Internal bubble detection method for polaroid production based on image processing

Abstract

The invention discloses an internal bubble detection method for polaroid production based on image processing, and particularly relates to the technical field of polaroid detection. According to the method, transmission, low-angle light and multi-polarization-angle polarization images are collected, multi-dimensional fusion features such as circularity and gray scale comparison are extracted, category subordinate degree judgment is carried out in combination with a defect sample set, similar defects such as internal bubbles and surface dust and scratches are effectively distinguished, and misjudgment and missed judgment are reduced; meanwhile, the positions of the bubbles are accurately positioned through processing such as local threshold segmentation and closed operation, and a reliable foundation is laid for subsequent evaluation and tracing.

Description

Technical Field

[0001] This invention relates to the field of polarizer inspection technology, and more specifically, to a method for detecting internal bubbles in polarizer production based on image processing. Background Technology

[0002] As a core component of LCD panels, polarizers directly determine the display effect and quality stability of terminal display products. During the production process, internal bubble defects are easily generated due to factors such as temperature control deviations. These defects will seriously affect the optical performance of polarizers, leading to problems such as bright spots and dark spots in terminal display products, and reducing the product qualification rate.

[0003] Currently, online detection methods based on image processing have gradually replaced manual visual inspection, but the following technical bottlenecks still exist:

[0004] On the one hand, the detection accuracy of a single type of image is limited, making it difficult to effectively distinguish internal bubbles from similar defects such as surface dust, scratches, and crystal points, which can easily lead to misjudgment or missed judgment.

[0005] On the other hand, the detection process can only identify and locate defects, but cannot establish a correlation between defects and production process parameters. It is difficult to trace the source of bubble defects, which means that hidden processes in the production process cannot be discovered and adjusted in a timely manner, and the defect incidence rate cannot be fundamentally reduced. The level of intelligence in detection is low.

[0006] To address this, an internal bubble detection method based on image processing for polarizer production was developed. Summary of the Invention

[0007] To overcome the above-mentioned deficiencies of the prior art, embodiments of the present invention provide an internal bubble detection method for polarizer production based on image processing.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] An image processing-based method for detecting internal bubbles in polarizer production includes:

[0010] Collect multiple feature images of the same polarizer region to form a feature image set;

[0011] The feature image set includes transmission images, low-angle light images, and polarization images; the polarization images contain images acquired at different polarization angles.

[0012] The feature image set is analyzed, and after locating the suspicious areas, the multi-dimensional fused feature vector of each suspicious area is extracted. Based on the fused feature vector, the internal bubble defects are identified and their locations are determined.

[0013] Based on the determination of defect types in different suspicious areas, the areas of suspicious areas belonging to the same defect type are accumulated, and then weighted and fused according to the defect weights set for different defect types to output the defect evaluation coefficient of the polarizer.

[0014] The process involves determining the physical coordinates of the start and end points of each process segment along the length of the polarizer. It also defines the intervals between the start and end points of different process segments in the coordinate system as the process influence zone, and then dividing the internal bubble defects into process influence zones. Based on these divisions, the process with potential problems is identified. The process parameters of the potential problems within the current production time period of the polarizer are extracted and analyzed for correlation. The severity of the potential problems is determined and sent to management personnel.

[0015] Specifically, The logic for locating suspicious areas in the steps;

[0016] In the transmission image, obtain the local threshold of the image. ;

[0017] According to the threshold Generate a binary mask image ;

[0018] Represented as: ; This represents the grayscale value at coordinates (x, y);

[0019] right After performing a erosion-dilation closing operation, connected components are marked, and the bounding rectangle of each connected component is extracted as a suspicious region.

[0020] Specifically, The process involves identifying and locating internal bubble defects.

[0021] Extract multi-dimensional fused feature vectors for each suspicious region, including transmission features, low-angle light features, and polarization features;

[0022] For each suspicious region, the transmission characteristics, low-angle light characteristics, and polarization characteristics are combined with the fusion feature vectors of various pre-constructed defect sample sets to determine the category affiliation between each suspicious region and various defect sample sets.

