Online marking method and device for TAC film surface microdefects and storage medium

Abstract

The invention discloses an online marking method and device for TAC film surface microdefects and a storage medium, and relates to the technical field of film microdefect marking, and the method comprises the following steps: setting a light scene, obtaining a to-be-detected TAC film image based on the light scene, and marking the to-be-detected TAC film image as a real-time TAC film image; obtaining a first type defect area and a first to-be-classified defect area based on the real-time TAC film image, the first gray threshold and the second gray threshold; constructing a real-time reference value based on the first defect area to be classified; obtaining a crystal point threshold value based on the TAC film image containing the crystal point; dividing the first to-be-classified defect area into a second type defect area and a third type defect area based on a real-time reference value and a crystal point threshold value; marking the first type of defect area, the second type of defect area and the third type of defect area on the TAC film surface differently; the method is used for solving the problem that defect traceability cannot be timely found subsequently due to the fact that an existing membrane microdefect marking technology cannot identify defect types.

Description

Technical Field

[0001] This invention relates to the field of membrane micro-defect marking technology, specifically to an online marking method, device, and storage medium for micro-defects on the surface of a TAC membrane. Background Technology

[0002] As the core protective material for polarizers in liquid crystal displays, the surface optical uniformity of TAC film directly determines the imaging quality and grade of the final display product. Micro-defects can affect the imaging quality and grade of the final display product, so micro-defect detection and calibration are necessary.

[0003] Traditional manual inspection of membrane micro-defects is subjective and inefficient. Therefore, image-based defect identification has been proposed. Traditional inspection systems typically output simple pass / fail signals. However, different types of defects have vastly different causes, impacts on product function, and required handling methods. For example, a dust defect that can be blown away and a crystal point indicating a process abnormality have drastically different implications for production. Failure to differentiate between them leads to a lack of targeted marking, rejection, and process feedback, resulting in resource waste and difficulties in tracing the source of problems. Existing membrane micro-defect marking technologies fail to identify defect types, hindering timely defect detection and tracing. Summary of the Invention

[0004] This invention aims to at least partially solve one of the technical problems in the prior art. By setting a lighting scene, an image of the TAC film to be inspected is acquired based on the lighting scene and marked as a real-time TAC film image. A first grayscale threshold and a second grayscale threshold are obtained based on the defect-free TAC film image. A first type of defect region and a first type of defect region to be classified are obtained based on the real-time TAC film image, the first grayscale threshold, and the second grayscale threshold. A real-time reference value is constructed based on the first type of defect region to be classified. A crystal point threshold is obtained based on the TAC film image containing crystal points. The first type of defect region to be classified is divided into a second type of defect region and a third type of defect region based on the real-time reference value and the crystal point threshold. The first type of defect region, the second type of defect region, and the third type of defect region are marked differently on the TAC film surface. This solves the problem that existing membrane micro-defect marking technologies fail to identify defect types, leading to the inability to promptly detect and trace the source of defects.

[0005] To achieve the above objectives, this application provides an online marking method for micro-defects on the TAC film surface, comprising the following steps: Define the lighting scene, acquire the TAC membrane image to be detected based on the lighting scene, and label it as the real-time TAC membrane image; The first grayscale threshold and the second grayscale threshold are obtained based on the defect-free TAC film image; The first type of defect region and the first defect region to be classified are obtained based on the real-time TAC film image, the first grayscale threshold and the second grayscale threshold. A real-time reference value is constructed based on the first defect region to be classified. Crystal point thresholds are obtained from TAC film images containing crystal points. Based on real-time reference values ​​and crystal point thresholds, the first defect region to be classified is divided into a second type of defect region and a third type of defect region. The first type of defect area, the second type of defect area, and the third type of defect area are marked differently on the TAC film surface.

[0006] Furthermore, setting up the lighting scene includes the following sub-steps: The lighting scenario involves placing a uniform light source under the TAC film to be inspected, and ensuring that the illumination intensity of the film surface area is consistent in the real-time TAC film image.

