Wood board edge sealing tape glue failure measuring method and system and storage medium

Abstract

The invention discloses a wood board edge band glue failure measuring method and system and a storage medium, and the method comprises the steps: obtaining a wood board image, and carrying out the preprocessing based on the wood board image, and obtaining a processed image; equally-spaced segmentation is carried out on the glue failure area based on the processed image to obtain an interested area, the height of the interested area is ten pixel points, and the width of the interested area is the width value of the processed image; obtaining a glue failure connected domain range, wherein the glue failure connected domain range comprises local self-adaptive threshold segmentation and vertical direction projection to obtain the glue failure connected domain range; and after a glue failure connected domain range is obtained, carrying out glue failure measurement calculation by adopting a local extreme point and an edge point pair. According to the method, the influence of various interference problems in the industrial field on the glue failure measurement can be effectively solved, the universality of an image algorithm is improved by adopting the thought of equally dividing the glue failure area, and when the edge extreme points of the wood board and the edge band are calculated, the robustness of the wood board and the edge band can be improved by adopting the multiple filtering thought for screening.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and more specifically to a method, system and storage medium for measuring debonding of edge banding of a wooden board. Background Art

[0002] In the industrial field, debonding of wood board edge banding is a frequent problem, but measuring the maximum distance of debonding faces many challenges. Although visual algorithm technology is changing with each passing day, the current technical solutions still have certain limitations and shortcomings. Specifically, AI segmentation and labeling tasks are arduous and accuracy is difficult to guarantee, while traditional algorithms have poor robustness when dealing with complex and changing environments.

[0003] Especially in industrial manufacturing environments, due to the different processes of wood board edge banding machines, various situations may occur when the edge banding debonds, such as arched debonding between the wood board and the edge banding, local one-sided debonding, and different imaging positions of the edge banding, which will lead to separated debonding, resulting in poor versatility in visual measurement of debonding. In addition, the diversity and complexity of wood board patterns easily confuse the edges of the wood board and the edge banding, making conventional wood board edge banding measurement methods lack sufficient robustness in practical applications. In addition, due to material problems, some wood boards are prone to chipping at the edge banding, which seriously interferes with the measurement of debonding. Therefore, there is currently no public and highly robust visual measurement algorithm for wood board edge banding in the industrial field. Summary of the invention

[0004] The purpose of the present invention is to provide a method, system and storage medium for measuring debonding of edge bands of wooden boards, which can effectively solve the influence of various interference problems existing in the industrial field on debonding measurement. Among them, the idea of ​​equidistantly dividing the debonding area is adopted to improve the versatility of the image algorithm. When calculating the extreme points of the edges of wooden boards and edge bands, the idea of ​​multiple filtering is adopted for screening to improve the robustness of the edges of wooden boards and edge bands.

[0005] A first aspect of the present invention provides a method for measuring debonding of a wood board edge band, comprising the following steps: Acquire a wooden board image, and perform preprocessing based on the wooden board image to obtain a processed image; Based on the processed image, the debonding area is divided at equal intervals to obtain an area of ​​interest, wherein the area of ​​interest is ten pixels high and the width is the width of the processed image; Acquiring the scope of the debonding connected domain, which includes local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain; After obtaining the range of the debonding connected domain, local extreme points and edge point pairs are used to perform debonding measurement calculations.

[0006] In this solution, the step of obtaining the wooden board image and performing preprocessing based on the wooden board image to obtain a processed image specifically includes: The wooden board image is normalized, grayscale scaled, and contrast adjusted to obtain an adjusted image; Gaussian filtering and noise removal are performed on the adjusted image to obtain the processed image.

[0007] In this solution, the scope of the debonding connected domain is obtained, which includes local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain, specifically including: Performing a segmentation operation on the grayscale image of the region of interest, using a local mean to perform local adaptive segmentation and performing a closing operation to connect each edge region; Performing vertical projection on the grayscale image of the region of interest and calculating the mean value to simplify it into a one-dimensional vector; A connected domain analysis is performed on the one-dimensional vector to obtain a range of each connected domain, wherein the connected domain range corresponds to a range of edge extreme value points.

