Plate defect detection system and method and storage medium

Abstract

The invention discloses a plate defect detection system and method and a storage medium, and relates to the technical field of image detection. Comprising the steps of capturing a plate image of a plate under an optical framework, obtaining a strain force distribution image in the plate according to the plate image, obtaining a transmission image of the plate, and obtaining an abnormal area in the transmission image, a stacked plate image, the strain force distribution image and the transmission image according to a positioning method. According to the plate image, the strain force distribution image and the transmission image obtained under the optical architecture, the area for positioning the abnormal position in the projection image according to the gray value is constructed, and the internal stress quantity of the abnormal area position is determined in combination with the strain force distribution image; and finally, the internal stress quantity is compared with the existing defect type characteristic database, so that the possible defects of the plate in the use stage are detected, and guarantee is provided for after-sales and subsequent replacement and maintenance of the plate.

Description

Technical Field

[0001] The present invention relates to the technical field of image detection, and in particular to a plate defect detection system, method and storage medium. Background Art

[0002] During the application of sheet materials, due to improper operation in the production, transportation or processing links, defects such as cracks, scratches and stains may appear on the surface of the sheet materials. These defects not only damage the appearance quality of the sheet materials, but may also have a negative impact on their performance and safety. Therefore, it is particularly necessary to develop a sheet material defect detection system and its supporting methods and storage media.

[0003] After searching, the Chinese invention patent with publication number "CN115018826A" discloses "a method and system for fabric defect detection based on image recognition". This application determines each defect point and its position in the surface image of the fabric to be detected, determines the concentration coefficient of each defect point relative to the overall defect points, and determines the defect indication amount of the fabric to be detected based on the concentration coefficient of each defect point in the surface image of the fabric to be detected relative to the overall defect points and the gray value of each defect point, and then determines whether there are defects of the corresponding defect type in the fabric to be detected.

[0004] In addition, the Chinese invention patent with publication number "CN117173171A" discloses "a material surface detection method, device and system based on image processing". This application detects the material surface through a laser detection head, obtains a color three-dimensional map of the material surface in a three-dimensional coordinate system, performs image recognition on the color three-dimensional map according to the visible light wavelength range, and obtains several first target areas. For each first target area, a cross section is made to obtain a target triangular surface, and a two-dimensional line graph is generated according to the coordinates of the circumcenter point of the target triangular surface. The first target point is selected from the two-dimensional line graph, and a straight line fitting is performed on the circumcenter point of the target triangular surface according to the first target point to obtain a defect range. The defect range is marked in the two-dimensional line graph, and the corresponding color three-dimensional map is selected according to the defect range for display, so that the viewing angle of the cross section of the defect area is presented to the defect range in the form of a two-dimensional line graph, and the color three-dimensional map of the surface of the defect area is displayed at the same time, which more intuitively presents the surface state of the defect location of the detected material.

[0005] In view of the transparent plate with good sealing and intact surface, when the edge of the plate is squeezed during its use, the inside of the plate will be subjected to uneven force, thereby generating internal stress. These internal stresses generated during use will cause the material properties of the stress concentration area inside the plate to change during use, which may induce the plate to rupture or deform. However, in actual applications, the two patent methods disclosed above and similar methods are difficult to effectively detect defects that may be caused by stress concentration inside the transparent plate. In addition, these methods rely on image recognition technology, which may be more effective for detecting surface defects, but have limited effects on potential defects caused by internal stress inside the plate. Therefore, the purpose of the present invention is to provide a system, method and storage medium that can comprehensively and accurately detect surface and internal defects of a plate to address the shortcomings of the prior art. Summary of the invention

[0006] The object of the present invention is to provide a plate defect detection system, method and storage medium to solve the problems raised in the above background technology.

