Bubble detection method and system for carbon fiber prepreg production

Abstract

The invention relates to the technical field of image analysis, in particular to a bubble detection method and system for carbon fiber prepreg production, and the method comprises the steps: obtaining a target surface image corresponding to carbon fiber prepreg to be detected, generating a basically defect-free unit based on the target surface image, and shooting the basically defect-free unit in different postures; dividing a suspected bubble area and a non-bubble area; clustering is carried out on the suspected bubble areas; determining a form deviation degree corresponding to each suspected bubble area; screening out a suspected bubble region set corresponding to each suspected bubble region from the target cluster; and performing bubble trend analysis processing on each suspected bubble region based on the form deviation degree corresponding to each suspected bubble region and the suspected bubble region set, and judging whether the suspected bubble region is a real bubble region based on the interference conformity. According to the invention, by analyzing the target surface image, bubble detection is realized, and the accuracy of bubble detection is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image analysis, and in particular to a bubble detection method and system for producing carbon fiber prepreg. Background Art

[0002] Detecting bubbles in carbon fiber prepreg is often a key step in ensuring the performance of composite materials. Bubbles tend to significantly reduce the mechanical properties of composite materials, such as interlaminar shear strength and fatigue resistance, resulting in a decrease in the structural load-bearing capacity or even failure. In high-precision fields such as aerospace and automobiles, even tiny bubbles may cause stress concentration and endanger overall safety. Bubble detection can often optimize impregnation process parameters, improve resin distribution uniformity, and reduce defect formation. In addition, strict bubble quality inspection can often reduce product scrap rates and save production costs, while meeting the stringent standards for material reliability in high-end applications, and has important engineering value for ensuring the quality of end products.

[0003] At present, when performing defect detection on an object, the method usually adopted is: segmenting the defect area from the object image according to the gray value. However, when segmenting the bubble area from the carbon fiber prepreg image according to the gray value, the following technical problems often occur: Since the grayscale differences between the bubble area, local resin accumulation area and loss area of ​​the carbon fiber prepreg are often not large, if only the difference in grayscale values ​​is considered when detecting the bubble area, it may often cause misjudgment of the bubble pixels, resulting in poor accuracy of bubble detection. Summary of the invention

[0004] In order to solve the technical problem of poor accuracy of bubble detection, the present invention proposes a bubble detection method and system for carbon fiber prepreg production.

[0005] In a first aspect, the present invention provides a bubble detection method for carbon fiber prepreg production, the method comprising: Obtain a target surface image corresponding to the carbon fiber prepreg to be inspected, generate a basic defect-free unit based on the target surface image, and shoot the basic defect-free unit in different postures to obtain a possible state unit; According to the similarity between the preset window corresponding to the pixel point in the target surface image and the possible units in different states, the suspected bubble area and the non-bubble area are divided from the target surface image; Based on the non-bubble area, the suspected bubble area is clustered to obtain the target cluster; Based on the morphological differences between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs, the degree of morphological deviation corresponding to each suspected bubble region is determined; Based on the area difference between each suspected bubble region and other suspected bubble regions in its target cluster, a set of suspected bubble regions corresponding to each suspected bubble region is screened out from the target cluster; Based on the morphological deviation degree corresponding to each suspected bubble region and the set of suspected bubble regions, bubble trend analysis is performed on each suspected bubble region to obtain the interference compliance corresponding to each suspected bubble region, and based on the interference compliance, it is determined whether the suspected bubble region is a real bubble region.

[0006] Combined with the above first aspect, in a possible implementation manner, the generating of the basically defect-free unit based on the target surface image includes: Based on the target surface image, the warp and weft yarns are separated by a direction selective filter, and the cross region in the target surface image is extracted as the basic unit; After extracting the basic unit, a standard defect-free template is generated by geometric rule merging; Based on the standard defect-free template, the cross region in the standard defect-free template is extracted as the basically defect-free unit by a direction selective filter.

[0007] Combined with the above first aspect, in a possible implementation manner, the photographing of the basically defect-free unit in different postures to obtain the state possible units includes: By adjusting the Euler angles, photographing of the basically defect-free unit in different postures is realized, and the state possible units corresponding to different Euler angles are obtained.

[0008] Combined with the above first aspect, in a possible implementation manner, the dividing of the suspected bubble regions and non-bubble regions from the target surface image according to the similarity between the preset window corresponding to the pixel point in the target surface image and different state possible units includes: The NCC coefficient between the preset window corresponding to each pixel point in the target surface image and each state possible unit is determined as the target similarity between each pixel point and each state possible unit; The state possible unit with the largest target similarity with the pixel point is screened out from all state possible units as the target matching unit corresponding to the pixel point; The continuous pixel points with the same target matching unit in the target surface image form a matching analysis region; According to the difference situation between each matching analysis region and its adjacent matching analysis regions, the abnormal distortion index corresponding to each matching analysis region is determined; If the abnormal distortion index corresponding to the matching analysis region is greater than the preset abnormal distortion threshold, the matching analysis region is determined as a suspected bubble region; If the abnormal distortion index corresponding to the matching analysis region is less than or equal to the preset abnormal distortion threshold, the matching analysis region is determined as a non-bubble region.

[0009] Combined with the first aspect above, in a possible implementation manner, the determining the abnormal distortion index corresponding to each matching analysis region according to the difference situation between each matching analysis region and its adjacent matching analysis region includes: Determine the Euler angles corresponding to the target matching units of the pixel points in each matching analysis region as the target Euler angles corresponding to each matching analysis region; Determine any one matching analysis region as a marked region, and determine each matching analysis region adjacent to the marked region as a reference region; Determine the difference between the target Euler angle corresponding to the marked region and the target Euler angles corresponding to each reference region as the target difference between the marked region and each reference region; Determine the angle difference index corresponding to the marked region according to the target differences between the marked region and all reference regions, where the target difference and the angle difference index have a positive correlation; Determine the abnormal distortion index corresponding to the marked region according to the target similarity between all pixel points in the marked region and their corresponding target matching units, and the angle difference index corresponding to the marked region, where the target similarity and the abnormal distortion index have a negative correlation, and the angle difference index and the abnormal distortion index have a positive correlation.

[0010] Combined with the first aspect above, in a possible implementation manner, the clustering the suspected bubble regions based on the non-bubble regions to obtain target clusters includes: Determine each edge pixel point of each non-bubble region as a marked pixel point, perform template matching with the target matching unit corresponding to the marked pixel point as a matching template, and obtain the target matching region corresponding to the marked pixel point; Update the target matching units corresponding to all pixel points in the target matching region corresponding to the marked pixel point to the target matching unit corresponding to the marked pixel point; Determine the region formed by the target matching regions corresponding to all marked pixel points as a reference region, and determine each pixel point in the reference region as a reference point; Determine the region formed by all pixel points except the reference points in each suspected bubble region as a candidate region; Record the reference points and the pixel points in the non-bubble regions as temporary pixel points, and determine the mean value of the Euler angles corresponding to the latest target matching units of all temporary pixel points in the preset window corresponding to each edge pixel point of each candidate region as the Euler representative angle corresponding to each edge pixel point of each candidate region; Obtain the state possible units corresponding to the Euler representative angles of each edge pixel point of each candidate region as the target possible units corresponding to each edge pixel point of each candidate region; Perform template matching with the target possible units corresponding to the edge pixel points of the candidate region as the matching templates to obtain the target matching regions corresponding to the edge pixel points of the candidate region; Update the target matching units corresponding to all pixel points within the target matching regions corresponding to the edge pixel points of the candidate region to the target possible units corresponding to this edge pixel point; Determine the region composed of the target matching regions corresponding to all edge pixel points of all candidate regions as the temporary region, and determine each pixel point within the temporary region as a temporary point; Update the candidate region to the region composed of all pixel points within it except the temporary points, and repeat the candidate region update step until the finally updated candidate region is empty, so as to realize the update of the target matching units corresponding to each pixel point within each suspected bubble region; Cluster the updated target matching units corresponding to all pixel points within all suspected bubble regions, and determine each obtained clustering cluster as a reference cluster; Determine the number of target matching units belonging to each reference cluster among the updated target matching units corresponding to all pixel points within each suspected bubble region as the overall attribution number between each suspected bubble region and each reference cluster; Screen out the reference cluster with the largest overall attribution number between the suspected bubble region from all reference clusters as the best reference cluster corresponding to the suspected bubble region; Divide the suspected bubble regions with the same corresponding best reference cluster into the same target cluster.