[0007] Furthermore, obtaining the first grayscale threshold and the second grayscale threshold based on the defect-free TAC film image includes the following sub-steps: The defect-free TAC membrane image is converted to grayscale to obtain a grayscale image, which is then marked as a historical normal membrane grayscale image. Obtain the grayscale values ​​of pixels in the historical normal membrane grayscale image and mark them as historical normal membrane grayscale values; Obtain the first number of historical normal membrane grayscale values; The number of identical historical normal membrane grayscale values ​​is marked as the number of historical normal membranes; A Cartesian coordinate system was established with the historical normal membrane grayscale value as the horizontal axis data and the historical normal membrane quantity as the vertical axis data, and it was marked as the grayscale value distribution coordinate system. The historical normal membrane grayscale values ​​and the corresponding number of historical normal membranes are plotted as data points on the grayscale value distribution coordinate system. Mark the data points in the grayscale value distribution coordinate system as normal membrane grayscale value distribution points; Obtain the range length of the x-coordinate of all normal membrane grayscale value distribution points, and label it A1; Create a line segment of length A2 on the horizontal axis of the grayscale value distribution coordinate system that can move left and right, and mark it as the first search line segment; The first abnormal threshold is calculated as: H1 = k1 × (A2 ÷ A1) × D1; where H1 is the first abnormal threshold, k1 is the first quantity ratio, and D1 is the first quantity.

[0008] Furthermore, obtaining the first grayscale threshold and the second grayscale threshold based on the defect-free TAC film image also includes the following sub-steps: The sum of the number of historical normal membrane gray values ​​corresponding to all historical normal membrane coordinate points directly above the first search line segment is marked as the first judgment number. Align the left endpoint of the first search line segment with the historical normal membrane coordinate point with the smallest horizontal coordinate, then move the first horizontal line segment to the right, and determine the size of the first judgment number and the first abnormal threshold. Stop moving the first search line segment when the first judgment number is greater than or equal to the first abnormal threshold. Obtain the historical normal membrane gray value corresponding to the left endpoint of the first search line segment at this time, and mark it as the first gray value threshold. Align the right endpoint of the first search line segment with the historical normal membrane coordinate point with the largest horizontal coordinate, then move the first horizontal line segment to the left, and determine the size of the first judgment number and the first abnormal threshold. Stop moving the first search line segment when the first judgment number is greater than or equal to the first abnormal threshold. Obtain the historical normal membrane gray value corresponding to the right endpoint of the first search line segment at this time, and mark it as the second gray value threshold.

[0009] Furthermore, obtaining the first type of defect region and the first defect region to be classified based on the real-time TAC film image, the first grayscale threshold, and the second grayscale threshold includes the following sub-steps: The real-time TAC membrane image is converted to grayscale to obtain a grayscale image, which is then labeled as the real-time membrane grayscale image. The grayscale values ​​of pixels in the real-time membrane grayscale image are marked as real-time membrane grayscale values; Real-time membrane gray values ​​that are less than the first gray value threshold are marked as first abnormal gray values; Real-time membrane gray values ​​that are greater than the second gray value threshold are marked as second abnormal gray values; The continuous region composed of the first abnormal gray values ​​is marked as the first type of defect region; The continuous region composed of the second abnormal gray values ​​is marked as the first defect region to be classified.

[0010] Furthermore, constructing real-time reference values ​​based on the first defect region to be classified includes the following sub-steps: Obtain the contour of each independent first defect region to be classified and mark it as the first abnormal contour. Obtain a second number of coordinate points on the first abnormal contour and mark them as the coordinate points of the first contour. The contour fitting function is constructed as: (x1-a1) 2 +(y2-a2) 2 =r 2 Where x1 and y2 are the abscissa and ordinate values ​​of the contour fitting function, respectively, and a1, a2, and r are constants; The first contour coordinate point is fitted with a contour fitting function to obtain the specific value of r; Obtain the distances from all coordinate points of the first contour to point (a1, a2), and mark them as the first distances; The mean of the absolute values ​​of the differences between each first distance and r is obtained and marked as the real-time reference value.