[0023] The defect of the corresponding type with the lowest category affiliation is identified as the defect in the suspicious area.

[0024] The suspicious areas that are determined to be internal bubble defects are extracted, and their locations are obtained based on the pixel information of the transmission image.

[0025] Specifically, the calculation logic for transmission characteristics, low-angle light characteristics, and polarization characteristics;

[0026] Transmission characteristics include roundness and grayscale contrast;

[0027] Gray-scale contrast is obtained by calculating the absolute difference between the average gray-scale value of pixels in the suspicious area and the average gray-scale value of pixels in the background area;

[0028] Low-angle light characteristics are represented by the contrast ratio between bright and dark fields;

[0029] The absolute difference between the mean gray level of the suspicious area in the low-angle light image and the mean gray level of the background area is calculated. The ratio is calculated by using the absolute difference as the numerator and the gray level contrast of the transmitted image as the denominator, and the contrast ratio of the bright and dark fields is obtained.

[0030] The polarization characteristic is curve smoothness;

[0031] For different polarization angles Calculate the average gray level of the suspicious area at each angle, and after normalization, fit the curve. Calculate the root mean square of the second derivative of the fitted curve as the curve smoothness.

[0032] Specifically, The steps involve identifying potential hazards in the process.

[0033] The process-affected zone includes the coating process-affected zone, the stretching process-affected zone, and the curing process-affected zone.

[0034] The number of bubble defect pixels in different process influence areas is counted, and the actual area is converted based on the image resolution. The process influence area with the highest bubble defect area is selected and its process type is identified as the potential process.

[0035] The potential hazards in the process include coating hazards, stretching hazards, and curing hazards, which correspond to the coating process influence zone, stretching process influence zone, and curing process influence zone, respectively.

[0036] Specifically, The parameters for the association analysis in the steps specifically include:

[0037] The process parameters associated with potential coating defects include slurry viscosity, doctor blade pressure, and coating speed.

[0038] The process parameters associated with potential stretching hazards include stretching zone temperature, stretching ratio, and stretching speed.

[0039] The process parameters associated with potential curing hazards include curing oven temperature and curing oven air velocity.

[0040] Specifically, If the potential hazard is in the coating process, then the degree of hazard in the potential hazard process is determined.

[0041] Extract the slurry viscosity, doctor blade pressure and coating speed at each time point during the coating process of the polarizer, calculate the average slurry viscosity, average doctor blade pressure and average coating speed.

[0042] After comprehensively processing the average viscosity of the slurry, the average pressure of the doctor blade, and the average coating speed in conjunction with the coating process standards, the coating process hazard coefficient is output.

[0043] If the coating process hazard coefficient is higher than the set coating threshold hazard coefficient, calculate the difference between the coating process hazard coefficient and the coating threshold hazard coefficient. Use the difference as the numerator and the coating threshold hazard coefficient as the denominator to calculate the ratio as the degree of hazard.

[0044] Specifically, If the potential hazard in the process is a tensile hazard, then the degree of hazard in the process is determined.

[0045] Extract the stretching zone temperature of the polarizer during the stretching time period; the stretching zone includes the inlet, middle section and outlet.

[0046] For the inlet zone temperature, middle zone temperature, and outlet zone temperature at each time point within the stretching period, calculate the average values ​​and output the average inlet temperature, average middle zone temperature, and average outlet temperature; after calculating the standard deviation of the average inlet temperature, average middle zone temperature, and average outlet temperature, output the interval temperature difference.

[0047] The stretching process risk coefficient is output by combining the temperature difference within the stretching time range with the stretching ratio and stretching speed, and then combining the stretching process standard.

[0048] If the hazard coefficient of the stretching process is higher than the set hazard coefficient of the stretching threshold, the difference between the hazard coefficient of the stretching process and the hazard coefficient of the stretching threshold is calculated. The difference is used as the numerator and the hazard coefficient of the stretching threshold is used as the denominator to calculate the ratio as the degree of hazard.

[0049] Specifically, If the potential hazard is a solidified hazard during the process, then the degree of hazard of the potential hazard is determined.