[0011] Combined with the above first aspect, in a possible implementation manner, determining the morphological offset degree corresponding to each suspected bubble region based on the morphological difference situation between each suspected bubble region and other suspected bubble regions in its target cluster includes: Determine the distance between every two edge pixel points of each suspected bubble region as the reference distance to obtain the reference distance set corresponding to each suspected bubble region; Determine the maximum value in the reference distance set corresponding to each suspected bubble region as the target long side representative distance corresponding to each suspected bubble region; Determine the line segment connecting the two edge pixel points corresponding to the largest reference distance in the reference distance set corresponding to each suspected bubble region as the target line segment corresponding to each suspected bubble region; Determine the angle corresponding to the target line segment corresponding to each suspected bubble region as the target representative angle corresponding to each suspected bubble region; For any point within the suspected bubble region, draw a perpendicular line to the target line segment corresponding to the suspected bubble region, denoted as the target perpendicular line corresponding to the suspected bubble region. Slide the target perpendicular line on the suspected bubble region, and denote the line segment obtained by the intersection of the target perpendicular line and the boundary of the suspected bubble region during the sliding process as the candidate line segment, and obtain the candidate line segment set corresponding to the suspected bubble region; Determine the maximum value in the candidate line segment set corresponding to each suspected bubble region as the target short side representative distance corresponding to each suspected bubble region; Determine the ratio of the target long side representative distance to the target short side representative distance corresponding to each suspected bubble region as the shape feature factor corresponding to each suspected bubble region; Based on the difference between the shape feature factor corresponding to each suspected bubble region and the shape feature factors corresponding to other suspected bubble regions in its target cluster, and the difference between the target representative angle corresponding to each suspected bubble region and the target representative angles corresponding to other suspected bubble regions in its target cluster, determine the morphological deviation degree corresponding to each suspected bubble region.

[0012] Combined with the above first aspect, in a possible implementation manner, the screening of the suspected bubble region set corresponding to each suspected bubble region from the target cluster based on the area difference situation between each suspected bubble region and other suspected bubble regions in its target cluster includes: Determine any one suspected bubble region as the marked suspected region, and determine each suspected bubble region in the target cluster to which the marked suspected region belongs except the marked suspected region as the reference suspected region; Determine the generation consistency factor between the marked suspected region and each reference suspected region according to the distance and area difference between the marked suspected region and each reference suspected region; If the generation consistency factor between the marked suspected region and the reference suspected region is greater than the preset consistency threshold, then determine the reference suspected region as the generation consistent region; The marked suspected region and all the generation consistent regions form the suspected bubble region set corresponding to the marked suspected region.

[0013] Combined with the above first aspect, in a possible implementation manner, the bubble trend analysis process for each suspected bubble region based on the morphological deviation degree and the suspected bubble region set corresponding to each suspected bubble region to obtain the interference compliance corresponding to each suspected bubble region includes: Determine any one suspected bubble region as the marked suspected region, and perform linear fitting on the central points of all suspected bubble regions in the suspected bubble region set corresponding to the marked suspected region to obtain the target line; Draw a perpendicular line to the target straight line through the center point of the suspected marked area, and record it as a reference perpendicular line; Taking the reference vertical line as a dividing line, dividing the set of suspected bubble regions corresponding to the suspected marked regions into two subsets; The average value of the area of ​​all suspected bubble regions in each subset is determined as the area representative factor corresponding to each subset; The subset with the larger area representative factor among the two subsets is determined as the set far from the fiber, and the subset with the smaller area representative factor among the two subsets is determined as the set close to the fiber; Determine the proportion of the suspected bubble regions in the remote fiber set whose corresponding area is larger than the area of ​​the suspected marked region as the first target proportion corresponding to the suspected marked region; Determine the proportion of the suspected bubble area whose corresponding area in the near-fiber set is smaller than the area of ​​the suspected marked area as the second target proportion corresponding to the suspected marked area; Determine the distance from the center point of the suspected mark area to the target straight line as the mark distance corresponding to the suspected mark area; The interference compliance corresponding to the marked suspected area is determined according to the degree of morphological deviation, the first target proportion, the second target proportion and the marking distance corresponding to the marked suspected area, wherein the degree of morphological deviation and the marking distance are positively correlated with the interference compliance, and the first target proportion and the second target proportion are negatively correlated with the interference compliance.

[0014] In a second aspect, the present invention provides a bubble detection system for carbon fiber prepreg production, comprising a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement the method in the first aspect or any possible implementation of the first aspect. Specifically, the system comprises: The image and unit acquisition module is used to acquire the target surface image corresponding to the carbon fiber prepreg to be inspected, generate a basic defect-free unit based on the target surface image, and shoot the basic defect-free unit in different postures to obtain a possible state unit; A region division module is used to divide the target surface image into suspected bubble regions and non-bubble regions according to the similarity between the preset window corresponding to the pixel point in the target surface image and the possible units in different states; The regional clustering module is used to cluster the suspected bubble areas based on the non-bubble areas to obtain the target clusters; A morphological deviation degree determination module is used to determine the morphological deviation degree corresponding to each suspected bubble region based on the morphological difference between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs; An area screening module, configured to screen out a set of suspected bubble areas corresponding to each suspected bubble area from a target cluster based on the area difference between each suspected bubble area and other suspected bubble areas in the target cluster to which it belongs; An analysis, processing and judgment module, configured to perform bubble trend analysis and processing on each suspected bubble area based on the morphological deviation degree and the set of suspected bubble areas corresponding to each suspected bubble area, obtain the interference compliance degree corresponding to each suspected bubble area, and determine whether the suspected bubble area is a real bubble area based on the interference compliance degree.

[0015] In a third aspect, a server is provided, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the device executes the method in the first aspect or any possible implementation manner of the first aspect.

[0016] In a fourth aspect, a computer program product is provided, which includes: computer program code, when the computer program code runs on a computer, the computer is enabled to execute the method in the first aspect or any possible implementation manner of the first aspect.

[0017] In a fifth aspect, a computer-readable storage medium is provided, which stores computer program code, and when the computer program code runs on a computer, the computer is enabled to execute the method in the first aspect or any possible implementation manner of the first aspect.

[0018] The present invention has the following beneficial effects: A bubble detection method for carbon fiber prepreg production according to the present invention realizes bubble detection by analyzing a target surface image, solves the technical problem of poor accuracy of bubble detection, and improves the accuracy of bubble detection. Specifically, when performing bubble detection, the present invention comprehensively considers multiple law characteristics related to bubbles, such as morphological deviation degree and interference compliance degree, etc., so as to relatively objectively realize bubble detection and improve the accuracy of bubble detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.