[0011] Furthermore, obtaining the crystal point threshold based on the TAC film image containing crystal points includes the following sub-steps: The TAC film image containing crystal points is treated as a real-time TAC film image to obtain real-time reference values, which are then marked as crystal point reference values. Obtain the third number of crystal point reference values; The number of identical crystal point reference values ​​is marked as the crystal point reference quantity; A Cartesian coordinate system is established with crystal point reference values ​​as the horizontal axis data and the number of crystal point references as the vertical axis data, and it is marked as the reference value distribution coordinate system. Plot the crystal point reference value and the corresponding crystal point reference quantity as data points on the reference value distribution coordinate system; Mark the data points in the reference value distribution coordinate system as normal crystal point distribution points; Obtain the range length of the x-coordinates of all normal crystal point distribution points, and label it A3; Create a line segment of length A4 on the horizontal axis of the reference value distribution coordinate system, which can move left and right, and mark it as the second search line segment; The second abnormal threshold is calculated as: H2 = k2 × (A4 ÷ A3) × D3; where H2 is the second abnormal threshold, k2 is the second quantity ratio, and D3 is the third quantity. The sum of the number of reference values ​​of all normal crystal point distribution points directly above the second search line segment is marked as the second judgment number. Align the right endpoint of the second search line segment with the normal crystal point distribution point with the largest horizontal coordinate, then move the second horizontal line segment to the left, and determine the size of the second judgment number and the second abnormal threshold. Stop moving the second search line segment when the second judgment number is greater than or equal to the second abnormal threshold. Obtain the crystal point reference value corresponding to the right endpoint of the second search line segment at this time, and mark it as the crystal point threshold.

[0012] Furthermore, dividing the first defect region to be classified into a second type of defect region and a third type of defect region based on real-time reference values ​​and crystal point thresholds includes the following sub-steps: If the real-time reference value is less than or equal to the crystal point threshold, the first defect region to be classified will be marked as the second type of defect region. If the real-time reference value is greater than the crystal point threshold, the first defect region to be classified will be marked as the third type of defect region.

[0013] This application provides an electronic device, including a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method described above are performed.

[0014] This application provides a storage medium on which a computer program is stored, which, when executed by a processor, performs the steps of the method described above.

[0015] The beneficial effects of this invention are as follows: This invention sets a lighting scene, acquires an image of the TAC film to be detected based on the lighting scene, and marks it as a real-time TAC film image; it acquires a first grayscale threshold and a second grayscale threshold based on the defect-free TAC film image; it acquires a first type of defect region and a first type of defect region to be classified based on the real-time TAC film image, the first grayscale threshold, and the second grayscale threshold; it constructs a real-time reference value based on the first type of defect region to be classified; it acquires a crystal point threshold based on the TAC film image containing crystal points; it divides the first type of defect region to be classified into a second type of defect region and a third type of defect region based on the real-time reference value and the crystal point threshold; and it marks the first type of defect region, the second type of defect region, and the third type of defect region differently on the TAC film surface. The advantage is that it can further classify the defect types and facilitate timely discovery and tracing of defect sources. The present invention constructs real-time reference values ​​based on the first defect region to be classified. Its advantage lies in that it further divides micro-defects based on crystal point characteristics, which can facilitate timely discovery and traceability of defect sources. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the steps of the method of the present invention; Figure 2 This is a schematic diagram of the normal film grayscale value distribution points of the present invention; Figure 3 This is a schematic diagram of r and points (a1, a2) in this invention; Figure 4 This is a schematic diagram of the normal crystal point distribution points of the present invention. Detailed Implementation

[0017] 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.

[0018] Example 1, please refer to Figure 1 As shown, this application provides an online marking method for micro-defects on a TAC film surface, comprising the following steps: Step S1: Set the lighting scene, acquire the TAC membrane image to be detected based on the lighting scene, and mark it as a real-time TAC membrane image; Step S1 includes the following sub-steps: Step S101: The lighting scene is to place a uniform light source under the TAC film to be tested, and to make the illumination intensity of the film surface area in the real-time TAC film image consistent; so that the gray value of the normal film is within a certain range, which facilitates the analysis of the micro-defect area, and the gray value of the defect is different from the gray value of the normal film.

