding to each pixel in the liquid crystal display according to a preset test scheme, and generate a standard gradient value matrix; the standard gradient value represents the gray gradient value corresponding to the pixel under preset normal working conditions; Obtain a test image of the liquid crystal display under test conditions; Construct an actual gradient value matrix based on the test image; Determine the defective pixel position area of the liquid crystal display by comparing the standard gradient value matrix and the actual gradient value matrix.

4. The method for detecting defective pixels of a liquid crystal display according to claim 1, wherein: Construct a pixel connection weight matrix, including: Obtain the physical distance between pixels and the sharing situation of driving lines; Set the connection weight of pixels with a physical distance less than the threshold and sharing the driving line to N1; Set the connection weight of pixels with a physical distance less than the threshold but not sharing the driving line to N2; Set the connection weight of pixels with a physical distance exceeding the threshold to N3.

5. The method for detecting defective pixels of a liquid crystal display according to claim 4, wherein: N1 is greater than N2, and N2 is greater than N3.

6. The method for detecting defective pixels of a liquid crystal display according to claim 4, wherein: Set the path diffusion priority of the path with a connection weight of N1 to P1, set the path diffusion priority of the path with a connection weight of N2 to P2, and set the path diffusion priority of the path with a connection weight of N3 to P3.

7. The method for detecting defective pixels of a liquid crystal display according to claim 6, wherein: P1 is greater than P2, and P2 is greater than P3.

8. The method for detecting defective pixels of a liquid crystal display according to claim 6, wherein: When the water level rises to a preset height, detect the minimum pixel distance between adjacent sub-regions, and establish a watershed line according to the minimum pixel distance, including: Calculate the minimum pixel distance between the sub-regions formed during the immersion process; When the minimum pixel distance between adjacent sub-regions is less than the threshold, obtain the position information of the pixel points of the adjacent sub-regions in the two-dimensional pixel coordinate system constructed in S2; According to the position information, obtain the connection relationship between the pixel points of the adjacent sub-regions and the gate lines and data lines in the driving line layout data in S2; According to the position information and the connection relationship between the pixel points and the gate lines and data lines, obtain the physical connection path of the pixel points of the adjacent sub-regions determined in S2, and the physical connection path is used to determine whether there is a direct physical path between the adjacent sub-regions; When the adjacent sub-regions share the same driving line and there is a direct physical path on the physical connection path, do not establish a watershed line; When the adjacent sub-regions do not share the same driving line or there is no direct physical path on the physical connection path, establish a watershed line.

9. A defective pixel detection system for a liquid crystal display, characterized in that Comprising: At least one processing unit; configured to execute instructions to implement the method for detecting defective pixels of a liquid crystal display according to any one of claims 1 to 8.

Images