[0020] Figure 1Flow chart of a method for detecting air bubbles in the production of carbon fiber prepreg according to the present invention; Figure 2 Schematic diagram of the composition structure of a system for detecting air bubbles in the production of carbon fiber prepreg according to the present invention; Figure 3 Schematic diagram of the structure of a computer device according to the present invention. Specific embodiments

[0021] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific embodiments, structures, features and their effects of the technical solutions proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0023] Detecting air bubbles in carbon fiber prepreg is a key link to ensure the performance of composite materials. Air bubbles will significantly reduce the mechanical properties of composite materials, such as interlaminar shear strength and fatigue resistance, resulting in a decrease in the structural load-bearing capacity and even failure. In high-precision fields such as aerospace and automotive, tiny air bubbles may cause stress concentration and endanger the overall safety. Through detection, the impregnation process parameters can be optimized, the resin distribution uniformity can be improved, and the formation of defects can be reduced. In addition, strict quality inspection can reduce the product rejection rate, save production costs, and at the same time meet the stringent standards for material reliability in high-end applications, which has important engineering value for ensuring the quality of end products. The present invention analyzes the causes of air bubbles in carbon fiber prepreg, as well as the morphological and distribution changes caused by resin flow, judges the suspected air bubble areas, excludes the interference of other situations, and obtains accurate air bubble areas.

[0024] Refer to Figure 1 , which shows the flow of some embodiments of a method for detecting air bubbles in the production of carbon fiber prepreg according to the present invention. The method for detecting air bubbles in the production of carbon fiber prepreg includes the following steps: Step S1, obtain the target surface image corresponding to the carbon fiber prepreg to be detected, generate a basically defect-free unit based on the target surface image, and take pictures of the basically defect-free unit in different postures to obtain state-possible units.

[0025] Among them, the carbon fiber prepreg to be detected can be the carbon fiber prepreg to be subjected to bubble detection. The carbon fiber prepreg is formed by compounding epoxy resin on carbon fiber through high-pressure and high-temperature technology. A composite material made of materials such as carbon fiber yarn, epoxy resin, and release paper through processes such as coating, hot pressing, cooling, film laminating, and coiling is called carbon fiber prepreg, also known as carbon fiber prepreg cloth. The target surface image can be the surface image of the carbon fiber prepreg to be detected.

[0026] As an example, this step may include the following steps: The first step, obtaining the target surface image corresponding to the carbon fiber prepreg to be detected may include the following sub-steps: The first sub-step, through a line array camera, collect the surface image of the carbon fiber prepreg to be detected to obtain a high-resolution image.

[0027] It should be noted that an industrial-grade line array camera with a resolution ≥ 8k can be selected to ensure that the single-line pixel density meets the microscopic detection requirements of carbon fiber texture, such as clearly resolving carbon fibers with a diameter ≤ 7μm.

[0028] The second sub-step, through a polarization camera, collect images of the carbon fiber prepreg to be detected in different polarization directions to obtain multi-angle images.

[0029] It should be noted that a polarization camera equipped with a micro-polarizer array can be used to simultaneously capture images in four polarization directions of 0°, 45°, 90°, and 135° to obtain three-channel information of the degree of linear polarization (DOLP), angle of polarization (AOP), and light intensity. By sending a signal through an encoder, the line array camera and the polarization camera perform synchronous acquisition of images.

[0030] The third sub-step, fuse the high-resolution image and the multi-angle image to obtain the target surface image.

[0031] It should be noted that the high-resolution image of the linear array camera and the multi-angle images of the polarization camera can be spatially aligned by using a feature point matching algorithm to ensure pixel-level correspondence. Among them, the feature point matching algorithm can be, but is not limited to, the SIFT (Scale-Invariant Feature Transform) algorithm. The image acquisition times of the linear array camera and the polarization camera are strictly ensured to be the same through timestamp verification. The obtained images are decomposed into a multi-scale space by using the nonsubsampled contourlet transform (NSCT). The texture energy information of the linear array camera is retained in the low-frequency sub-band, and the edge details of the polarization image are fused in the high-frequency sub-band. The images obtained by the linear array camera and the polarization camera are fused to finally obtain an enhanced planar image of the surface of the carbon fiber prepreg to be detected, denoted as the target surface image.

[0032] In the second step, based on the above target surface image, the warp and weft yarns are separated by a direction selective filter, and the cross region in the above target surface image is extracted as the basic unit.

[0033] For example, first, a gray threshold can be set according to the gray value to binarize the target surface image. Among them, the gray threshold can be a pre-set threshold, mainly used for image binarization, and it can be 80. Then, the weft yarns can be horizontally segmented and the warp yarns can be vertically segmented to extract the cross blocks as texture primitives, that is, the basic units.

[0034] In the third step, after the basic units are extracted, a standard defect-free template is generated by merging according to geometric rules. This method can retain texture features and simplify data.

[0035] In the fourth step, based on the standard defect-free template, the cross region in the above standard defect-free template is extracted as the basic defect-free unit by a direction selective filter.

[0036] For example, a multi-directional Gabor filter is used to extract texture feature vectors, and then classified by the fuzzy C-means algorithm to identify the basic units of the prepreg pattern, denoted as the basic defect-free units. Among them, the direction selective filter can be, but is not limited to, a multi-directional Gabor filter.

[0037] It should be noted that the pattern of the carbon fiber prepreg itself often has certain texture rules, and the texture changes have a certain degree of repeatability. When impregnating the fibers with resin, bubbles may be generated at multiple process stages of the operation. The presence of bubbles often causes distortion of the fiber image in some areas of the image, damaging the regularity of the pattern. Therefore, the suspected bubble areas can be screened based on the damage of the pattern in the obtained surface image of the prepreg. The basic defect-free unit can often represent the smallest repeating unit of the pattern when no distortion occurs.

[0038] The fifth step is to take pictures of the basic defect-free unit in different postures by adjusting the Euler angles, and obtain the state possible units corresponding to different Euler angles.

[0039] Among them, the state possible unit can be the basic defect-free unit under different Euler angles.

[0040] It should be noted that the basic unit of the pattern is photographed in different postures, and the basic unit images in different postures are obtained as the state possible units. Specifically, described by Euler angles, Euler angles are one of the most intuitive and commonly used methods to describe the posture of an object, and it is adjusted by three rotation angles. Euler angles usually include the yaw angle (Yaw) around the horizontal axis, the pitch angle (Pitch) around the vertical axis, and the roll angle around the vertical axis.

[0041] It should be noted that different state possible units can represent the smallest repeating unit in different postures.

[0042] Step S2, according to the similarity between the preset window corresponding to the pixel point in the target surface image and different state possible units, divide the suspected bubble area and the non-bubble area from the target surface image.

[0043] Among them, the preset window can be a pre-set window, and it can be a 7×7 window.

[0044] As an example, this step may include the following steps: The first step is to determine the NCC (Normalized Cross Correlation) coefficient between the preset window corresponding to each pixel point in the above target surface image and each state possible unit as the target similarity between each pixel point and each state possible unit.

[0045] It should be noted that the matching result of the normalized cross correlation (NCC) is represented by a correlation coefficient matrix, and the coefficient value range is [-1, 1].

[0046] The second step is to screen out the state possible unit with the largest target similarity with the pixel point from all state possible units as the target matching unit corresponding to the pixel point.

[0047] It should be noted that the target matching unit corresponding to a pixel point can, to a certain extent, represent the pose of the pixel point.

[0048] In the third step, consecutive pixel points with the same target matching unit in the above target surface image are used to form a matching analysis region.

[0049] Among them, the target matching units corresponding to all pixel points within the matching analysis region can be the same.

[0050] For example, if the target matching units corresponding to two adjacent pixel points in the target surface image are the same, these two adjacent pixel points can be divided into the same matching analysis region, thereby enabling the division of all adjacent pixel points in the target surface image, and further enabling the division of all pixel points in the target surface image.