[0019] Step S2 involves obtaining a first grayscale threshold and a second grayscale threshold based on the defect-free TAC film image; Step S2 includes the following sub-steps: Step S201: The defect-free TAC membrane image is converted to grayscale to obtain a grayscale image, which is then marked as a historical normal membrane grayscale image. Step S202: Obtain the grayscale values ​​of pixels in the historical normal film grayscale image and mark them as historical normal film grayscale values; Step S203: Obtain a first number of historical normal membrane grayscale values; in order to obtain the distribution range of historical normal membrane grayscale values, for example, the first number is 8 million; Step S204: The number of identical historical normal film grayscale values ​​is marked as the number of historical normal films; Step S205: Establish a Cartesian coordinate system with the historical normal membrane grayscale value as the horizontal axis data and the historical normal membrane quantity as the vertical axis data, and mark it as the grayscale value distribution coordinate system; Step S206: Plot the historical normal membrane grayscale values ​​and the corresponding number of historical normal membranes as data points on the grayscale value distribution coordinate system; in order to observe the distribution of historical normal membrane grayscale values; Step S207: Mark the data points in the grayscale value distribution coordinate system as normal membrane grayscale value distribution points; For practical applications, please refer to Figure 2 As shown, the distribution points of the normal membrane grayscale values ​​are plotted.

[0020] Step S208: Obtain the range length of the abscissa of all normal membrane grayscale value distribution points, marked as A1; please refer to... Figure 2 As shown, for example, if the distance between the horizontal axis 210 and 212 is 2cm, then A1 is 20cm; Step S209: Establish a line segment of length A2 on the horizontal axis of the gray value distribution coordinate system, which can move left and right, and mark it as the first search line segment; in order to observe the distribution of gray values ​​of historical normal films, A2 should be less than A1, so A2 is 1cm; Step S210, calculate the first anomaly threshold as: H1=k1×(A2÷A1)×D1; where H1 is the first anomaly threshold, k1 is the first quantity ratio, and D1 is the first quantity; the first anomaly threshold is to obtain the area with fewer historical normal membrane gray values, (A2÷A1)×D1 represents the average number of historical normal membrane gray values ​​on the first search line segment, so k1 is set to be less than 1, for example, k1 is 0.15; In practical applications, the first abnormal threshold is calculated as: H1 = 0.15 × (1 ÷ 20) × 800 = 6, so the first abnormal threshold is 6.

[0021] Step S211: Mark the sum of the number of historical normal membrane gray values ​​corresponding to all historical normal membrane coordinate points directly above the first search line segment as the first judgment number; Step S212: Align the left endpoint of the first search line segment with the historical normal membrane coordinate point with the smallest horizontal coordinate, then move the first horizontal line segment to the right, and determine the size of the first judgment number and the first abnormal threshold. Stop moving the first search line segment when the first judgment number is greater than or equal to the first abnormal threshold; obtain the historical normal membrane gray value corresponding to the left endpoint of the first search line segment at this time, and mark it as the first gray value threshold; delete historical normal membrane gray values ​​that are abnormally small, and then obtain the smallest normal historical membrane gray value. Step S213: Align the right endpoint of the first search line segment with the historical normal membrane coordinate point with the largest horizontal coordinate, then move the first horizontal line segment to the left, and determine the size of the first judgment number and the first abnormal threshold. Stop moving the first search line segment when the first judgment number is greater than or equal to the first abnormal threshold; obtain the historical normal membrane gray value corresponding to the right endpoint of the first search line segment at this time, and mark it as the second gray value threshold; delete the historical normal membrane gray values ​​that are abnormally large, and then obtain the largest normal historical membrane gray value. For practical applications, please refer to Figure 2 As shown, the left endpoint of the first search segment is aligned with the historical normal membrane coordinate point with the smallest horizontal coordinate. Then, the first horizontal segment is moved to the right. The movement of the first search segment stops when the number of first judgments is 8, which is greater than the first anomaly threshold of 6. The gray value of the historical normal membrane corresponding to the left endpoint of the first search segment at this time is 211, so the first gray value threshold is 211. The right endpoint of the first search segment is aligned with the historical normal membrane coordinate point with the largest horizontal coordinate. Then, the first horizontal segment is moved to the left. The size of the first judgment count and the first anomaly threshold is compared. The movement of the first search segment stops when the number of first judgments is 8, which is greater than the first anomaly threshold of 6. The gray value of the historical normal membrane corresponding to the right endpoint of the first search segment at this time is 229, so the second gray value threshold is 229.