[0008] In this solution, calculating the extreme points includes: Taking a first-order derivative of the one-dimensional vector to obtain positive and negative extreme point vectors; A minimum threshold of extreme points is set to filter the positive and negative extreme point vectors, and target extreme points are screened in combination with the connected domain range, wherein the filtering also includes filtering the wood board grain interference and non-background areas to obtain a list of extreme points on the edge of the wood board.

[0009] In this solution, the method further includes filtering the board edge abnormal points from the board edge extreme value point list, specifically including: A straight line fitting algorithm is used for the list of extreme point values ​​on the edge of the plank using a random sampling consistency algorithm, wherein when the number of fitted straight line point sets exceeds a preset proportion of the total number of point sets, the straight line is considered valid; The fixed straight line angle range is 60-120 degrees in the vertical direction; The point whose distance from the fixed extreme edge point to the fitting straight line is within the preset pixel is regarded as a valid point.

[0010] In this scheme, each valid pair of edge points in the list of extreme value points on the edge of the wooden board after filtering outliers on the edge of the wooden board is traversed to find the maximum distance as the maximum distance for debonding.

[0011] A second aspect of the present invention further provides a system for measuring the debonding of a wood board edge band, comprising a memory and a processor, wherein the memory comprises a program for measuring the debonding of a wood board edge band, and when the program for measuring the debonding of a wood board edge band is executed by the processor, the following steps are implemented: Acquire a wooden board image, and perform preprocessing based on the wooden board image to obtain a processed image; Based on the processed image, the debonding area is divided at equal intervals to obtain an area of ​​interest, wherein the area of ​​interest is ten pixels high and the width is the width of the processed image; Acquiring the scope of the debonding connected domain, which includes local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain; After obtaining the range of the debonding connected domain, local extreme points and edge point pairs are used to perform debonding measurement calculations.

[0012] In this solution, the step of obtaining the wooden board image and performing preprocessing based on the wooden board image to obtain a processed image specifically includes: The wooden board image is normalized, grayscale scaled, and contrast adjusted to obtain an adjusted image; Gaussian filtering and noise removal are performed on the adjusted image to obtain the processed image.

[0013] In this solution, the scope of the debonding connected domain is obtained, which includes local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain, specifically including: Performing a segmentation operation on the grayscale image of the region of interest, using a local mean to perform local adaptive segmentation and performing a closing operation to connect each edge region; Performing vertical projection on the grayscale image of the region of interest and calculating the mean value to simplify it into a one-dimensional vector; A connected domain analysis is performed on the one-dimensional vector to obtain a range of each connected domain, wherein the connected domain range corresponds to a range of edge extreme value points.

[0014] In this solution, calculating the extreme points includes: Taking a first-order derivative of the one-dimensional vector to obtain positive and negative extreme point vectors; A minimum threshold of extreme points is set to filter the positive and negative extreme point vectors, and target extreme points are screened in combination with the connected domain range, wherein the filtering also includes filtering the wood board grain interference and non-background areas to obtain a list of extreme points on the edge of the wood board.

[0015] In this solution, the method further includes filtering the board edge abnormal points from the board edge extreme value point list, specifically including: A straight line fitting algorithm is used for the list of extreme point values ​​on the edge of the plank using a random sampling consistency algorithm, wherein when the number of fitted straight line point sets exceeds a preset proportion of the total number of point sets, the straight line is considered valid; The fixed straight line angle range is 60-120 degrees in the vertical direction; The point whose distance from the fixed extreme edge point to the fitting straight line is within the preset pixel is regarded as a valid point.

[0016] In this scheme, each valid pair of edge points in the list of extreme value points on the edge of the wooden board after filtering outliers on the edge of the wooden board is traversed to find the maximum distance as the maximum distance for debonding.

[0017] A third aspect of the present invention provides a computer-readable storage medium, which includes a program for a method for measuring debonding of a wooden board edge banding of a machine. When the program for measuring debonding of a wooden board edge banding is executed by a processor, the steps of a method for measuring debonding of a wooden board edge banding as described in any one of the above items are implemented.