[0051] It should be noted that the matching analysis region can represent positions with the same pose.

[0052] In the fourth step, determining the abnormal distortion index corresponding to each matching analysis region according to the difference between each matching analysis region and its adjacent matching analysis region can include the following sub-steps: In the first sub-step, the Euler angles corresponding to the target matching units of the pixel points within each matching analysis region are determined as the target Euler angles corresponding to each matching analysis region.

[0053] In the second sub-step, any one matching analysis region is determined as the marked region, and each matching analysis region adjacent to the above marked region is determined as the reference region.

[0054] It should be noted that the method for obtaining the matching analysis regions adjacent to the marked region can be: each matching analysis region except the marked region is determined as the calibration region, the union of the preset neighborhoods corresponding to all edge pixel points of the marked region is determined as the region around the mark, and the matching analysis regions having intersections with the region around the mark are recorded as the reference regions. Among them, the preset neighborhood can be a pre-set neighborhood, which can be a 5×5 neighborhood.

[0055] In the third sub-step, the difference between the target Euler angle corresponding to the above marked region and the target Euler angles corresponding to each reference region is determined as the target difference between the marked region and each reference region.

[0056] It should be noted that the greater the target difference between the marked region and the reference region, the more likely it indicates that the poses between the marked region and the reference region are more different.

[0057] For example, the formula for determining the target difference between the marked region and the reference region can be: ; wherein, is the target difference between the marked area and the i-th reference area. i is the serial number of the reference area. is the absolute value function. is the yaw angle included in the target Euler angles corresponding to the marked area. is the yaw angle included in the target Euler angles corresponding to the i-th reference area. is the pitch angle included in the target Euler angles corresponding to the marked area. is the pitch angle included in the target Euler angles corresponding to the i-th reference area. is the roll angle included in the target Euler angles corresponding to the marked area. is the roll angle included in the target Euler angles corresponding to the i-th reference area.

[0058] The fourth sub-step is to determine the angle difference index corresponding to the marked area according to the target differences between the marked area and all reference areas.

[0059] wherein, the target difference can have a positive correlation with the angle difference index.

[0060] It should be noted that when the angle difference index corresponding to the marked area is larger, it often indicates that the posture between the marked area and most reference areas is more different, often indicates that the posture between the marked area and the surrounding areas is more different, often indicates that there is more disconnection between the marked area and the surrounding areas, and often indicates that the marked area is more likely to be a bubble area that destroys the regular distribution of the pattern.

[0061] For example, the formula for determining the angle difference index corresponding to the marked area can be: ; wherein, B is the angle difference index corresponding to the marked area. n is the number of reference areas. i is the serial number of the reference area. is the target difference between the marked area and the i-th reference area.

[0062] The fifth sub-step is to determine the abnormal distortion index corresponding to the marked area according to the target similarity between all pixel points in the marked area and their corresponding target matching units, and the angle difference index corresponding to the marked area.

[0063] wherein, the target similarity can have a negative correlation with the abnormal distortion index. The angle difference index can have a positive correlation with the abnormal distortion index.

[0064] For example, the formula for determining the abnormal distortion index corresponding to the marked area can be: ; wherein, Q is the abnormal distortion index corresponding to the marked area. is a normalization function. B is the angular difference index corresponding to the marked area. is the exponential function with the natural constant as the base. N is the number of pixel points in the marked area. j is the serial number of the pixel points in the marked area. is the target similarity between the j-th pixel point in the marked area and its corresponding target matching unit.

[0065] It should be noted that when B is larger, it often indicates that the marked area is more likely to be a bubble area that destroys the regular distribution of the pattern. When is larger, it often indicates that the pixel points in the marked area match their corresponding target matching units better, and it often indicates that the pixel points in the marked area are more likely to be normal pixel points without bubbles. Therefore, when Q is larger, it often indicates that the marked area is more likely to be a bubble area.

[0066] Step 5, if the abnormal distortion index corresponding to the matching analysis area is greater than the preset abnormal distortion threshold, then determine the matching analysis area as a suspected bubble area.

[0067] Among them, the preset abnormal distortion threshold can be the maximum abnormal distortion index allowed when it is considered that no bubbles occur in the matching analysis area, and it can be 0.7.

[0068] Step 6, if the abnormal distortion index corresponding to the matching analysis area is less than or equal to the preset abnormal distortion threshold, then determine the matching analysis area as a non-bubble area.

[0069] Step S3, cluster the suspected bubble areas based on the non-bubble areas to obtain the target clusters.

[0070] As an example, this step may include the following steps: Step 1, determine each edge pixel point of each non-bubble area as a marked pixel point, perform template matching with the target matching unit corresponding to the marked pixel point as the matching template, and obtain the target matching area corresponding to the marked pixel point.

[0071] Among them, the target matching area can be the area that matches the target matching unit, that is, the area that is the same as the target matching unit.

[0072] Step 2, update the target matching units corresponding to all pixel points in the target matching area corresponding to the marked pixel point to the target matching unit corresponding to the marked pixel point.

[0073] It should be noted that in actual situations, since the generation of bubbles often disrupts the regularity of the pattern, the generation of bubbles often changes the postures of pixel points. The target matching unit corresponding to the marked pixel point can represent the posture of the pixel point without bubbles. The target matching region corresponding to the marked pixel point can represent the matching region with the same posture as the marked pixel point when there are no bubbles. Therefore, when there are no bubbles, the postures corresponding to all pixel points within the target matching region often are the same as the posture of the marked pixel point. Therefore, the updated target matching unit corresponding to the pixel point can, to a certain extent, represent the posture that the pixel point should exhibit when there are no bubble defects.

[0074] In the third step, the region formed by the target matching regions corresponding to all marked pixel points is determined as the reference region, and each pixel point within the above-mentioned reference region is determined as a reference point.

[0075] It should be noted that the updated target matching units corresponding to all pixel points within the reference region can, to a certain extent, represent the postures that these pixel points should exhibit when there are no bubble defects.

[0076] In the fourth step, the region formed by all pixel points except the reference points within each suspected bubble region is determined as the candidate region.

[0077] In the fifth step, the reference points and the pixel points within the non-bubble region are denoted as temporary pixel points, and the mean value of the Euler angles corresponding to the latest target matching units of all temporary pixel points within the preset window corresponding to each edge pixel point of each candidate region is determined as the Euler representative angle corresponding to each edge pixel point of each candidate region.

[0078] It should be noted that in actual situations, since the pattern of the carbon fiber prepreg has a certain regularity, the pixel postures within a certain region often are similar to the postures of surrounding pixels when there are no defects. Therefore, the Euler representative angle corresponding to the edge pixel point of the candidate region can, to a certain extent, represent the angle corresponding to the posture of this edge pixel point.

[0079] For example, the formula for determining the Euler representative angle corresponding to the edge pixel point can be: ; where is the yaw angle included in the Euler representative angle corresponding to the edge pixel point. is the pitch angle included in the Euler representative angle corresponding to the edge pixel point. is the roll angle included in the Euler representative angle corresponding to the edge pixel point. m is the number of temporary pixel points within the preset window corresponding to the edge pixel point. a is the serial number of the temporary pixel point within the preset window corresponding to the edge pixel point. is the yaw angle included in the Euler angles corresponding to the latest target matching unit corresponding to the a-th temporary pixel point within the preset window corresponding to the edge pixel point. is the pitch angle included in the Euler angles corresponding to the latest target matching unit corresponding to the a-th temporary pixel point within the preset window corresponding to the edge pixel point. is the roll angle included in the Euler angles corresponding to the latest target matching unit corresponding to the a-th temporary pixel point within the preset window corresponding to the edge pixel point.