[0022] Step S3 involves obtaining the first type of defect region and the first defect region to be classified based on the real-time TAC film image, the first grayscale threshold, and the second grayscale threshold; Step S3 includes the following sub-steps: Step S301: The real-time TAC membrane image is converted to grayscale to obtain a grayscale image, which is then marked as the real-time membrane grayscale image. Step S302: Mark the gray values ​​of the pixels in the real-time membrane grayscale image as the real-time membrane grayscale values; Step S303: Mark real-time membrane gray values ​​that are less than the first gray value threshold as first abnormal gray values; Step S304: Mark real-time membrane gray values ​​that are greater than the second gray value threshold as second abnormal gray values; Step S305: Mark the continuous area composed of the first abnormal gray values ​​as the first type of defect area; if there are adhering substances on the film surface, such as dust, the adhering substances will block the light source, resulting in a low gray value, so it can be classified as the first type of defect area. Step S306: Mark the continuous area composed of the second abnormal gray values ​​as the first defect area to be classified; the defect area itself will cause light leakage or light concentration, resulting in an increase in gray value, so the first defect area to be classified can be marked. In practical applications, real-time membrane gray values ​​less than 211 are marked as the first abnormal gray value; real-time membrane gray values ​​greater than 229 are marked as the second abnormal gray value.

[0023] Step S4: Construct real-time reference values ​​based on the first defect region to be classified; Step S4 includes the following sub-steps: Step S401: Obtain the contour of each independent first defect region to be classified and mark it as the first abnormal contour; Step S402: Obtain a second number of coordinate points on the first abnormal contour and mark them as the coordinate points of the first contour; the second number of first contour coordinate points facilitates function fitting, for example, the second number is 20; Step S403, construct the contour fitting function as: (x1-a1) 2 +(y2-a2) 2 =r 2 Where x1 and y2 are the abscissa and ordinate values ​​of the contour fitting function, respectively, and a1, a2 and r are constants; the contour fitting function is the equation of a circle. If it is a defect of its own nature, the contour is usually irregular, while the contour of the crystal point is close to a circle; the raw materials and internal processes of the crystal point are related to internal problems, while the scratches are related to external factors, so they can be further divided. Step S404: Fit the first contour coordinate points with a contour fitting function to obtain the specific value of r; r is the radius of the contour fitting function. Step S405: Obtain the distances from all first contour coordinate points to point (a1, a2), and mark them as the first distances; point (a1, a2) is the center of the contour fitting function; therefore, the first distance is the distance from the first contour coordinate point to the center of the contour fitting function. Step S406: Obtain the mean of the absolute values ​​of the differences between each first distance and r, and mark it as the real-time reference value; the real-time reference value represents the value that is not close to a circle. The smaller the real-time reference value, the closer it is to a circle; the larger the real-time reference value, the less similar it is to a circle. For practical applications, please refer to Figure 3 As shown, r is obtained as 0.10nm, and point (a1, a2) is point (0.3, 0.3); the real-time reference value is obtained as 0.010nm.

[0024] Step S5: Obtain the crystal point threshold based on the TAC film image containing crystal points; Step S5 includes the following sub-steps: Step S501: Treat the TAC film image containing crystal points as a real-time TAC film image to obtain real-time reference values, and mark them as crystal point reference values; Step S502: Obtain a third number of crystal point reference values; to obtain the range of crystal point reference values, for example, the third number is 800; Step S503: Mark the number of identical crystal point reference values ​​as the crystal point reference number; Step S504: Establish a Cartesian coordinate system with the crystal point reference value as the horizontal axis data and the number of crystal point references as the vertical axis data, and mark it as the reference value distribution coordinate system; Step S505: Plot the crystal point reference value and the corresponding crystal point reference quantity as data points on the reference value distribution coordinate system; Step S506: Mark the data points in the reference value distribution coordinate system as normal crystal point distribution points; in order to observe the distribution of crystal point reference values; For practical applications, please refer to Figure 4 As shown, the normal crystal point distribution points were obtained.

[0025] Step S507: Obtain the range length of the abscissa of all normal crystal point distribution points, marked as A3; please refer to... Figure 4 As shown, for example, if the distance between the horizontal axis 0.002 and 0.004 is 2cm, then A1 is 22cm; Step S508: Establish a line segment of length A4 on the horizontal axis of the reference value distribution coordinate system, which can move left and right, and mark it as the second search line segment; the second search line segment is 1cm long because A4 is used to analyze the distribution of crystal point reference values. Step S509, calculate the second anomaly threshold as: H2 = k2 × (A4 ÷ A3) × D3; where H2 is the second anomaly threshold, k2 is the second quantity ratio, and D3 is the third quantity; the second anomaly threshold is to obtain the area with fewer crystal point reference values, and (A4 ÷ A3) × D3 represents the average number of crystal point reference values ​​on the second search line segment, so k2 is set to be less than 1, for example, k2 is 0.15; In practical applications, the second abnormal threshold is calculated as: H1 = 0.15 × (1 ÷ 21) × 800 = 5.7. The result is rounded to one decimal place, so the second abnormal threshold is 5.7.