[0007] In a first aspect, the present invention provides a plate defect detection method, comprising: Capturing the plate image under the optical framework; Obtaining a strain force distribution image inside the plate based on the plate image; Acquire a transmission image of the plate; According to the positioning method, the abnormal area in the transmission image is obtained; Stacking plate images, strain distribution images and transmission images, and obtaining the position of the abnormal area in the plate image and the strain distribution characteristics of the plate at the abnormal area; Identify the defect types of the plate based on the strain distribution characteristics; The positioning method comprises: According to the specification parameters of the transmission image, a two-dimensional coordinate system is constructed in the transmission image, wherein the abscissa axis of the two-dimensional coordinate system represents the width direction of the plate, and the ordinate axis represents the length direction of the plate; In the constructed two-dimensional coordinate system, a plurality of grid units are set, each grid unit corresponds to a detection area on the plate, and the grid unit is an M×N matrix, where M is the number of rows and N is the number of columns; For different types of plates, the corresponding grayscale values ​​are retrieved and obtained from Internet resources and set as threshold parameters; Obtain the pixel gray value of each grid cell in the transmission image; Compare the ratios of different pixel gray values ​​to threshold parameters; Based on the comparison results, the grid cells whose pixel grayscale values ​​exceed the threshold parameters are identified as abnormal areas.

[0008] As a further preferred embodiment of the present technical solution, the calculation formula of the pixel gray value at the grid unit is: ; Used to represent the pixel gray value at the grid unit. It is used to indicate the position of the grid unit in the two-dimensional coordinate system. Used to indicate The red channel intensity value at Used to indicate The green channel intensity value at Used to indicate The blue channel intensity value at , the coefficients 0.299, 0.587 and 0.114 are based on the sensitivity weights of the human body to red, green and blue.

[0009] As a further preferred embodiment of the present technical solution, the positioning method includes: According to the specification parameters of the transmission image, a two-dimensional coordinate system is constructed in the transmission image, wherein the abscissa axis of the two-dimensional coordinate system represents the width direction of the plate, and the ordinate axis represents the length direction of the plate; In the constructed two-dimensional coordinate system, a plurality of grid units are set, and an identification code is constructed in each grid unit, and the identification code is used to display the clarity of the transmission image at the grid unit; Obtain a first-level grid unit, where the first-level grid unit is the one with the clearest identification code among multiple grid units; Obtain a secondary grid unit, where the secondary grid unit is the one with the most ambiguous identification code among the multiple grid units; Get the limit distance between the primary grid unit and the secondary grid unit; The abnormal area is established based on the primary grid unit, the secondary grid unit and the limit distance.

[0010] As a further preferred embodiment of the present technical solution, the formula for obtaining the clarity includes: ;in To indicate clarity, Used to The gradient value of the position is converted into a global average statistic, Used to indicate The gradient value of the position, where , used to indicate location The horizontal gradient value of Used to indicate location Vertical gradient.

[0011] As a further preferred embodiment of the present technical solution, the stacking method of the plate image, the strain distribution image and the transmission image includes: Determine a corresponding processing ratio based on the size specifications of the plate image, the strain distribution image, and the transmission image, and the processing ratio is used to perform a scaling or enlargement operation on any one of the plate image, the strain distribution image, and the transmission image; Overlay processing is performed on the scaled or expanded plate image, strain distribution image and transmission image; The overlay processing method is to overlay the strain distribution image and the transmission image on the plate image with a transparency parameter. The transparency parameter is set according to the needs, aiming to make the image information clearly displayed without interfering with each other.

[0012] As a further preferred embodiment of the present technical solution, the method for identifying the defect type of the plate according to the strain force distribution characteristics includes: According to the type of plate, the strain density threshold range of the corresponding defect type is obtained in the network; Construct a defect type feature database based on strain force density threshold range; Correlate the defect type feature database with the strain force distribution features of the abnormal area; According to the correlation results, the defect type of the plate is determined.

[0013] As a further preferred embodiment of the present technical solution, the method for associating the type feature library with the strain distribution features at the abnormal area includes: Obtain the total amount of strain in the abnormal area; Measure the total area of ​​the abnormal region; The strain value per unit area is calculated based on the total area and the total amount of strain force. The strain value per unit area is used to reflect the strain force density in the abnormal area. Comparing the strain force value per unit area with the strain force density threshold interval; According to the comparison result, the position information of the unit area strain force value within the strain force density threshold range is determined; According to the position information, the corresponding defect type is identified in the defect type feature database.