[0080] Step 6: Obtain the state possible units corresponding to the Euler representative angles of each edge pixel point of each candidate region as the target possible units corresponding to each edge pixel point of each candidate region.

[0081] Among them, the state possible unit corresponding to the Euler representative angle can be the basic defect-free unit under the Euler representative angle.

[0082] It should be noted that the target possible unit corresponding to the edge pixel point of the candidate region can, to a certain extent, characterize the attitude that the edge pixel point should exhibit when no bubble defect occurs.

[0083] Step 7: Perform template matching with the target possible unit corresponding to the edge pixel point of the candidate region as the matching template to obtain the target matching region corresponding to the edge pixel point of the candidate region.

[0084] Among them, the target matching region corresponding to the edge pixel point can be the region that matches the target possible unit corresponding to the edge pixel point, that is, the region that is the same as the target possible unit corresponding to the edge pixel point.

[0085] Step 8: Update the target matching units corresponding to all pixel points within the target matching region corresponding to the edge pixel point of the candidate region to the target possible unit corresponding to the edge pixel point.

[0086] Step 9: Determine the region composed of the target matching regions corresponding to all edge pixel points of all candidate regions as the temporary region, and determine each pixel point within the above temporary region as a temporary point.

[0087] Step 10: Update the candidate region to the region composed of all pixel points within it except the temporary points, and repeat the candidate region update step until the finally updated candidate region is empty, so as to realize the update of the target matching units corresponding to each pixel point within each suspected bubble region.

[0088] It should be noted that, based on the target matching units corresponding to the pixel points in the non-bubble region, the update of the target matching units corresponding to the pixel points in the suspected bubble region is realized, and the updated target matching units corresponding to the pixel points in the suspected bubble region are obtained. Moreover, the updated target matching units corresponding to the pixel points can, to a certain extent, characterize the posture that the pixel points should exhibit when no bubble defect occurs.

[0089] Among them, the candidate region update step may include the following sub-steps: The first sub-step: Obtain the state possible units corresponding to the Euler representative angles of each edge pixel point of each latest candidate region as the target possible units corresponding to each edge pixel point of each latest candidate region.

[0090] Among them, the method for obtaining the Euler representative angle corresponding to the edge pixel point of the latest candidate region may be: Denote each pixel point in the target surface image except for all the latest candidate regions as a reference point; Determine the mean value of the Euler angles corresponding to the latest target matching units of all the reference points within the preset window corresponding to the edge pixel point of the latest candidate region as the Euler representative angle corresponding to this edge pixel point.

[0091] The second sub-step: Use the target possible units corresponding to the edge pixel points of the latest candidate region as a matching template to perform template matching to obtain the target matching regions corresponding to the edge pixel points of the latest candidate region.

[0092] The third sub-step: Update the target matching units corresponding to all the pixel points within the target matching regions corresponding to the edge pixel points of the latest candidate region to the target possible units corresponding to this edge pixel point.

[0093] The fourth sub-step: Update the temporary region to the region composed of the target matching regions corresponding to all the edge pixel points of all the latest candidate regions, and update the temporary point to the pixel point within the latest temporary region.

[0094] The fifth sub-step: Update the candidate region to the region composed of all the pixel points within it except for the latest temporary point.

[0095] The eleventh step: Through the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm, cluster the updated target matching units corresponding to all the pixel points within all the suspected bubble regions, and determine each obtained cluster as a reference cluster.

[0096] In the twelfth step, the number of target matching units belonging to each reference cluster among the updated target matching units corresponding to all pixels in each suspected bubble area is determined as the overall belonging number between each suspected bubble area and each reference cluster.

[0097] In the thirteenth step, the reference cluster with the largest number of overall affiliations with the suspected bubble area is selected from all reference clusters as the best reference cluster corresponding to the suspected bubble area.

[0098] In the fourteenth step, the suspected bubble areas with the same corresponding best reference cluster are divided into the same target cluster.

[0099] It should be noted that the distribution trends of the suspected bubble areas in the same target cluster are often similar when the influence of bubbles is excluded, which often indicates that the suspected bubble areas in the same target cluster may be located in the same part of the carbon fiber prepreg, which often indicates that the suspected bubble areas in the same target cluster can be analyzed as a whole.

[0100] Step S4, determining the degree of morphological deviation corresponding to each suspected bubble region based on the morphological differences between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs.

[0101] It should be noted that bubbles in the same part tend to have similar shapes due to the same reasons. For example, the components in the resin that are not fully volatilized expand under heat during the curing process to form bubbles. When the resin flows along the fiber axis, the bubbles are stretched into long strips; while when flowing horizontally, the bubbles tend to be closer to a circle. Among them, the components in the resin that are not fully volatilized can be water. Therefore, the resin gaps or resin accumulation interferences similar to bubbles can be judged and eliminated by the difference in the shapes of the suspected bubble areas in a single target cluster.

[0102] As an example, this step may include the following steps: In the first step, the distance between every two edge pixels of each suspected bubble area is determined as the reference distance, and a reference distance set corresponding to each suspected bubble area is obtained.

[0103] The reference distance set corresponding to the suspected bubble region may include: the distances between all different edge pixel points of the suspected bubble region.

[0104] In the second step, the maximum value in the reference distance set corresponding to each suspected bubble area is determined as the representative distance of the target long side corresponding to each suspected bubble area.

[0105] In the third step, the line segment connecting the two edge pixels corresponding to the largest reference distance in the reference distance set corresponding to each suspected bubble area is determined as the target line segment corresponding to each suspected bubble area.

[0106] Among them, the length of the target line segment corresponding to the suspected bubble area can be equal to the maximum reference distance in the reference distance set corresponding to the suspected bubble area.

[0107] In the fourth step, the angle corresponding to the target line segment of each suspected bubble area is determined as the target representative angle corresponding to each suspected bubble area.

[0108] Among them, the value range of the angle corresponding to the target line segment can be [0°, 360°]. The angle corresponding to the target line segment can be equal to: the included angle formed by rotating counterclockwise from the horizontal right direction to the target line segment.

[0109] In the fifth step, through any point in the suspected bubble area, a perpendicular line to the target line segment corresponding to the suspected bubble area is drawn, denoted as the target perpendicular line corresponding to the suspected bubble area, and the target perpendicular line is slid on the suspected bubble area. The line segment obtained by the intersection of the target perpendicular line and the boundary of the suspected bubble area during the sliding process is denoted as the candidate line segment, and the candidate line segment set corresponding to the suspected bubble area is obtained.

[0110] It should be noted that the candidate line segment can be: the line segment formed by taking the two intersection points between the target perpendicular line and the boundary of the suspected bubble area as endpoints.

[0111] In the sixth step, the maximum value in the candidate line segment set corresponding to each suspected bubble area is determined as the target short side representative distance corresponding to each suspected bubble area.

[0112] In the seventh step, the ratio of the target long side representative distance to the target short side representative distance corresponding to each suspected bubble area is determined as the shape feature factor corresponding to each suspected bubble area.

[0113] In the eighth step, according to the difference between the shape feature factor corresponding to each suspected bubble area and the shape feature factors corresponding to other suspected bubble areas in its target cluster, and the difference between the target representative angle corresponding to each suspected bubble area and the target representative angles corresponding to other suspected bubble areas in its target cluster, the morphological deviation degree corresponding to each suspected bubble area is determined.