[0050] Extract the temperature and wind speed of the polarizer at each time point during the curing time in the curing oven, calculate the average value of each time point, and output the average temperature and average wind speed in the oven.

[0051] After comprehensively processing the average temperature and average wind speed inside the furnace in conjunction with the furnace curing standard, the curing process hazard coefficient is output.

[0052] If the hazard coefficient of the curing process is higher than the hazard coefficient of the set curing threshold, the difference between the hazard coefficient of the curing process and the hazard coefficient of the curing threshold is calculated. The difference is used as the numerator and the hazard coefficient of the curing threshold is used as the denominator to calculate the ratio as the degree of hazard.

[0053] The technical effects and advantages of this invention are as follows:

[0054] (1) By collecting transmitted, low-angle light and multi-polarization images, multi-dimensional fusion features such as circularity and grayscale contrast are extracted. Combined with the defect sample set, the category degree is determined, which effectively distinguishes internal bubbles from similar defects such as surface dust and scratches, reducing misjudgment and missed judgment. At the same time, the bubble position is accurately located through local threshold segmentation, closing operation and other processing, laying a reliable foundation for subsequent evaluation and traceability.

[0055] (2) By dividing the affected areas of coating, stretching and curing processes, and linking the corresponding process parameters to conduct hidden danger analysis, the degree of hidden danger is quantified and fed back to the management personnel. This not only can quickly locate the source of the bubble, but also provide data support for the adjustment of process parameters, helping enterprises to optimize the production process in a timely manner, reduce the occurrence rate of bubble defects from the root, and improve the product qualification rate. Attached Figure Description

[0056] Figure 1 This is a flowchart of the internal bubble detection method for polarizer production based on image processing according to the present invention. Detailed Implementation

[0057] 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 embodiments of the present invention, and not all embodiments. 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.

[0058] like Figure 1 As shown, the method for detecting internal bubbles in polarizer production based on image processing includes:

[0059] Image acquisition and processing: When the polarizer passes through the inspection station, multiple feature images of the same polarizer area are acquired as a feature image set;

[0060] The feature image set includes transmission images, low-angle light images, and polarization images; the polarization images contain images acquired at different polarization angles.

[0061] The feature images are constructed by sequentially triggering different lighting conditions; the lighting conditions include vertical transmission lighting, low-angle ring lighting, and polarized lighting.

[0062] Intelligent defect identification and processing: Based on the pre-built defect classification logic, the feature image set is analyzed, and after locating the suspicious area, the multi-dimensional fusion feature vector of each suspicious area is extracted. Based on the fusion feature vector, the internal bubble defect is identified and its location is determined.

[0063] Specifically:

[0064] Gaussian filtering is applied to the transmission image to reduce noise.

[0065] In the transmission image, obtain the local threshold of the image. ;

[0066] This threshold is a dynamically changing value. For each pixel in the image or its local neighborhood, all gray values ​​below a certain threshold are considered acceptable. The pixels will be initially identified as potential defects and thus marked as foreground in the subsequent binarization step.

[0067] Based on local mean and variance: ;in Indicates The arithmetic mean of the gray values ​​of all pixels in the local neighborhood centered on the center (implemented by defining a sliding window of size N×N, such as 15×15, which reflects the overall brightness level of the local area).

[0068] Represented by pixels The standard deviation of all pixel gray values ​​within a local neighborhood centered on the center (quantifies the dispersion or "noise" level of gray values ​​in that local area).

[0069] This is a preset scaling factor; the range is limited to 0.5-3.0; (set by technical personnel to determine the leniency of the segmentation conditions).

[0070] According to the threshold Generate a binary mask image ;

[0071] Represented as: ; This represents the grayscale value at coordinates (x, y);

[0072] Pixels assigned a value of 1 constitute the foreground region, representing potential defect points whose grayscale values ​​are significantly lower than the local adaptive threshold; pixels assigned a value of 0 constitute the background region.

[0073] right A erosion-dilation closing operation is performed to smooth the boundaries and fill small holes; then, after connecting component labeling, the bounding rectangle of each connected component is extracted as a suspicious region.