[0014] In the second aspect, in order to improve the above technical solution, the present invention further proposes a plate defect detection system. It should be noted that the plate defect detection system adopts the above-mentioned plate defect detection method and includes: An image capture module, used to capture images of the plate under an optical framework, including a plate image, a strain distribution image, and a transmission image; An image processing module, used for processing the captured image; The abnormal region detection module detects abnormal regions based on the transmission image and using a positioning method; An image stacking module, used to stack the plate image, strain distribution image and transmission image together; A defect type identification module is used to analyze the strain force distribution characteristics of the abnormal area and determine the defect type based on the strain force distribution characteristics; The user interface and interaction module is used to provide an interactive interface between the user and the system, so that the user can intuitively view the type and location of defects on the board and perform corresponding interactive steps; The data storage and management module is used to store the captured image abnormal area information and the identification results of the defect types.

[0015] On the third aspect, in order to further complete the above technical solution, the present invention also proposes a storage medium of a plate defect detection system. It should be noted that a storage medium of a plate defect detection system is equipped with the aforementioned plate defect detection system.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The plate defect detection system, method and storage medium are based on the plate image, strain distribution image and transmission image obtained under the optical framework, and locate the area of ​​abnormal position in the projection image according to the grayscale value, and determine the amount of internal stress in the abnormal area in combination with the strain distribution image. Finally, by comparing the amount of internal stress with the existing defect type feature database, it realizes the detection of defects that may appear in the plate during the use stage, and provides guarantee for the after-sales and subsequent replacement and maintenance of the plate. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of the steps of the method disclosed in the present invention; Figure 2 is an auxiliary illustration of step S100 of the present invention; Figure 3 is an auxiliary illustration of step S401 of the present invention; Figure 4 It is an auxiliary illustration diagram of step S402.1 to step S402.4 of the present invention; Figure 5 This is a module composition diagram of the system disclosed in the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Before exploring the solution proposed by the present invention, it must be made clear that the plates detected by the present invention are transparent materials, including glass, acrylic plates and transparent plastic plates. It should be supplemented that in the specific implementation of the present invention, when the transparent material plates are in use, when the edges of the transparent material plates are subjected to pressure, the inside of the transparent material plates will be subjected to the action of non-uniform forces, thereby generating internal stress. In addition, the internal stress generated by these transparent material plates during use will also cause the material properties of the stress concentration areas inside the transparent material plates to change after experiencing high and low temperature environments, which may cause the transparent material plates to further rupture or deform. Therefore, the main purpose of the technical method proposed by the present invention is to detect defects that may occur in the manufacturing or use of these transparent material plates, especially potential defects caused by internal stress.

[0020] It should be noted that the reference Figure 1 It can be seen that the present invention proposes a plate defect detection method, which specifically includes steps S100 to S600.

[0021] The method comprises: Step S100: capturing a plate image of the plate under an optical framework.

[0022] It should be noted that in the present invention, the specific form of the optical architecture can be referred to Figure 2 In addition, the plate image acquired by the optical architecture is used to show the observer the detailed morphological structure and surface features of the plate.

[0023] Specifically, refer to Figure 2 It can be seen that the optical architecture includes: A white spectrum screen acting as a background is used to provide a uniform and bright background light source to better highlight the defects on the surface of the board.

[0024] A black reflective screen is placed perpendicular to the underside of the white spectrum screen to capture light reflected from the edge of the sheet.

[0025] Optical photographic equipment (HD camera) is used to obtain plate images and subsequent strain distribution images and transmission images.

[0026] Step S200: obtaining a strain distribution image inside the plate according to the plate image.

[0027] It should be clearly pointed out that in the present invention, reference Figure 2The strain distribution image inside the plate is obtained by placing the plate to be tested above a black reflective screen and in close contact with a white spectrum screen. In the presence of stress inside the plate, the propagation path of light inside the plate will be offset. Therefore, through the white spectrum screen and the black reflective screen, the refraction or reflection of light caused by stress inside the plate can be captured and displayed intuitively. Subsequently, optical photography equipment (high-definition camera) is used to record and store it.

[0028] Step S300: Acquire a transmission image of the plate.

[0029] It should be pointed out that the projected image of the plate in the present invention is a white spectrum screen image obtained by an optical photographic device through the plate to be inspected.

[0030] In addition, during the actual detection process, a number of marking patterns serving as reference objects may be set in the white spectrum screen in the optical structure, and the proportion of these marking patterns in the white spectrum screen does not exceed one tenth of the total proportion.

[0031] Step S400: Acquire the abnormal area in the transmission image according to the positioning method.

[0032] It should be noted that, in the present invention, the abnormal area in step S400 is used to indicate an area where defects or internal stress concentration exist in the plate.