[0018] The present invention discloses a method, system and storage medium for measuring debonding of edge bands of wooden boards, which can effectively solve the influence of various interference problems existing in the industrial field on debonding measurement. The idea of ​​equally dividing the debonding area is adopted to improve the versatility of the image algorithm. When calculating the extreme points of the edges of wooden boards and edge bands, the idea of ​​multiple filtering is adopted for screening to improve the robustness of the edges of wooden boards and edge bands. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram showing the steps of a method for measuring debonding of a wood board edge band of the present invention is shown; Figure 2 A schematic diagram of a wooden board image showing a method for measuring debonding of a wooden board edge band of the present invention is shown; Figure 3 A schematic diagram of a processed image of a method for measuring debonding of a wood board edge banding according to the present invention is shown; Figure 4A A local adaptive threshold segmentation result diagram of a method for measuring debonding of a wood board edge band of the present invention is shown; Figure 4B A schematic diagram showing the display result after the closed operation of a method for measuring the debonding of a wood board edge band of the present invention is shown; Figure 4C A schematic diagram of ROI binarization after equidistant segmentation of a method for measuring debonding of a wood board edge band of the present invention is shown; Figure 5 A one-dimensional vector projection schematic diagram of a method for measuring debonding of a wood board edge band of the present invention is shown; Figure 6 A schematic diagram of the distribution of derivative extreme value points of a method for measuring debonding of a wood board edge band of the present invention is shown; Figure 7 A schematic diagram showing the maximum distribution of the absolute values ​​of the extreme points of a method for measuring the debonding of a wood board edge band according to the present invention is shown; Figure 8 A schematic diagram of the distribution of minimum value vectors of a method for measuring debonding of a wood board edge band of the present invention is shown; Fig. 9 A schematic diagram of the maximum value vector distribution of a method for measuring the debonding of a wood board edge band according to the present invention is shown; Fig.10 A schematic diagram of chipping interference of a method for measuring debonding of a wood board edge band of the present invention is shown; Fig.11 A flow chart of a method for measuring debonding of a wood board edge band of the present invention is shown; Fig. 12A A schematic diagram showing the display of arched debonding in a method for measuring debonding of a wood board edge band of the present invention is shown; Fig. 12B A schematic diagram showing one-side warping and debonding of a method for measuring debonding of a wood board edge band of the present invention is shown; Fig. 12C A schematic diagram showing the display of a separation type debonding of a method for measuring debonding of a wood board edge band of the present invention is shown; Fig.12D A schematic diagram showing the display of wood board stripe interference in a method for measuring debonding of wood board edge banding according to the present invention is shown; Fig.12E A schematic diagram showing the display of chipping interference in a method for measuring debonding of a wood board edge band of the present invention is shown; Fig.13 A block diagram of a system for measuring debonding of a wood board edge band of the present invention is shown. DETAILED DESCRIPTION

[0020] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0022] Figure 1 A flow chart of a method for measuring debonding of a wood board edge band of the present application is shown.

[0023] like Figure 1 As shown, the present application discloses a method for measuring the debonding of a wood board edge band, comprising the following steps: S102, acquiring a wooden board image, and performing preprocessing based on the wooden board image to obtain a processed image; S104, dividing the debonding area at equal intervals based on the processed image to obtain an area of ​​interest, wherein the area of ​​interest is ten pixels high and the width is the width of the processed image; S106, obtaining the range of the debonding connected domain, which includes local adaptive threshold segmentation and vertical projection to obtain the range of the debonding connected domain; S108, after obtaining the range of the debonding connected domain, using local extreme points and edge point pairs to perform debonding measurement calculation.

[0024] It should be noted that, in this embodiment, AI detection is used for initial positioning, and the network model adopts the open source YOLO8 model for detection. Only a certain amount of annotation is used to stably obtain the debonding area in the wooden board, and it is assumed that the size of the debonding area is h*w (height*width, and the size of each debonding area is different). The subsequent measurement part is difficult to annotate with the AI ​​segmentation model, so the traditional method is used for measurement and acquisition. The obtained wooden board image is as follows: Figure 2 shown.