[0074] For each suspicious region's pixel coordinate set Using a pre-calibrated affine transformation matrix, its coordinates are mapped onto the low-angle light image and polarization image, and completely corresponding image patches are extracted, including the transmission image patch, the low-angle light image patch, and the polarization image patch, denoted as . ;

[0075] Supplementary explanation, among which This represents an image patch extracted from a transmission image;

[0076] This represents an image patch extracted from a low-angle light image;

[0077] This represents image patches extracted from polarized images at different polarization angles.

[0078] i is the number of the suspicious area.

[0079] The multi-dimensional fused feature vector includes transmission features, low-angle light features, and polarization features;

[0080] Transmission characteristics include roundness and grayscale contrast;

[0081] Circularity is determined by the formula The calculation yielded, where Indicates area, Indicates the perimeter.

[0082] Gray-scale contrast is obtained by calculating the absolute difference between the average gray-scale value of pixels in the suspicious area and the average gray-scale value of pixels in the background area;

[0083] Low-angle light characteristics are represented by the contrast ratio between bright and dark fields;

[0084] The absolute difference between the mean gray level of the suspicious area in the low-angle light image and the mean gray level of the background area is calculated. The ratio is calculated by using the absolute difference as the numerator and the gray level contrast of the transmitted image as the denominator, and the contrast ratio of the bright and dark fields is obtained.

[0085] Contrast ratio between light and dark scenes: ;in It represents the absolute difference between the mean gray level in a low-angle light image and the mean gray level of the background area; Indicates the grayscale contrast of a transmitted image;

[0086] For real bubbles, this ratio is usually >1 due to edge refraction; for surface dust, this ratio tends to be close to 1 or <1.

[0087] The polarization characteristic is curve smoothness;

[0088] For different polarization angles Calculate the average gray level of the suspicious area at each angle, and then fit the curve after normalization. Where A is the modulation depth. B is the phase angle, and B is the background intensity. The average gray level is normalized and linearized using trigonometric identities; the root mean square of the second derivative of the fitted curve is calculated as the curve smoothness.

[0089] For roundness, grayscale contrast, light and dark field contrast ratio, and curve smoothness;

[0090] For each suspicious region, the circularity, grayscale contrast, light-dark field contrast ratio, and curve smoothness are comprehensively processed in conjunction with the fusion feature vectors of various pre-constructed defect sample sets to determine the category affiliation between each suspicious region and various defect sample sets.

[0091] Collect a large number of defect samples of known categories (labeled as "internal bubbles", "surface dust", "scratches", "crystal points", etc.) and extract their fused feature vectors to form a sample set.

[0092] The specific process of comprehensive processing is as follows:

[0093] The circularity, grayscale contrast, light-dark contrast ratio, and curve smoothness of each suspicious area were marked as follows: ;

[0094] Various defect sample sets are fused with feature vectors that include roundness, grayscale contrast, brightness-to-dark field contrast ratio, and curve smoothness. express;

[0095] After normalization, the formula is used. The output, after weighted calculation, shows the category affiliation between each suspicious region and various defect sample sets. ;in The weighting coefficients are set, and their sum is one.

[0096] The defect of the corresponding type with the lowest category affiliation is identified as the defect in the suspicious area.

[0097] The suspicious areas that are determined to be internal bubble defects are extracted, and their positions are obtained based on the pixel information of the transmission image;

[0098] That is, the region confidence coordinates And convert them into physical coordinates in the polarizer width and travel direction.

[0099] Defect comprehensive assessment and processing: Based on the determination of defect types in different suspicious areas, the areas of suspicious areas belonging to the same defect type are accumulated, and the defect weights set for different defect types are comprehensively processed to output the defect assessment coefficient of the polarizer.

[0100] That is, using formulas The defect assessment coefficient is calculated. ;in This represents the number of different defect types, where N is the total number of defect types. The result after accumulating and normalizing the additive value of suspicious areas for different defect types. Defect weights are assigned to different defect types.