[0114] For example, the formula for determining the morphological deviation degree corresponding to the suspected bubble area can be: ; among them, is the morphological deviation degree corresponding to the f-th suspected bubble area in the h-th target cluster. h is the serial number of the target cluster. f and x are the serial numbers of different suspected bubble areas in the h-th target cluster. is the number of suspected bubble areas in the h-th target cluster. is the absolute value function. is the shape feature factor corresponding to the fth suspected bubble region in the hth target cluster. is the shape feature factor corresponding to the xth suspected bubble region in the hth target cluster. is the target representative angle corresponding to the fth suspected bubble region in the hth target cluster. is the target representative angle corresponding to the xth suspected bubble region in the hth target cluster.

[0115] It should be noted that when When it is larger, it often means that the shapes and change trends of other suspected bubble regions in the hth target cluster are less similar to the fth suspected bubble region, and the fth suspected bubble region is less similar to other suspected bubble regions in the same position. It often means that the fth suspected bubble region is more likely to be a resin gap or resin accumulation interference region similar to bubbles.

[0116] Step S5, based on the area difference between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs, a set of suspected bubble regions corresponding to each suspected bubble region is screened out from the target cluster.

[0117] It should be noted that due to problems with the weaving process, the finished product may have insufficient tension or inconsistent gaps, which may result in bubbles being formed in subsequent processing. The bubbles formed may have different sizes due to the different gap sizes. Therefore, the embodiment of the present invention analyzes whether two bubbles are generated in the same gap or in a relatively close gap based on the difference in the size of the bubbles formed and the distance between them.

[0118] As an example, this step may include the following steps: In the first step, any suspected bubble region is determined as a marked suspected region, and each suspected bubble region in the target cluster to which the marked suspected region belongs, except for the marked suspected region, is determined as a reference suspected region.

[0119] In the second step, the consistency factor between the above-mentioned marked suspected area and each reference suspected area is determined based on the distance and area difference between the above-mentioned marked suspected area and each reference suspected area.

[0120] For example, the formula corresponding to the generated consistency factor between the marked suspected area and the reference suspected area can be: ;in, Is to mark the suspected area and The consistency factor between the reference suspected regions. It is the serial number of the reference suspected area. It is an exponential function with a natural constant as its base. is the absolute value function. S is the area of the marked suspected region. is the area of the th reference suspected region. is the distance between the marked suspected region and the th reference suspected region, which can be equal to the distance between the center point of the marked suspected region and the center point of the

[0121] It should be noted that when is larger, it often indicates that the area difference between the marked suspected region and the th reference suspected region is larger. When is larger, it often indicates that the distance between the marked suspected region and the th reference suspected region is larger. Therefore, when is larger, it often indicates that the distance between the marked suspected region and the th reference suspected region is smaller, and the area difference is smaller; it often indicates that the marked suspected region and the th reference suspected region are more likely to be located in the same gap or adjacent gaps.

[0122] In the third step, if the generated consistency factor between the above-mentioned marked suspected region and the reference suspected region is greater than the preset consistency threshold, the reference suspected region is determined as the generated consistent region.

[0123] Among them, the preset consistency threshold can be the maximum generated consistency factor allowed when it is considered that the regions are located in different gaps or non-adjacent gaps, and it can be 0.7.

[0124] In the fourth step, the above-mentioned marked suspected region and all generated consistent regions form the set of suspected bubble regions corresponding to the marked suspected region.

[0125] It should be noted that the gaps where the suspected bubble regions in the set of suspected bubble regions corresponding to the marked suspected region are located are often the same as or close to the gap where the marked suspected region is located.

[0126] Step S6: Based on the morphological deviation degree corresponding to each suspected bubble region and the set of suspected bubble regions, perform bubble trend analysis processing on each suspected bubble region to obtain the interference compliance corresponding to each suspected bubble region, and based on the interference compliance, determine whether the suspected bubble region is a real bubble region.

[0127] It should be noted that during the resin impregnation process, in the area close to the fibers, due to the dense fiber bundles, the resin flow is blocked, the micro-impregnation is difficult, and air bubbles are likely to be retained. In the area far from the fibers, the macroscopic resin flow is smoother, and there is more space for air bubbles to aggregate or merge. Therefore, the size of the air bubbles usually gradually increases from the area close to the fibers to the area far from the fibers. The distribution of air bubbles generated in the same fiber gap is affected by the resin flow and usually shows a linear distribution. Thus, it is possible to judge the conformity of air bubbles in the suspected air bubble area by the size distribution of the suspected air bubble area within a single set of suspected air bubbles and the reasonableness of the trend of the air bubbles.

[0128] As an example, this step may include the following steps: First step, determine any suspected air bubble area as the marked suspected area, and perform a linear fitting on the center points of all suspected air bubble areas in the set of suspected air bubble areas corresponding to the above-mentioned marked suspected area to obtain the target line.

[0129] Second step, draw a perpendicular line to the above-mentioned target line through the center point of the above-mentioned marked suspected area, and denote it as the reference perpendicular line.

[0130] Third step, using the above-mentioned reference perpendicular line as the dividing line, divide the set of suspected air bubble areas corresponding to the above-mentioned marked suspected area into two subsets.

[0131] Among them, one subset may include: all the suspected air bubble areas in the set of suspected air bubble areas except the marked suspected area on one side of the reference perpendicular line; the other subset may include: all the suspected air bubble areas in the set of suspected air bubble areas except the marked suspected area on the other side of the reference perpendicular line.

[0132] Fourth step, determine the average value of the areas of all suspected air bubble areas in each subset as the area representative factor corresponding to each subset.

[0133] Fifth step, determine the subset with the larger corresponding area representative factor among the two subsets as the set far from the fibers, and determine the subset with the smaller corresponding area representative factor among the two subsets as the set close to the fibers.

[0134] Sixth step, determine the proportion of the suspected air bubble areas in the above-mentioned set far from the fibers whose corresponding areas are greater than the area of the above-mentioned marked suspected area as the first target proportion corresponding to the above-mentioned marked suspected area.

[0135] For example, the formula for determining the first target proportion corresponding to the marked suspected area can be: ; where is the first target proportion corresponding to the marked suspected area. is the number of suspected bubble regions far from the fiber assembly, where the corresponding area is larger than the area of the marked suspected region. is the total number of all suspected bubble regions far from the fiber assembly.

[0136] It should be noted that when is larger, it often indicates that is closer to , it often indicates that the area of the suspected bubble regions far from the fiber assembly is larger than the marked suspected region, and it often indicates that the distribution of the suspected bubble regions far from the fiber assembly is more likely to conform to the bubble size change distribution.

[0137] Step 7: Determine the second target proportion corresponding to the above-mentioned marked suspected region as the proportion of the suspected bubble regions near the fiber assembly where the corresponding area is smaller than the area of the above-mentioned marked suspected region.

[0138] For example, the formula for determining the second target proportion corresponding to the marked suspected region can be: ; where is the second target proportion corresponding to the marked suspected region. is the number of suspected bubble regions near the fiber assembly where the corresponding area is smaller than the area of the marked suspected region. is the total number of all suspected bubble regions near the fiber assembly.

[0139] It should be noted that when is larger, it often indicates that is closer to , it often indicates that the area of the suspected bubble regions near the fiber assembly is smaller than the marked suspected region, and it often indicates that the distribution of the suspected bubble regions near the fiber assembly is more likely to conform to the bubble size change distribution.

[0140] Step 8: Determine the marked distance corresponding to the above-mentioned marked suspected region as the distance from the center point of the above-mentioned marked suspected region to the above-mentioned target line.

[0141] Step 9: Determine the interference compliance corresponding to the above-mentioned marked suspected region according to the morphological deviation degree, first target proportion, second target proportion, and marked distance corresponding to the above-mentioned marked suspected region.

[0142] Among them, both the morphological deviation degree and the marked distance can have a positive correlation with the interference compliance. Both the first target proportion and the second target proportion can have a negative correlation with the interference compliance.