[0025] According to an embodiment of the present invention, the acquiring of the wooden board image and the preprocessing of the wooden board image to obtain the processed image specifically include: The wooden board image is normalized, grayscale scaled, and contrast adjusted to obtain an adjusted image; Gaussian filtering and noise removal are performed on the adjusted image to obtain the processed image.

[0026] It should be noted that, in this embodiment, the wooden board image is normalized, the grayscale value range is scaled to the "0-255" area, and the contrast is enhanced to prevent the wooden board area from being dark due to poor lighting, and the image is Gaussian filtered. After normalization, the noise will also be amplified, so it is necessary to eliminate the influence of small Gaussian noise. Figure 3 As shown, a schematic diagram of the processed image is shown.

[0027] According to an embodiment of the present invention, obtaining the range of the debonding connected domain includes local adaptive threshold segmentation and vertical projection to obtain the range of the debonding connected domain, specifically including: Performing a segmentation operation on the grayscale image of the region of interest, using a local mean to perform local adaptive segmentation and performing a closing operation to connect each edge region; Performing vertical projection on the grayscale image of the region of interest and calculating the mean value to simplify it into a one-dimensional vector; A connected domain analysis is performed on the one-dimensional vector to obtain a range of each connected domain, wherein the connected domain range corresponds to a range of edge extreme value points.

[0028] It should be noted that, in this embodiment, because the debonding area of ​​the edge band may appear in various forms in the camera field of view, including "arched debonding", "local unilateral debonding" and "separated debonding", the global image segmentation method is often unable to accurately locate the debonding area correctly. Therefore, in this embodiment, the image is divided into debonding areas at equal intervals, wherein each small ROI (region of interest) after segmentation is 10 pixels high and as wide as the width w of the preprocessed image (the last row of ROIs is not calculated if the height is less than 10 pixels).

[0029] Furthermore, the grayscale image of each small ROI is segmented, and the local mean method is used for local adaptive segmentation, which can segment the local edge information, filter most of the non-edge parts, and perform a closed operation to connect each edge area, where: Figure 4A is the local adaptive threshold segmentation result, Figure 4B To display the result after the closing operation, Figure 4C It is a binary image of a small ROI in a certain layer after equidistant segmentation.

[0030] Furthermore, the grayscale image of the region of interest is projected vertically to calculate the mean value to simplify it into a one-dimensional vector, wherein each small ROI binary image is projected vertically to calculate the mean value to simplify it into a one-dimensional vector of size 1*w, as shown in the following example: Figure 5 shown.

[0031] Furthermore, a connected domain analysis is performed on the one-dimensional vector to obtain the range of each connected domain, wherein the connected domain range corresponds to the edge extreme value point range, so as to filter out some interference extreme value points caused by patterns or backgrounds.

[0032] According to an embodiment of the present invention, calculating the extreme point includes: Taking a first-order derivative of the one-dimensional vector to obtain positive and negative extreme point vectors; A minimum threshold of extreme points is set to filter the positive and negative extreme point vectors, and target extreme points are screened in combination with the connected domain range, wherein the filtering also includes filtering the wood board grain interference and non-background areas to obtain a list of extreme points on the edge of the wood board.

[0033] It should be noted that, in this embodiment, the first-order derivative of the one-dimensional vector is performed to obtain positive and negative extreme point vectors, which can be divided into minimum value vectors, maximum value vectors and absolute value maximum value vectors, such as Figure 6 As shown in the figure, the distribution of derivative extreme points (the extreme point values ​​in the white area are greater than 0, and the extreme point values ​​in the black area are less than 0). It can be seen that the locations where the extreme points appear are concentrated in the wooden board and edge banding areas.