[0101] Process parameter correlation analysis: Determine the physical coordinates of the start and end points of each process segment along the length direction of the polarizer (i.e., machine direction MD, Y-axis). Define the intervals between the start and end points of different process segments in the coordinate system as the process influence zone. Divide the process influence zone for the location of internal bubble defects (if a bubble location is within the interval between the start and end points of the coating process influence zone, it is divided into the coating process influence zone). Based on the division results, identify potential processes, extract the process parameters of potential processes within the current production time period of the polarizer, perform correlation analysis, determine the degree of danger of potential processes, and send the results to management personnel.

[0102] The key functional sections and their sequence in the continuous polarizer production line are clearly defined. The production line includes: coating section → stretching section → curing section.

[0103] Physical markers are set on the production line or encoder information is used to determine the physical coordinates of the start and end points of each process segment along the length of the polarizer.

[0104] In the system's coordinate system, the interval from Y_start to Y_end corresponds to the "coating process influence zone", the interval from Y_stretch_start to Y_stretch_end corresponds to the "stretching process influence zone", and so on.

[0105] Specifically:

[0106] The process-affected zone includes the coating process-affected zone, the stretching process-affected zone, and the curing process-affected zone.

[0107] If a bubble is located in the coating process-affected area, it is classified as a coating section-related bubble.

[0108] If a bubble is located in the stretching process influence zone, it is classified as a stretching section associated bubble.

[0109] If a bubble is located in the curing process-affected area, it is classified as a curing-related bubble.

[0110] The number of bubble defect pixels in different process influence areas is counted, and the actual area is converted based on the image resolution. The process influence area with the highest bubble defect area is selected and its process type is identified as the potential process.

[0111] The hidden dangers include coating process hazards, tensile process hazards, and curing process hazards, which correspond to the coating process influence zone, tensile process influence zone, and curing process influence zone, respectively.

[0112] Extract the process parameters of the potentially problematic processes within the current production time period corresponding to the polarizer and perform correlation analysis to determine the degree of potential problems in the problematic processes;

[0113] The process parameters associated with potential coating defects include slurry viscosity, doctor blade pressure, and coating speed.

[0114] The process parameters associated with potential stretching hazards include stretching zone temperature, stretching ratio, and stretching speed.

[0115] The process parameters associated with potential curing hazards include curing oven temperature and curing oven air velocity;

[0116] If the potential problem lies in the coating process;

[0117] Extract the slurry viscosity, doctor blade pressure and coating speed at each time point during the coating process of the polarizer, calculate the average slurry viscosity, average doctor blade pressure and average coating speed.

[0118] High viscosity: The slurry has poor fluidity, and air is easily trapped in the gap between the doctor blade and the substrate. After coating, the air is trapped in the wet film and forms bubbles. Low viscosity: The slurry has excessive leveling properties, resulting in uneven thickness of the wet film in some areas. Differences in solvent evaporation rates lead to pinhole-like bubbles.

[0119] Excessive pressure: Slight deformation of the substrate surface causes turbulence during slurry extrusion, trapping air and forming bubbles; Insufficient pressure: Unstable slurry layer thickness control results in voids in the wet film, which shrink after drying, forming bubbles.

[0120] Too fast a speed: The relative shear force between the doctor blade and the slurry increases, increasing the probability of air entrapment; Too slow a speed: The wet film stays in the coating area for too long, the solvent evaporates prematurely, and after the surface cures, the internal solvent cannot be discharged, forming closed-cell bubbles.

[0121] After comprehensively processing the average viscosity of the slurry, the average pressure of the doctor blade, and the average coating speed in conjunction with the coating process standards, the coating process hazard coefficient is output.

[0122] The coating process standards include standard slurry viscosity, standard doctor blade pressure, and standard coating speed.

[0123] Using formula Calculate the risk factor of the coating process ;in , as well as These represent the average viscosity of the slurry, the average pressure of the doctor blade, and the average coating speed, respectively. , as well as These represent the standard slurry viscosity, standard doctor blade pressure, and standard coating speed, respectively. , as well as The weighting coefficients are set, and their sum is one.