[0143] For example, the formula for determining the interference compliance corresponding to the marked suspected region can be: ; where F is the interference compliance corresponding to the marked suspected region. is a normalization function. D is the marker distance corresponding to the marker suspected region. H is the degree of morphological deviation corresponding to the marker suspected region. It is the proportion of the first target corresponding to the marked suspected area. It is the proportion of the second target corresponding to the marked suspected area. It is a pre-set factor greater than 0, mainly used to prevent the denominator from being 0, which can be 0.001.

[0144] It should be noted that when H is larger, it often indicates that the suspected marked area is more likely not to be a real bubble area, and it often indicates that the suspected marked area is more likely to be a resin gap or resin accumulation interference area similar to a bubble. When the value is larger, it often indicates that the distribution of suspected bubbles in the area far away from the fiber collection is more likely to conform to the distribution of bubble size changes. The larger the value, the more likely the distribution of the suspected bubble area near the fiber assembly is to conform to the distribution of bubble size changes. When D is smaller, the more likely the marked suspected area is to conform to the linear distribution of bubble changes. Therefore, the larger the value of F, the more likely the marked suspected area is not a real bubble area, and the more likely the marked suspected area is to be a resin gap or resin accumulation interference area similar to a bubble.

[0145] The tenth step is to determine whether the suspected bubble area is a real bubble area based on the interference conformity.

[0146] For example, if the interference conformity corresponding to the suspected bubble region is less than or equal to the preset interference threshold, the suspected bubble region can be determined to be a real bubble region. The preset interference threshold can be the maximum interference conformity allowed when the suspected bubble region is considered to have bubbles, which can be 0.6.

[0147] refer to Figure 2 Based on the same inventive concept as the above method embodiment, the present invention provides a bubble detection system for carbon fiber prepreg production, the system comprising a processor and a memory, the processor being used to process instructions stored in the memory to implement a bubble detection method for carbon fiber prepreg production, which may specifically include: The image and unit acquisition module 201 is used to acquire the target surface image corresponding to the carbon fiber prepreg to be inspected, generate a basically defect-free unit based on the target surface image, and shoot the basically defect-free unit in different postures to obtain a state possible unit; A region division module 202 is used to divide the target surface image into suspected bubble regions and non-bubble regions according to the similarity between the preset window corresponding to the pixel point in the target surface image and the possible units in different states; The region clustering module 203 is configured to cluster the suspected bubble regions based on the non-bubble regions to obtain target clusters; The morphological deviation degree determination module 204 is configured to determine the morphological deviation degree corresponding to each suspected bubble region based on the morphological difference between each suspected bubble region and other suspected bubble regions in its target cluster; The region screening module 205 is configured to screen out the set of suspected bubble regions corresponding to each suspected bubble region from the target cluster based on the area difference between each suspected bubble region and other suspected bubble regions in its target cluster; The analysis, processing and judgment module 206 is configured to perform bubble trend analysis and processing on each suspected bubble region based on the morphological deviation degree and the set of suspected bubble regions corresponding to each suspected bubble region, obtain the interference compliance corresponding to each suspected bubble region, and judge whether the suspected bubble region is a real bubble region based on the interference compliance.

[0148] Figure 3 It is a schematic structural diagram of a computer device provided by an embodiment of the present invention. Exemplarily, as Figure 3 shown, the computer device 300 includes: a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program 303, the computer device can execute any one of the foregoing bubble detection methods for carbon fiber prepreg production.

[0149] Based on the same inventive concept as the above method embodiment, the present invention provides a server, including a memory and a processor. The memory is used to store executable program codes, and the processor is used to call and run the executable program codes from the memory, so that the device executes any one of the foregoing bubble detection methods for carbon fiber prepreg production.

[0150] Based on the same inventive concept as the above method embodiment, the present invention provides a computer program product, which includes: computer program codes. When the computer program codes run on a computer, the computer is enabled to execute any one of the foregoing bubble detection methods for carbon fiber prepreg production.

[0151] Based on the same inventive concept as the above method embodiment, the present invention provides a computer-readable storage medium, which stores computer program codes. When the computer program codes run on a computer, the computer is enabled to execute any one of the foregoing bubble detection methods for carbon fiber prepreg production.

[0152] In summary, when performing bubble detection, the present invention comprehensively considers multiple law features related to bubbles, such as the degree of morphological deviation and the degree of interference compliance, etc., thereby relatively objectively realizing bubble detection and improving the accuracy of bubble detection.

[0153] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the various embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A bubble detection method for carbon fiber prepreg production, characterized in that: The following steps are involved: Obtain a target surface image corresponding to the carbon fiber prepreg to be inspected, generate a basic defect-free unit based on the target surface image, and shoot the basic defect-free unit in different postures to obtain a possible state unit; According to the similarity between the preset window corresponding to the pixel point in the target surface image and the possible units in different states, the suspected bubble area and the non-bubble area are divided from the target surface image; Based on the non-bubble area, the suspected bubble area is clustered to obtain the target cluster; Based on the morphological differences between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs, the degree of morphological deviation corresponding to each suspected bubble region is determined; Based on the area difference between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs, a set of suspected bubble regions corresponding to each suspected bubble region is screened out from the target cluster; Based on the degree of morphological deviation corresponding to each suspected bubble area and the set of suspected bubble areas, a bubble trend analysis is performed on each suspected bubble area to obtain the interference conformity corresponding to each suspected bubble area, and based on the interference conformity, it is determined whether the suspected bubble area is a real bubble area.

2. The bubble detection method for carbon fiber prepreg production according to claim 1, characterized in that: The method of generating a substantially defect-free unit based on the target surface image comprises: Based on the target surface image, separating the warp yarn and the weft yarn by a direction selective filter, and extracting the intersection area in the target surface image as a basic unit; After extracting the basic units, the standard defect-free template is generated by merging them through geometric rules; Based on a standard defect-free template, a cross region in the standard defect-free template is extracted as a basic defect-free unit through a direction selective filter.

3. The bubble detection method for carbon fiber prepreg production according to claim 1, characterized in that: The method of photographing the basically defect-free unit in different postures to obtain the possible state units includes: By adjusting the Euler angle, the different postures of the basically defect-free unit can be photographed, and the possible state units corresponding to different Euler angles can be obtained.

4. The bubble detection method for carbon fiber prepreg production according to claim 3 is characterized in that: The method of dividing the target surface image into suspected bubble areas and non-bubble areas according to the similarity between the preset window corresponding to the pixel point in the target surface image and the possible units in different states includes: Determine the NCC coefficient between the preset window corresponding to each pixel point in the target surface image and each state possible unit as the target similarity between each pixel point and each state possible unit; Filter out the state possible unit with the greatest target similarity with the pixel from all the state possible units as the target matching unit corresponding to the pixel; The same continuous pixel points as the corresponding target matching units in the target surface image constitute a matching analysis area; According to the difference between each matching analysis area and its adjacent matching analysis area, the abnormal distortion index corresponding to each matching analysis area is determined; If the abnormal distortion index corresponding to the matching analysis area is greater than the preset abnormal distortion threshold, the matching analysis area is determined as a suspected bubble area; If the abnormal distortion index corresponding to the matching analysis area is less than or equal to the preset abnormal distortion threshold, the matching analysis area is determined as a non-bubble area.