[0034] Furthermore, the minimum threshold of extreme points is set to filter the positive and negative extreme point vectors. By setting the minimum threshold of extreme points, relatively small extreme points are initially filtered to roughly confirm the edge positions of the wooden board and the edge band. Figure 7 As shown, the first bright spot is the edge of the wooden board, and the second and third bright spots are the edges on both sides of the edge banding.

[0035] It should be noted that for some wooden board areas, the extreme point vector still contains multiple extreme points where the gradient changes too quickly, which require further filtering. Therefore, the filtering also includes filtering the wooden board grain interference and non-background areas to obtain a list of extreme points on the edge of the wooden board.

[0036] Specifically, when striped patterns appear, the judgment of extreme points is prone to a lot of interference. Taking the wooden board area facing left, the edge band area facing right, and the wooden board pattern similar to vertical stripes as an example, when obtaining the extreme points at the edge of the wooden board area, due to the interference of the wooden board pattern, the extreme points may fall at various positions in the wooden board area, and the best edge extreme points cannot be located.

[0037] Therefore, from left to right as the positive direction, collect the extreme point vectors of the edge change from bright to dark, that is, the minimum value vector. This vector is the "extreme point of the edge of the board to be selected". Due to the interference of the pattern, multiple extreme points will be generated, such as Figure 8 The second bright spot in the area is the actual extreme point of the edge of the board; from right to left as the positive direction, collect the extreme point vectors of the edge change from bright to dark, which is the "extreme point of the edge band to be selected". Due to the interference of the pattern in the board area, interference extreme points will also be generated. Fig. 9 The first bright spot in the area viewed from right to left is the actual edge extreme point of the edge banding. Then, by setting the distance threshold of adjacent extreme points to dis, if the distance between multiple extreme points in the "extreme points of the edge of the board to be selected" is within dis, the rightmost extreme point is retained and the other extreme points are cleared; similarly, if the distance between the extreme points in the "extreme points of the edge of the edge band to be selected" is within dis, the leftmost extreme point is retained and the other extreme points are cleared. This step can effectively remove the extreme point interference caused by most adjacent textures, but there may still be some extreme points of irregular textures, and some textures cannot be removed. Therefore, the background area pixel mean is introduced as a reference.

[0038] Specifically, among the extreme value points on the edge of the board to be selected, the right side of the extreme value point on the edge of the board that is actually debonding should be the background area, where the pixel mean is low. Assuming that the pixel mean is lower than a certain threshold in a certain area to the right of an extreme value point encountered for the first time, it is considered to be a valid extreme value point on the edge, and the valid extreme value points of each small ROI in the debonding area are collected to form a "list of extreme value points on the edge of the board". The extreme value points encountered for the first time are used mainly to prevent recognition on the edge banding; and the edge banding does not have various patterns, and the basic color remains consistent. Therefore, among the extreme value points on the edge of the edge banding to be selected, the left side of the edge banding should be the background area, where the pixel mean is low. Assuming that the pixel mean is lower than a certain value in a certain area to the left of an extreme value point encountered for the first time, the extreme value point is considered to be a valid extreme value point on the edge of the edge banding, and the positions of the valid extreme value points of each small ROI in the debonding area are collected to form a "list of extreme value points on the edge of the edge banding".

[0039] According to an embodiment of the present invention, the method further includes filtering the board edge abnormal points from the board edge extreme value point list, specifically including: A straight line fitting algorithm is used for the list of extreme point values ​​on the edge of the plank using a random sampling consistency algorithm, wherein when the number of fitted straight line point sets exceeds a preset proportion of the total number of point sets, the straight line is considered valid; The fixed straight line angle range is 60-120 degrees in the vertical direction; The point whose distance from the fixed extreme edge point to the fitting straight line is within the preset pixel is regarded as a valid point.