[0026] Step S510: Mark the sum of the number of reference values ​​of all normal crystal point distribution points directly above the second search line segment as the second judgment number; Step S511: Align the right endpoint of the second search line segment with the normal crystal point distribution point with the largest horizontal coordinate, then move the second horizontal line segment to the left, and determine the size of the second judgment number and the second abnormal threshold. Stop moving the second search line segment when the second judgment number is greater than or equal to the second abnormal threshold; obtain the crystal point reference value corresponding to the right endpoint of the second search line segment at this time, and mark it as the crystal point threshold; exclude individual excessively large crystal point reference values, and then obtain a more reasonable and accurate crystal point threshold. For practical applications, please refer to Figure 4 As shown, the right endpoint of the second search line segment is aligned with the normal crystal point distribution point with the largest horizontal coordinate. Then, the second horizontal line segment is moved to the left. The size of the second judgment number and the second abnormal threshold is judged. When the second judgment number is 6, which is greater than the second abnormal threshold of 5.7, the movement of the second search line segment is stopped. The crystal point reference value corresponding to the right endpoint of the second search line segment at this time is obtained as 0.021, so the crystal point threshold is 0.021.

[0027] Step S6: Based on the real-time reference value and the crystal point threshold, the first defect region to be classified is divided into a second type of defect region and a third type of defect region; Step S6 includes the following sub-steps: Step S601: If the real-time reference value is less than or equal to the crystal point threshold, the first defect region to be classified is marked as a second type of defect region; the outline is close to a circle, so the defect is very likely a crystal point defect. Step S602: If the real-time reference value is greater than the crystal point threshold, the first defect area to be classified is marked as the third type of defect area; the outline is not close to a circle, and the second type of defect area is a defect caused by external reasons such as scratches; therefore, the first defect area to be classified can be divided into the second type of defect area and the third type of defect area.

[0028] Step S7: Mark the first type of defect area, the second type of defect area, and the third type of defect area on the TAC film surface differently; Marking the first type of defect area, the second type of defect area, and the third type of defect area on the TAC film surface differently based on the image facilitates timely discovery and tracing of defects.

[0029] Example 2: This application also provides an electronic device, which may include: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, the steps of an online marking method for micro-defects on a TAC film surface are performed to achieve the following functions: setting a lighting scene; acquiring an image of the TAC film to be detected based on the lighting scene and marking it as a real-time TAC film image; acquiring a first grayscale threshold and a second grayscale threshold based on a defect-free TAC film image; acquiring a first type of defect region and a first type of defect region to be classified based on the real-time TAC film image, the first grayscale threshold, and the second grayscale threshold; constructing a real-time reference value based on the first type of defect region to be classified; acquiring a crystal point threshold based on a TAC film image containing crystal points; dividing the first type of defect region to be classified into a second type of defect region and a third type of defect region based on the real-time reference value and the crystal point threshold; and marking the first type of defect region, the second type of defect region, and the third type of defect region differently on the TAC film surface.

[0030] Furthermore, when the logical instructions in the aforementioned memory can be 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 hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0031] Example 3: This application also provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute an online marking method for micro-defects on a TAC film surface provided by the above methods. The method includes: setting a lighting scene; acquiring an image of the TAC film to be detected based on the lighting scene and marking it as a real-time TAC film image; acquiring a first grayscale threshold and a second grayscale threshold based on a defect-free TAC film image; acquiring a first type of defect region and a first type of defect region to be classified based on the real-time TAC film image, the first grayscale threshold, and the second grayscale threshold; constructing a real-time reference value based on the first type of defect region to be classified; acquiring a crystal point threshold based on a TAC film image containing crystal points; dividing the first type of defect region to be classified into a second type of defect region and a third type of defect region based on the real-time reference value and the crystal point threshold; and marking the first type of defect region, the second type of defect region, and the third type of defect region differently on the TAC film surface.