[0033] It should be supplemented that, in step S400, the positioning method includes: step S401 to step S406.

[0034] Step S401: constructing a two-dimensional coordinate system in the transmission image according to the specification parameters of the transmission image.

[0035] refer to Figure 3 It can be seen that in step S401, the abscissa axis of the two-dimensional coordinate system represents the width direction of the plate, and the ordinate axis represents the length direction of the plate.

[0036] Step S402: setting a plurality of grid units in the constructed two-dimensional coordinate system.

[0037] It should be noted that in step S402, each grid unit corresponds to a detection area on the plate, and the grid unit is an M×N matrix, where M is the number of rows and N is the number of columns.

[0038] Step S403: for different types of plates, corresponding grayscale values ​​are retrieved from Internet resources and obtained and set as threshold parameters.

[0039] Step S404: Obtain the pixel grayscale value of each grid unit in the transmission image.

[0040] Step S405: comparing the ratios of different pixel grayscale values ​​to the threshold parameters.

[0041] Step S406: Based on the comparison result, the grid cells whose pixel grayscale values ​​exceed the threshold parameter are identified as abnormal areas.

[0042] It should be clear that the comparison result in step S406 is the ratio in step S405. In addition, it should be noted that steps S401 to S406 disclosed in the present invention obtain the grayscale values ​​of different detection intervals in the grid unit and compare them with the preset threshold parameters, so as to accurately locate the areas where defects or internal stress concentrations exist in the plate, which has higher accuracy and efficiency than traditional manual inspection or simple machine vision inspection.

[0043] In addition, during the actual operation, in order to improve the accuracy of positioning, in step S402 of the present invention, an identification code can be constructed in each grid unit, and the identification code is used to display the clarity of the transmission image at the grid unit. By obtaining the primary grid unit (i.e., the grid unit with the highest clarity) and the secondary grid unit (i.e., the grid unit with the lowest clarity) and calculating the limiting distance between them, the scope of the abnormal area can be further narrowed and the positioning accuracy can be improved.

[0044] Specifically, it includes: step S402.1 to step S402.4.

[0045] Step S402.1: Obtain a first-level grid unit, where the first-level grid unit is the one with the clearest identification code among multiple grid units.

[0046] Step S402.2: Obtain a secondary grid unit, where the secondary grid unit is the one with the most ambiguous identification code among the multiple grid units.

[0047] Step S402.3: Obtain the limit distance between the primary grid unit and the secondary grid unit.

[0048] Step S402.4: Establish an abnormal area based on the primary grid unit, the secondary grid unit and the limit distance.

[0049] Supplementary information, see Figure 4 As shown, a circular area is constructed in the abnormal area during the execution of steps S402.1 to S402.4, wherein the abnormal area is established by obtaining the limiting distance between the primary grid unit and the secondary grid unit and the interaction position between the two-dimensional coordinate system as the intersection point, and constructing a circular area with the intersection point, wherein the circular area covers the primary grid unit and the secondary grid unit.

[0050] In addition, it is necessary to supplement step S404 that the calculation formula of the pixel gray value at the grid unit in the present invention is: ;It should be clear that Used to represent the pixel gray value at the grid unit. It is used to represent the position of the grid cell in the two-dimensional coordinate system. Used to indicate The red channel intensity value at Used to indicate The green channel intensity value at Used to indicate The blue channel intensity value at , the coefficients 0.299, 0.587 and 0.114 are based on the sensitivity weights of the person to red, red and blue.

[0051] It should be noted that in the calculation formula for pixel grayscale value, , as well as The acquisition is achieved through the color sensor of the existing technology installed in the high-speed camera, where the color sensor can accurately capture the intensity value of each grid unit in the red, green and blue channels, thereby ensuring the accuracy and reliability of the pixel grayscale value calculation.

[0052] In addition, it is necessary to supplement step S402.1 that the formula for obtaining the clarity includes: ;in To indicate clarity, Used to The gradient value of the position is converted into a global average statistic, Used to indicate The gradient value of the position, where , used to indicate location The horizontal gradient value of Used to indicate location Vertical gradient.