[0040] It should be noted that in this embodiment, the "board edge extreme point list" is the first extreme point encountered in each small ROI. However, when there is a "chip" or damage in the edge area of ​​the board, the first extreme point encountered at this time may be the chip area inside the board, such as Fig.10 As shown, in order to prevent the interference of "collapse", the idea of ​​filtering the extreme points of the edge of the wooden board by straight line is proposed, that is, the edge points at the edge of the wooden board can be fitted into a straight line. Specifically, the "extreme point list of the edge of the wooden board" is fitted into a straight line algorithm using a random sampling consistency algorithm. When the number of fitted straight line point sets exceeds a preset proportion of the total number of point sets, the preset proportion is, for example, "60%". The straight line is considered valid, and the straight line angle range is fixed to 60-120 degrees in the vertical direction, and the point whose distance from the extreme edge point to the fitting straight line is within a preset pixel is a valid point. The preset pixel is, for example, "20 pixels". After the operation in this embodiment, abnormal points on the edge of the wooden board can be filtered out.

[0041] Finally, if Fig.11As shown, after completing the filtering of some textures and non-background areas, determine whether the ROI has reached the end. If so, filter the extreme points of the wooden board in a straight line, and traverse each valid group of edge point pairs in the list of extreme points of the wooden board edge after filtering the abnormal points on the wooden board edge to find the maximum distance as the maximum distance for debonding.

[0042] Further, it should be noted that there are arched debonding, unilateral warping debonding, and separation debonding types. In the processing process, there are also interferences of wood stripes and chipping. After actual measurement by the measurement method of the present application, different test results can be clearly obtained. Among them, Fig. 12A A schematic diagram showing the display of arched debonding is shown. Fig. 12B The schematic diagram of the display of one-side warping and debonding is shown. Fig. 12C A schematic diagram of a separation type debonding is shown. Fig.12D A schematic diagram showing the display of plank stripe interference is shown. Fig.12E A display schematic diagram showing chipping interference.

[0043] Fig.13 A block diagram of a system for measuring debonding of a wood board edge band of the present invention is shown.

[0044] like Fig.13 As shown, the present invention discloses a system for measuring the debonding of a wood board edge band, comprising a memory and a processor, wherein the memory comprises a program for measuring the debonding of a wood board edge band, and when the program for measuring the debonding of a wood board edge band is executed by the processor, the following steps are implemented: Acquire a wooden board image, and perform preprocessing based on the wooden board image to obtain a processed image; Based on the processed image, the debonding area is divided at equal intervals to obtain an area of ​​interest, wherein the area of ​​interest is ten pixels high and the width is the width of the processed image; Acquiring the scope of the debonding connected domain, which includes local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain; After obtaining the range of the debonding connected domain, local extreme points and edge point pairs are used to perform debonding measurement calculations.

[0045] It should be noted that, in this embodiment, AI detection is used for initial positioning, and the network model adopts the open source YOLO8 model for detection. Only a certain amount of annotation is used to stably obtain the debonding area in the wooden board, and it is assumed that the size of the debonding area is h*w (height*width, and the size of each debonding area is different). The subsequent measurement part is difficult to annotate with the AI ​​segmentation model, so the traditional method is used for measurement and acquisition. The obtained wooden board image is as follows: Figure 2 shown.

[0046] According to an embodiment of the present invention, the acquiring of the wooden board image and the preprocessing of the wooden board image to obtain the processed image specifically include: The wooden board image is normalized, grayscale scaled, and contrast adjusted to obtain an adjusted image; Gaussian filtering and noise removal are performed on the adjusted image to obtain the processed image.

[0047] It should be noted that, in this embodiment, the wooden board image is normalized, the grayscale value range is scaled to the "0-255" area, and the contrast is enhanced to prevent the wooden board area from being dark due to poor lighting, and the image is Gaussian filtered. After normalization, the noise will also be amplified, so it is necessary to eliminate the influence of small Gaussian noise. Figure 3 As shown, a schematic diagram of the processed image is shown.

[0048] According to an embodiment of the present invention, obtaining the range of the debonding connected domain includes local adaptive threshold segmentation and vertical projection to obtain the range of the debonding connected domain, specifically including: Performing a segmentation operation on the grayscale image of the region of interest, using a local mean to perform local adaptive segmentation and performing a closing operation to connect each edge region; Performing vertical projection on the grayscale image of the region of interest and calculating the mean value to simplify it into a one-dimensional vector; A connected domain analysis is performed on the one-dimensional vector to obtain a range of each connected domain, wherein the connected domain range corresponds to a range of edge extreme value points.