[0032] Example 4: This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it performs the steps of the above-described online marking method for micro-defects on a TAC film surface to achieve the following functions: setting a lighting scene; acquiring an image of the TAC film to be detected based on the lighting scene and marking it as a real-time TAC film image; acquiring a first grayscale threshold and a second grayscale threshold based on a defect-free TAC film image; acquiring a first type of defect region and a first type of defect region to be classified based on the real-time TAC film image, the first grayscale threshold, and the second grayscale threshold; constructing a real-time reference value based on the first type of defect region to be classified; acquiring a crystal point threshold based on a TAC film image containing crystal points; dividing the first type of defect region to be classified into a second type of defect region and a third type of defect region based on the real-time reference value and the crystal point threshold; and marking the first type of defect region, the second type of defect region, and the third type of defect region differently on the TAC film surface.

[0033] Based on the above description of the embodiments, the embodiments of the present invention can be provided as methods, systems, or computer program products. Based on this understanding, the above technical solutions, in essence or in terms of their contribution to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or certain parts of the embodiments.

[0034] In the embodiments provided in this application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of modules or units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces. The indirect coupling or communication connection between systems, modules, and units may be electrical, mechanical, or other forms.

[0035] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An on-line marking method of micro-defects on a TAC film surface, characterized by, The method comprises the following steps: Setting a light scene, obtaining a TAC film image to be detected based on the light scene, and marking as a real-time TAC film image; Obtaining a first gray threshold and a second gray threshold based on a defect-free TAC film image; Obtaining a first type of defect area and a first to-be-classified defect area based on the real-time TAC film image, the first gray threshold, and the second gray threshold; Constructing a real-time reference value based on the first to-be-classified defect area; Obtaining a crystal point threshold based on a TAC film image containing a crystal point; Dividing the first to-be-classified defect area into a second type of defect area and a third type of defect area based on the real-time reference value and the crystal point threshold; Marking the first type of defect area, the second type of defect area, and the third type of defect area on the TAC film respectively.

2. The method according to claim 1, wherein the method is characterized by, The step of setting a light scene comprises the following sub-steps: The light scene is to place a uniform light source under the TAC film to be detected, and to make the illumination intensity of the film surface area in the real-time TAC film image consistent.

3. The method of claim 2, wherein the TAC film surface microdefects are marked on line. The step of obtaining a first gray threshold and a second gray threshold based on a defect-free TAC film image comprises the following sub-steps: Performing gray processing on the defect-free TAC film image to obtain a gray image, and marking as a historical normal film gray image; Obtaining the gray values of the pixel points in the historical normal film gray image, and marking as historical normal film gray values; Obtaining a first number of historical normal film gray values; Marking the number of the same historical normal film gray values as the number of historical normal films; Establishing a plane rectangular coordinate system with the historical normal film gray values as the horizontal axis data and the number of historical normal films as the vertical axis data, and marking as a gray value distribution coordinate system; Plotting the historical normal film gray values and the corresponding number of historical normal films as data points on the gray value distribution coordinate system; Marking the data points in the gray value distribution coordinate system as normal film gray value distribution points; Obtaining the range length of the horizontal coordinates of all normal film gray value distribution points, and marking as A1; Establishing a line segment with a length of A2 and movable left and right on the horizontal axis of the gray value distribution coordinate system, and marking as a first search line segment; Calculating the first abnormal threshold value as: H1=k1×(A2÷A1)×D1; wherein H1 is the first abnormal threshold value, k1 is the first number ratio, and D1 is the first number.

4. The method of claim 3, wherein the TAC film surface micro-defects are marked on line. The step of obtaining a first gray threshold and a second gray threshold based on a defect-free TAC film image further comprises the following sub-steps: Marking the sum of the number of historical normal film gray values corresponding to all historical normal film coordinate points above the first search line segment as a first judgment number; Reconciling the left end point of the first search line segment with the historical normal film coordinate point with the smallest horizontal coordinate, then moving the first horizontal moving line segment to the right, and judging the size of the first judgment number and the first abnormal threshold value, and stopping moving the first search line segment when the first judgment number is greater than or equal to the first abnormal threshold value; Obtaining the historical normal film gray value corresponding to the left end point of the first search line segment at this time, and marking as the first gray threshold value; Reconciling the right end point of the first search line segment with the historical normal film coordinate point with the largest horizontal coordinate, then moving the first horizontal moving line segment to the left, and judging the size of the first judgment number and the first abnormal threshold value, and stopping moving the first search line segment when the first judgment number is greater than or equal to the first abnormal threshold value; Obtain the historical normal film gray value corresponding to the right end point of the first search line segment at this time, and mark it as a second gray threshold.