[0053] Specifically, in the present invention For grayscale value calculation, the grid units are traversed through the grayscale value calculation formula and the clarity acquisition formula to obtain the pixel grayscale value and clarity of each grid unit. On this basis, the primary grid unit and the secondary grid unit can be screened out. Further, by calculating the limit distance between the primary grid unit and the secondary grid unit, a more accurate abnormal area is constructed. This innovative method not only improves the accuracy of positioning, but also effectively reduces the cases of misjudgment and missed judgment, bringing a new solution to the field of plate defect detection.

[0054] Step S500: stacking the plate image, the strain distribution image and the transmission image, and obtaining the position of the abnormal area in the plate image and the strain distribution characteristics of the plate at the abnormal area.

[0055] It should be understood that the stacking method of the plate image, the strain distribution image and the transmission image in step S500 includes: step S501 to step S503.

[0056] Step S501: Determine a corresponding processing ratio based on the size specifications of the plate image, the strain distribution image and the transmission image.

[0057] It should be understood that the processing ratio disclosed in step S501 in the present invention is used to perform a scaling or enlargement operation on any one of the plate image, the strain distribution image and the transmission image.

[0058] Step S502: performing superposition processing on the scaled or enlarged plate image, strain distribution image and transmission image.

[0059] It should be clear that step S502 is used in the present invention to merge the processed plate image, strain distribution image and transmission image so as to simultaneously display the appearance characteristics, strain distribution and internal structure characteristics of the plate on one image, which helps to more intuitively observe and analyze the relationship between the abnormal area and the plate as a whole.

[0060] Step S503: The overlay processing method is to overlay the strain distribution image and the transmission image on the plate image using a transparency parameter.

[0061] It should be noted that the transparency parameter in step S503 is set according to the requirements, aiming to enable the image information to be displayed clearly without interfering with each other.

[0062] It should be added that, in the present invention, steps S501 to S503 are based on the images that have completed the stacking process to check for abnormal areas therein, and the abnormal areas are located in the plate image by integrating the acquired plate image, strain distribution image and transmission image. At the same time, by analyzing the strain distribution characteristics of the plate in the abnormal area and completely extracting the features, data support is provided for subsequent in-depth research on plate defects and targeted treatment measures.

[0063] Step S600: Identify the defect type of the plate according to the strain distribution characteristics.

[0064] It should be understood that the method for identifying the defect type of the plate according to the strain force distribution characteristics in step S600 includes: steps S601 to S604.

[0065] Step S601: According to the type of plate, a strain density threshold range corresponding to the defect type is obtained in the network.

[0066] Step S602: constructing a defect type feature database based on the strain force density threshold range.

[0067] Step S603: Associating the defect type feature database with the strain force distribution feature of the abnormal area.

[0068] Step S604: Determine the defect type of the plate according to the correlation result.

[0069] It should be understood that in the present invention, steps S601 to S604 realize automatic identification of plate defect types by constructing a defect type feature database and a matching relationship between the strain force characteristics of abnormal regions.

[0070] Specifically, first, according to the type of plate to be tested, the corresponding defect type and its strain density threshold range are retrieved and obtained from Internet resources. This step ensures that the data used in the subsequent analysis process is for the specific type of plate, thereby improving the accuracy of identification. Subsequently, a defect type feature database based on these strain density threshold ranges is constructed. The defect type feature database stores the correspondence between different defect types and strain characteristics, providing reliable data support for subsequent defect type identification. After obtaining the strain distribution characteristics of the abnormal area, these characteristics are correlated with the data in the defect type feature database for analysis. Through comparison and matching, the defect type of the plate is finally determined based on the correlation results. This not only realizes the automatic identification of defect types, but also greatly improves the efficiency and accuracy of identification, providing a strong technical guarantee for plate production and quality control.

[0071] It should be supplemented that the method for associating the comparison of the type feature library and the strain force distribution features at the abnormal area in step S603 includes: step S603.1-step S603.6.

[0072] Step S603.1: Obtain the total amount of strain force in the abnormal area.

[0073] Step S603.2: Measure the total area of ​​the abnormal region.

[0074] Step S603.3: Calculate the strain value per unit area according to the total area and the total amount of strain. The strain value per unit area is used to reflect the strain density of the abnormal area.

[0075] Step S603.4: Compare the strain force value per unit area with the strain force density threshold range.

[0076] Step S603.5: Determine the position information of the unit area strain force value within the strain force density threshold range according to the comparison result.