[0124] If the coating process hazard coefficient is higher than the set coating threshold hazard coefficient, calculate the difference between the coating process hazard coefficient and the coating threshold hazard coefficient, and use the difference as the numerator and the coating threshold hazard coefficient as the denominator to calculate the ratio as the degree of hazard.

[0125] If the coating process hazard coefficient is lower than the set coating threshold hazard coefficient, the hazard level will be output directly as the preset value; it can be set to 0.

[0126] If the potential hazard is a tensile hazard;

[0127] Extract the temperature of the stretching zone of the polarizer during the stretching time period;

[0128] The stretching section includes the inlet, middle section, and outlet; that is, the inlet, middle section, and outlet of the stretching oven.

[0129] The average values ​​of the inlet zone temperature, middle zone temperature, and outlet zone temperature at each time point during the stretching period are calculated, and the average inlet temperature, average middle zone temperature, and average outlet temperature are output.

[0130] After calculating the standard deviation of the inlet temperature, the middle temperature, and the outlet temperature, the interval temperature difference is output.

[0131] Excessive temperature difference: Uneven heating of the substrate and differences in shrinkage rate during stretching lead to the formation of bubbles.

[0132] The stretching process risk coefficient is output by combining the temperature difference within the stretching time range with the stretching ratio and stretching speed, and then combining the stretching process standard.

[0133] Excessive magnification: The substrate is stretched too much, the internal polymer chain arrangement is destroyed, forming tiny voids, which expand into bubbles in subsequent processes;

[0134] Too low a magnification: The substrate does not meet the orientation requirements of the process, and the internal residual stress is unevenly distributed, which induces bubbles;

[0135] Excessive speed: Sudden change in tensile stress in the substrate, local areas are "thinned", and residual solvent areas rupture to form bubbles; Large speed fluctuations: The substrate vibrates, and the force is uneven during the stretching process, resulting in periodically distributed bubbles.

[0136] The stretching process standards include standard temperature difference range, standard stretching ratio, and standard stretching speed.

[0137] Using formula Calculate the potential risk factor of the stretching process. ;in , , These represent the temperature difference within the interval, the stretching ratio, and the stretching speed, respectively. , These represent the standard temperature range difference, standard elongation ratio, and standard elongation speed, respectively. , , The weighting coefficients are set, and their sum is one.

[0138] If the hazard coefficient of the stretching process is higher than the set hazard coefficient of the stretching threshold, the difference between the hazard coefficient of the stretching process and the hazard coefficient of the stretching threshold is calculated. The difference is used as the numerator and the hazard coefficient of the stretching threshold is used as the denominator to calculate the ratio as the degree of hazard.

[0139] If the risk factor of the stretching process is lower than the set risk factor of the stretching threshold, the risk level will be output directly as the preset value; it can be set to 0.

[0140] If the potential hazard is a solidified hazard;

[0141] Extract the temperature and wind speed of the polarizer at each time point during the curing time in the curing oven, calculate the average value of each time point, and output the average temperature and average wind speed in the oven.

[0142] Too low a temperature: The curing reaction is insufficient, the degree of cross-linking of molecules inside the substrate is insufficient, and residual monomers and solvents form bubbles; Too high a temperature: The surface of the substrate cures rapidly, the internal heat cannot be dissipated, and the solvent vaporizes and is trapped inside the substrate, forming closed-cell bubbles.

[0143] Too slow an airflow: uneven heat distribution inside the furnace, large differences in the degree of curing of the substrate, resulting in bubbles; Too fast an airflow: a sudden drop in the surface temperature of the substrate, a large temperature difference between the inside and outside, uneven shrinkage, forming tiny bubbles.

[0144] After comprehensively processing the average temperature and average wind speed inside the furnace in conjunction with the furnace curing standard, the curing process hazard coefficient is output.

[0145] The furnace curing standards include standard furnace temperature and standard furnace wind speed;

[0146] Using formula Calculate the potential risk factor of the curing process ;in , These represent the average temperature inside the furnace and the average wind speed inside the furnace, respectively. , These represent the standard furnace temperature and the standard furnace wind speed, respectively. , The weighting coefficients are set, and their sum is one.