5. A bubble detection method for carbon fiber prepreg production according to claim 4, characterized in that: Determining the abnormal distortion index corresponding to each matching analysis region according to the difference between each matching analysis region and its adjacent matching analysis region includes: Determine the Euler angle corresponding to the target matching unit corresponding to the pixel point in each matching analysis area as the target Euler angle corresponding to each matching analysis area; Determine any matching analysis region as a marked region, and determine each matching analysis region adjacent to the marked region as a reference region; Determine the difference between the target Euler angle corresponding to the marked area and the target Euler angle corresponding to each reference area as the target difference between the marked area and each reference area; Determining an angle difference index corresponding to the marked area according to a target difference between the marked area and all reference areas, wherein the target difference is positively correlated with the angle difference index; According to the target similarity between all the pixels in the marked area and the corresponding target matching units, and the angle difference index corresponding to the marked area, the abnormal distortion index corresponding to the marked area is determined, wherein the target similarity is negatively correlated with the abnormal distortion index, and the angle difference index is positively correlated with the abnormal distortion index.

6. The bubble detection method for carbon fiber prepreg production according to claim 3 is characterized in that: The method clusters the suspected bubble areas based on the non-bubble areas to obtain target clusters, including: Each edge pixel point of each non-bubble area is determined as a marked pixel point, and a target matching unit corresponding to the marked pixel point is used as a matching template to perform template matching to obtain a target matching area corresponding to the marked pixel point; Update the target matching units corresponding to all pixels in the target matching area corresponding to the marked pixel to the target matching unit corresponding to the marked pixel; Determine an area formed by target matching areas corresponding to all marked pixels as a reference area, and determine each pixel in the reference area as a reference point; Determine the region consisting of all pixels except the reference point in each suspected bubble region as a candidate region; The reference point and the pixel points in the non-bubble area are recorded as temporary pixel points, and the average of the Euler angles corresponding to the latest target matching units corresponding to all temporary pixel points in the preset window corresponding to each edge pixel point of each candidate area is determined as the Euler representative angle corresponding to each edge pixel point of each candidate area; Obtain the state possible unit corresponding to the Euler representative angle corresponding to each edge pixel point of each candidate region as the target possible unit corresponding to each edge pixel point of each candidate region; The target possible unit corresponding to the edge pixel point of the candidate area is used as the matching template for template matching to obtain the target matching area corresponding to the edge pixel point of the candidate area; Update the target matching units corresponding to all pixels in the target matching area corresponding to the edge pixel of the candidate area to the target possible unit corresponding to the edge pixel; Determine an area formed by target matching areas corresponding to all edge pixels of all candidate areas as a temporary area, and determine each pixel point in the temporary area as a temporary point; The candidate region is updated to a region consisting of all pixels except the temporary point, and the candidate region updating step is repeated until the final updated candidate region is empty, thereby updating the target matching unit corresponding to each pixel in each suspected bubble region; Clustering the updated target matching units corresponding to all pixels in all suspected bubble areas, and determining each obtained cluster as a reference cluster; The number of target matching units belonging to each reference cluster in the updated target matching units corresponding to all pixels in each suspected bubble region is determined as the overall belonging number between each suspected bubble region and each reference cluster; Select the reference cluster with the largest number of overall affiliations with the suspected bubble area from all reference clusters as the best reference cluster corresponding to the suspected bubble area; The suspected bubble areas with the same corresponding best reference cluster are divided into the same target cluster.

7. The bubble detection method for carbon fiber prepreg production according to claim 1, characterized in that: The step of determining the degree of morphological deviation corresponding to each suspected bubble region based on the morphological difference between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs includes: The distance between every two edge pixels of each suspected bubble region is determined as a reference distance, and a reference distance set corresponding to each suspected bubble region is obtained; The maximum value in the reference distance set corresponding to each suspected bubble area is determined as the target long side representative distance corresponding to each suspected bubble area; The line segment connecting two edge pixels corresponding to the largest reference distance in the reference distance set corresponding to each suspected bubble region is determined as the target line segment corresponding to each suspected bubble region; The angle corresponding to the target line segment corresponding to each suspected bubble area is determined as the target representative angle corresponding to each suspected bubble area; Through any point in the suspected bubble region, a perpendicular line is drawn to the target line segment corresponding to the suspected bubble region, recorded as the target perpendicular line corresponding to the suspected bubble region, and the target perpendicular line is slid on the suspected bubble region, and the line segment obtained by the intersection of the target perpendicular line and the boundary of the suspected bubble region during the sliding process is recorded as a candidate line segment, thereby obtaining a set of candidate line segments corresponding to the suspected bubble region; The maximum value in the set of candidate line segments corresponding to each suspected bubble region is determined as the representative distance of the target short side corresponding to each suspected bubble region; The ratio of the target long side representative distance corresponding to each suspected bubble area to the target short side representative distance is determined as the shape characteristic factor corresponding to each suspected bubble area; The degree of morphological deviation corresponding to each suspected bubble region is determined based on the difference between the shape characteristic factors corresponding to each suspected bubble region and the shape characteristic factors corresponding to other suspected bubble regions in the target cluster to which it belongs, and the difference between the target representative angle corresponding to each suspected bubble region and the target representative angle corresponding to other suspected bubble regions in the target cluster to which it belongs.

8. The bubble detection method for carbon fiber prepreg production according to claim 1, characterized in that: The method of screening out a set of suspected bubble regions corresponding to each suspected bubble region from the target cluster based on the area difference between each suspected bubble region and other suspected bubble regions in the target cluster to which it belongs includes: Determine any suspected bubble region as a marked suspected region, and determine each suspected bubble region in the target cluster to which the marked suspected region belongs, except the marked suspected region, as a reference suspected region; Determining a generation consistency factor between the marked suspected region and each reference suspected region according to the distance and area difference between the marked suspected region and each reference suspected region; If the generation consistency factor between the marked suspected area and the reference suspected area is greater than a preset consistency threshold, the reference suspected area is determined as a generation consistent area; The suspected mark area and all the generated consistent areas are combined to form a set of suspected bubble areas corresponding to the suspected mark area.

9. The bubble detection method for carbon fiber prepreg production according to claim 1, characterized in that: Based on the morphological deviation degree corresponding to each suspected bubble region and the set of suspected bubble regions, each suspected bubble region is subjected to bubble trend analysis processing to obtain the interference compliance corresponding to each suspected bubble region, including: Determine any suspected bubble region as a marked suspected region, and perform straight line fitting on the center points of all suspected bubble regions in the set of suspected bubble regions corresponding to the marked suspected region to obtain a target straight line; Draw a perpendicular line to the target straight line through the center point of the suspected marked area, and record it as a reference perpendicular line; Taking the reference vertical line as a dividing line, dividing the set of suspected bubble regions corresponding to the suspected marked regions into two subsets; The average value of the area of ​​all suspected bubble regions in each subset is determined as the area representative factor corresponding to each subset; The subset with the larger area representative factor among the two subsets is determined as the set far from the fiber, and the subset with the smaller area representative factor among the two subsets is determined as the set close to the fiber; Determine the proportion of the suspected bubble regions in the remote fiber set whose corresponding area is larger than the area of ​​the suspected marked region as the first target proportion corresponding to the suspected marked region; Determine the proportion of the suspected bubble area whose corresponding area in the near-fiber set is smaller than the area of ​​the suspected marked area as the second target proportion corresponding to the suspected marked area; Determine the distance from the center point of the suspected mark area to the target straight line as the mark distance corresponding to the suspected mark area; The interference compliance corresponding to the marked suspected area is determined according to the degree of morphological deviation, the first target proportion, the second target proportion and the marking distance corresponding to the marked suspected area, wherein the degree of morphological deviation and the marking distance are positively correlated with the interference compliance, and the first target proportion and the second target proportion are negatively correlated with the interference compliance.

10. A bubble detection system for carbon fiber prepreg production, characterized in that: The invention comprises a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement a bubble detection method for producing carbon fiber prepreg cloth according to any one of claims 1 to 9.

Images