[0077] Step S603.6: Identify the corresponding defect type in the defect type feature database based on the position information.

[0078] It should be added that in step S603.1, the total amount of strain in the abnormal area is obtained through image sensor recognition. Specifically, in step S603.1, the image sensor can accurately capture the slight deformation of the plate when it is subjected to force, thereby calculating the total amount of strain. In addition, in step S603.2, the total area of ​​the abnormal area is measured using existing image processing technology, specifically by capturing the image of the surface of the plate with a high-resolution camera, and using image analysis software to automatically calculate the area of ​​the abnormal area.

[0079] It should also be added that in step S603.3, when calculating the strain value per unit area, the total strain value is divided by the total area of ​​the abnormal area, and the result is the strain value per unit area. This value can intuitively reflect the strain density of the abnormal area. In addition, in step S603.4, the calculated strain value per unit area is compared with the pre-set strain density threshold range to determine the range to which it belongs.

[0080] In addition, reference Figure 5 It can be seen that the present invention also proposes a plate defect detection system. It should be supplemented that a plate defect detection system uses the plate defect detection method disclosed above and includes: The image capture module is designed to obtain images of the sheet in the optical system, including the sheet itself, stress distribution, and transmission effects.

[0081] The image processing module is responsible for in-depth processing and analysis of the collected images.

[0082] It should be clear that the in-depth processing and analysis disclosed by the image processing module include steps such as image preprocessing, feature extraction, image segmentation and target recognition, which aim to accurately identify the defective area on the surface of the plate and further analyze the strain situation in the area. Specifically, through the image preprocessing step, the image noise can be effectively removed, the image contrast can be enhanced, and the accuracy of subsequent processing can be improved. The feature extraction step can extract the unique features of the defective area of ​​the plate, such as shape, size and color, to provide key information for subsequent defect identification. The image segmentation step divides the plate image into multiple areas, distinguishes the defective area from the normal area, and lays the foundation for subsequent strain analysis. The target recognition step is based on the results of feature extraction and image segmentation to accurately identify the type and location of defects on the surface of the plate.

[0083] The abnormal area detection module uses transmission images combined with positioning technology to identify defective areas in the plate.

[0084] Image stacking module, which effectively integrates plate image, stress distribution image and transmission image to form a unified image display; The defect type identification module determines the defect type on the plate by analyzing the stress distribution characteristics of the defect area; The user interface and interaction module provides an intuitive user interface, allowing users to clearly see the type and location of sheet defects and perform corresponding interactive operations; The data storage and management module is responsible for saving the collected image data and defect area information, while managing the identification results of defect classification to ensure the integrity and traceability of the data.

[0085] Finally, the present invention also proposes a storage medium for a plate defect detection system, wherein the storage medium for a plate defect detection system is used to carry the plate defect detection system disclosed above.

[0086] It should be noted that the storage medium in actual use is a hard disk, a USB flash drive, an optical disk, a device or apparatus that can store data. By storing the above-mentioned plate defect detection system, the computer equipment or intelligent device can be equipped with the above-mentioned plate defect detection system after reading the storage medium, and execute the above-mentioned plate defect detection method to realize the detection, analysis and management of plate defects.

[0087] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. A plate defect detection method, characterized in that: include: Capturing the plate image under the optical framework; Obtaining a strain force distribution image inside the plate based on the plate image; Acquire a transmission image of the plate; According to the positioning method, the abnormal area in the transmission image is obtained; Stacking plate images, strain distribution images and transmission images, and obtaining the position of the abnormal area in the plate image and the strain distribution characteristics of the plate at the abnormal area; Identify the defect types of the plate based on the strain distribution characteristics; The positioning method comprises: According to the specification parameters of the transmission image, a two-dimensional coordinate system is constructed in the transmission image, wherein the abscissa axis of the two-dimensional coordinate system represents the width direction of the plate, and the ordinate axis represents the length direction of the plate; In the constructed two-dimensional coordinate system, a plurality of grid units are set, each grid unit corresponds to a detection area on the plate, and the grid unit is an M×N matrix, where M is the number of rows and N is the number of columns; For different types of plates, the corresponding grayscale values ​​are retrieved and obtained from Internet resources and set as threshold parameters; Obtain the pixel gray value of each grid cell in the transmission image; Compare the ratios of different pixel gray values ​​to threshold parameters; Based on the comparison results, the grid cells whose pixel grayscale values ​​exceed the threshold parameters are identified as abnormal areas.