[0147] If the hazard coefficient of the curing process is higher than the hazard coefficient of the set curing threshold, the difference between the hazard coefficient of the curing process and the hazard coefficient of the curing threshold is calculated. The difference is used as the numerator and the hazard coefficient of the curing threshold is used as the denominator to calculate the ratio as the degree of hazard.

[0148] If the hazard coefficient of the curing process is lower than the set curing threshold hazard coefficient, the hazard level will be output directly as the preset value; it can be set to 0.

[0149] Defect data traceability settings: Each inspected polarizer is bound to a unique traceability ID, and its defect identification and classification results, defect assessment coefficient, potential process that causes bubbles and the degree of associated potential problems, and the production batch to which it belongs are recorded, forming a complete quality data chain;

[0150] The above formulas are all dimensionless calculations. Dimensionless calculations can be performed using various methods such as standardization, which will not be elaborated here. The formulas are derived from software simulations based on a large amount of collected data, and the preset parameters in the formulas can be set by those skilled in the art according to the actual situation.

[0151] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, ATA hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state ATA hard disk.

[0152] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0153] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0154] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0155] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0156] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0157] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they 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 a portion 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 a computer 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 ATA hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0158] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for detecting internal bubbles in polarizer production based on image processing, characterized in that, include: Collect multiple feature images of the same polarizer region to form a feature image set; The feature image set includes transmission images, low-angle light images, and polarization images; the polarization images contain images acquired at different polarization angles. The feature image set is analyzed, and after locating the suspicious areas, the multi-dimensional fused feature vector of each suspicious area is extracted. Based on the fused feature vector, the internal bubble defects are identified and their locations are determined. Based on the determination of defect types in different suspicious areas, the areas of suspicious areas belonging to the same defect type are accumulated, and then weighted and fused according to the defect weights set for different defect types to output the defect evaluation coefficient of the polarizer. The process involves determining the physical coordinates of the start and end points of each process segment along the length of the polarizer. It also defines the intervals between the start and end points of different process segments in the coordinate system as the process influence zone, and then dividing the internal bubble defects into process influence zones. Based on these divisions, the process with potential problems is identified. The process parameters of the potential problems within the current production time period of the polarizer are extracted and analyzed for correlation. The severity of the potential problems is determined and sent to management personnel.

2. The method for detecting internal bubbles in polarizer production based on image processing according to claim 1, characterized in that: The logic for locating suspicious areas in the steps; In the transmission image, obtain the local threshold of the image. ; According to the threshold Generate a binary mask image ; Represented as: ; This represents the grayscale value at coordinates (x, y); right After performing a erosion-dilation closing operation, connected components are marked, and the bounding rectangle of each connected component is extracted as a suspicious region.

3. The method for detecting internal bubbles in polarizer production based on image processing according to claim 2, characterized in that: The process involves identifying and locating internal bubble defects. Extract multi-dimensional fused feature vectors for each suspicious region, including transmission features, low-angle light features, and polarization features; For each suspicious region, the transmission characteristics, low-angle light characteristics, and polarization characteristics are combined with the fusion feature vectors of various pre-constructed defect sample sets to determine the category affiliation between each suspicious region and various defect sample sets. The defect of the corresponding type with the lowest category affiliation is identified as the defect in the suspicious area. The suspicious areas that are determined to be internal bubble defects are extracted, and their locations are obtained based on the pixel information of the transmission image.

4. The method for detecting internal bubbles in polarizer production based on image processing according to claim 3, characterized in that: The calculation logic for transmission characteristics, low-angle light characteristics, and polarization characteristics; Transmission characteristics include roundness and grayscale contrast; Gray-scale contrast is obtained by calculating the absolute difference between the average gray-scale value of pixels in the suspicious area and the average gray-scale value of pixels in the background area; Low-angle light characteristics are represented by the contrast ratio between bright and dark areas; The absolute difference between the mean gray level of the suspicious area in the low-angle light image and the mean gray level of the background area is calculated. The ratio is calculated by using the absolute difference as the numerator and the gray level contrast of the transmitted image as the denominator, and the contrast ratio of the bright and dark fields is obtained. The polarization characteristic is curve smoothness; For different polarization angles Calculate the average gray level of the suspicious area at each angle, and after normalization, fit the curve. Calculate the root mean square of the second derivative of the fitted curve as the curve smoothness.