5. The method of claim 4, wherein the TAC film surface microdefects are marked on line. Obtain the first kind of defect region and the first to-be-classified defect region based on the real-time TAC film image, the first gray threshold and the second gray threshold, including the following sub-steps: Perform gray processing on the real-time TAC film image to obtain a gray image, and mark it as a real-time film gray image; Mark the gray value of the pixel point in the real-time film gray image as a real-time film gray value; Mark the real-time film gray value less than the first gray threshold as a first abnormal gray value; Mark the real-time film gray value greater than the second gray threshold as a second abnormal gray value; Mark the continuous region composed of the first abnormal gray value as the first kind of defect region; Mark the continuous region composed of the second abnormal gray value as the first to-be-classified defect region.

6. The method of claim 5, wherein the method further comprises: Construct the real-time reference value based on the first to-be-classified defect region, including the following sub-steps: Obtain the contour of each independent first to-be-classified defect region, and mark it as a first abnormal contour; Obtain a second number of coordinate points on the first abnormal contour, and mark them as first contour coordinate points; The contour fitting function is constructed as: (x1-a1) 2 + (y2-a2) 2 = r 2 ; wherein x1 and y2 are respectively the horizontal coordinate value and the vertical coordinate value of the contour fitting function, and a1, a2 and r are constants; Fit the first contour coordinate points with a contour fitting function to obtain the specific value of r; Obtain the distance from all first contour coordinate points to point (a1, a2), and mark it as a first distance; Obtain the mean value of the absolute value of the difference between each first distance and r, and mark it as a real-time reference value.

7. The method of claim 6, wherein the method is characterized by: Obtain the crystal point threshold based on the crystal point-containing TAC film image, including the following sub-steps: Take the crystal point-containing TAC film image as the real-time TAC film image to obtain the real-time reference value, and mark it as a crystal point reference value; Obtain a third number of crystal point reference values; Mark the number of the same crystal point reference value as the crystal point reference number; Establish a plane rectangular coordinate system with the crystal point reference value as the horizontal axis data and the crystal point reference number as the vertical axis data, and mark it as a reference value distribution coordinate system; Plot the crystal point reference value and the corresponding crystal point reference number as data points in the reference value distribution coordinate system; Mark the data points in the reference value distribution coordinate system as normal crystal point distribution points; Obtain the range length of the horizontal coordinates of all normal crystal point distribution points, and mark it as A3; Establish a line segment with a length of A4 and movable left and right on the horizontal axis of the reference value distribution coordinate system, and mark it as a second search line segment; Calculate the second abnormal threshold value as: H2=k2×(A4÷A3)×D3; wherein H2 is the second abnormal threshold value, k2 is the second number ratio, and D3 is the third number; Mark the sum of the number of crystal point reference values corresponding to all normal crystal point distribution points directly above the second search line segment as a second judgment number; Make the right end point of the second search line segment coincide with the normal crystal point distribution point with the largest horizontal coordinate, and then move the second horizontal moving line segment to the left, and judge the size of the second judgment number and the second abnormal threshold value, and stop moving the second search line segment when the second judgment number is greater than or equal to the second abnormal threshold value. Obtain the crystal point reference value corresponding to the right end point of the second search line segment at this time, and mark it as a crystal point threshold.

8. The method of claim 7, wherein the method is characterized by: Divide the first to-be-classified defect region into a second kind of defect region and a third kind of defect region based on the real-time reference value and the crystal point threshold, including the following sub-steps: If the real-time reference value is less than or equal to the crystal point threshold value, the first defect region to be classified is marked as a second kind of defect region; If the real-time reference value is greater than the crystal point threshold value, the first defect region to be classified is marked as a third kind of defect region.

9. An electronic device, comprising: A computer program product comprising a processor and a memory, said memory storing computer readable instructions which, when executed by the processor, perform the steps of any of the methods of claims 1-8.

10. A storage medium having stored thereon a computer program, characterized in that The computer program, when executed by a processor, performs the steps of any of the methods of claims 1-8.

Images