2. A plate defect detection method according to claim 1, characterized in that: The calculation formula for the pixel grayscale value at the grid unit is: ; Used to represent the pixel gray value at the grid unit. It is used to indicate the position of the grid unit in the two-dimensional coordinate system. Used to indicate The red channel intensity value at Used to indicate The green channel intensity value at Used to indicate The blue channel intensity value at , the coefficients 0.299, 0.587 and 0.114 are based on the sensitivity weights of the human body to red, green and blue.

3. A plate defect detection method according to claim 1, characterized in that: The positioning method comprises: According to the specification parameters of the transmission image, a two-dimensional coordinate system is constructed in the transmission image, wherein the abscissa axis of the two-dimensional coordinate system represents the width direction of the plate, and the ordinate axis represents the length direction of the plate; In the constructed two-dimensional coordinate system, a plurality of grid units are set, and an identification code is constructed in each grid unit, and the identification code is used to display the clarity of the transmission image at the grid unit; Obtain a first-level grid unit, where the first-level grid unit is the one with the clearest identification code among multiple grid units; Obtain a secondary grid unit, where the secondary grid unit is the one with the most ambiguous identification code among the multiple grid units; Get the limit distance between the primary grid unit and the secondary grid unit; The abnormal area is established based on the primary grid unit, the secondary grid unit and the limit distance.

4. A plate defect detection method according to claim 3, characterized in that: The formula for obtaining clarity includes: ;in To indicate clarity, Used to The gradient value of the position is converted into a global average statistic, Used to indicate The gradient value of the position, where , used to indicate location The horizontal gradient value of Used to indicate location Vertical gradient.

5. A plate defect detection method according to claim 1, characterized in that: The stacking methods of plate images, strain distribution images and transmission images include: Determine a corresponding processing ratio based on the size specifications of the plate image, the strain distribution image, and the transmission image, and the processing ratio is used to perform a scaling or enlargement operation on any one of the plate image, the strain distribution image, and the transmission image; Overlay processing is performed on the scaled or expanded plate image, strain distribution image and transmission image; The overlay processing method is to overlay the strain distribution image and the transmission image on the plate image with a transparency parameter. The transparency parameter is set according to the needs, aiming to make the image information clearly displayed without interfering with each other.

6. A plate defect detection method according to claim 1, characterized in that: Methods for identifying the types of defects in plates based on strain distribution characteristics include: According to the type of plate, the strain density threshold range of the corresponding defect type is obtained in the network; Construct a defect type feature database based on strain force density threshold range; Correlate the defect type feature database with the strain force distribution features of the abnormal area; According to the correlation results, the defect type of the plate is determined.

7. A plate defect detection method according to claim 6, characterized in that: The correlation method of comparing the type feature library and the strain distribution characteristics at the abnormal area includes: Obtain the total amount of strain in the abnormal area; Measure the total area of ​​the abnormal region; The strain value per unit area is calculated based on the total area and the total amount of strain force. The strain value per unit area is used to reflect the strain force density in the abnormal area. Comparing the strain force value per unit area with the strain force density threshold interval; According to the comparison result, the position information of the unit area strain force value within the strain force density threshold range is determined; According to the position information, the corresponding defect type is identified in the defect type feature database.

8. A plate defect detection system, using a plate defect detection method according to any one of claims 1 to 6, characterized in that: include: An image capture module, used to capture images of the plate under an optical framework, including a plate image, a strain distribution image, and a transmission image; An image processing module, used for processing the captured image; The abnormal region detection module detects abnormal regions based on the transmission image and using a positioning method; An image stacking module, used to stack the plate image, strain distribution image and transmission image together; A defect type identification module is used to analyze the strain force distribution characteristics of the abnormal area and determine the defect type based on the strain force distribution characteristics; The user interface and interaction module is used to provide an interactive interface between the user and the system, so that the user can intuitively view the type and location of defects on the board and perform corresponding interactive steps; The data storage and management module is used to store the captured image abnormal area information and the identification results of the defect types.

9. A storage medium for a plate defect detection system, characterized in that: The invention is equipped with a plate defect detection system as described in claim 8.

Images