5. The method for detecting internal bubbles in polarizer production based on image processing according to claim 1, characterized in that: The process steps include identifying potential hazards in the manufacturing process; The process-affected zone includes the coating process-affected zone, the stretching process-affected zone, and the curing process-affected zone. The number of bubble defect pixels in different process influence areas is counted, and the actual area is converted based on the image resolution. The process influence area with the highest bubble defect area is selected and its process type is identified as the potential process. The potential hazards in the process include coating hazards, stretching hazards, and curing hazards, which correspond to the coating process influence zone, stretching process influence zone, and curing process influence zone, respectively.

6. The method for detecting internal bubbles in polarizer production based on image processing according to claim 5, characterized in that: The parameters for the association analysis in the steps specifically include: The process parameters associated with potential coating defects include slurry viscosity, doctor blade pressure, and coating speed. The process parameters associated with potential stretching hazards include stretching zone temperature, stretching ratio, and stretching speed. The process parameters associated with potential curing hazards include curing oven temperature and curing oven air velocity.

7. The method for detecting internal bubbles in polarizer production based on image processing according to claim 6, characterized in that: If the potential hazard is in the coating process, then the degree of hazard in the potential hazard process is determined. Extract the slurry viscosity, doctor blade pressure and coating speed at each time point during the coating process of the polarizer, calculate the average slurry viscosity, average doctor blade pressure and average coating speed. After comprehensively processing the average viscosity of the slurry, the average pressure of the doctor blade, and the average coating speed in conjunction with the coating process standards, the coating process hazard coefficient is output. If the coating process hazard coefficient is higher than the set coating threshold hazard coefficient, calculate the difference between the coating process hazard coefficient and the coating threshold hazard coefficient. Use the difference as the numerator and the coating threshold hazard coefficient as the denominator to calculate the ratio as the degree of hazard.

8. The method for detecting internal bubbles in polarizer production based on image processing according to claim 6, characterized in that: If the potential hazard in the process is a tensile hazard, then the degree of hazard in the process is determined. Extract the stretching zone temperature of the polarizer during the stretching time period; the stretching zone includes the inlet, middle section and outlet. For the inlet zone temperature, middle zone temperature, and outlet zone temperature at each time point within the stretching period, calculate the average values ​​and output the average inlet temperature, average middle zone temperature, and average outlet temperature; after calculating the standard deviation of the average inlet temperature, average middle zone temperature, and average outlet temperature, output the interval temperature difference. The stretching process risk coefficient is output by combining the temperature difference within the stretching time range with the stretching ratio and stretching speed, and then combining the stretching process standard. If the hazard coefficient of the stretching process is higher than the set hazard coefficient of the stretching threshold, the difference between the hazard coefficient of the stretching process and the hazard coefficient of the stretching threshold is calculated. The difference is used as the numerator and the hazard coefficient of the stretching threshold is used as the denominator to calculate the ratio as the degree of hazard.

9. The method for detecting internal bubbles in polarizer production based on image processing according to claim 6, characterized in that: If the potential hazard is a solidified hazard during the process, then the degree of hazard of the potential hazard is determined. Extract the temperature and wind speed of the polarizer at each time point during the curing time in the curing oven, calculate the average value of each time point, and output the average temperature and average wind speed in the oven. After comprehensively processing the average temperature and average wind speed inside the furnace in conjunction with the furnace curing standard, the curing process hazard coefficient is output. If the hazard coefficient of the curing process is higher than the hazard coefficient of the set curing threshold, the difference between the hazard coefficient of the curing process and the hazard coefficient of the curing threshold is calculated. The difference is used as the numerator and the hazard coefficient of the curing threshold is used as the denominator to calculate the ratio as the degree of hazard.

Images