[0049] It should be noted that, in this embodiment, because the debonding area of ​​the edge band may appear in various forms in the camera field of view, including "arched debonding", "local unilateral debonding" and "separated debonding", the global image segmentation method is often unable to accurately locate the debonding area correctly. Therefore, in this embodiment, the image is divided into debonding areas at equal intervals, wherein each small ROI (region of interest) after segmentation is 10 pixels high and as wide as the width w of the preprocessed image (the last row of ROIs is not calculated if the height is less than 10 pixels).

[0050] Furthermore, the grayscale image of each small ROI is segmented, and the local mean method is used for local adaptive segmentation, which can segment the local edge information, filter most of the non-edge parts, and perform a closed operation to connect each edge area, where: Figure 4A is the local adaptive threshold segmentation result, Figure 4B To display the result after the closing operation, Figure 4C It is a binary image of a small ROI in a certain layer after equidistant segmentation.

[0051] Furthermore, the grayscale image of the region of interest is projected vertically to calculate the mean value to simplify it into a one-dimensional vector, wherein each small ROI binary image is projected vertically to calculate the mean value to simplify it into a one-dimensional vector of size 1*w, as shown in the following example: Figure 5 shown.

[0052] Furthermore, a connected domain analysis is performed on the one-dimensional vector to obtain the range of each connected domain, wherein the connected domain range corresponds to the edge extreme value point range, so as to filter out some interference extreme value points caused by patterns or backgrounds.

[0053] According to an embodiment of the present invention, calculating the extreme point includes: Taking a first-order derivative of the one-dimensional vector to obtain positive and negative extreme point vectors; A minimum threshold of extreme points is set to filter the positive and negative extreme point vectors, and target extreme points are screened in combination with the connected domain range, wherein the filtering also includes filtering the wood board grain interference and non-background areas to obtain a list of extreme points on the edge of the wood board.

[0054] It should be noted that, in this embodiment, the first-order derivative of the one-dimensional vector is performed to obtain positive and negative extreme point vectors, which can be divided into minimum value vectors, maximum value vectors and absolute value maximum value vectors, such as Figure 6 As shown in the figure, the distribution of derivative extreme points (the extreme point values ​​in the white area are greater than 0, and the extreme point values ​​in the black area are less than 0). It can be seen that the locations where the extreme points appear are concentrated in the wooden board and edge banding areas.

[0055] Furthermore, the minimum threshold of extreme points is set to filter the positive and negative extreme point vectors. By setting the minimum threshold of extreme points, relatively small extreme points are initially filtered to roughly confirm the edge positions of the wooden board and the edge band. Figure 7 As shown, the first bright spot is the edge of the wooden board, and the second and third bright spots are the edges on both sides of the edge banding.

[0056] It should be noted that for some wooden board areas, the extreme point vector still contains multiple extreme points where the gradient changes too quickly, which require further filtering. Therefore, the filtering also includes filtering the wooden board grain interference and non-background areas to obtain a list of extreme points on the edge of the wooden board.

[0057] Specifically, when striped patterns appear, the judgment of extreme points is prone to a lot of interference. Taking the wooden board area facing left, the edge band area facing right, and the wooden board pattern similar to vertical stripes as an example, when obtaining the extreme points at the edge of the wooden board area, due to the interference of the wooden board pattern, the extreme points may fall at various positions in the wooden board area, and the best edge extreme points cannot be located.

[0058] Therefore, from left to right as the positive direction, collect the extreme point vectors of the edge change from bright to dark, that is, the minimum value vector. This vector is the "extreme point of the edge of the board to be selected". Due to the interference of the pattern, multiple extreme points will be generated, such as Figure 8 The second bright spot in the area is the actual extreme point of the edge of the board; from right to left as the positive direction, collect the extreme point vectors of the edge change from bright to dark, which is the "extreme point of the edge band to be selected". Due to the interference of the pattern in the board area, interference extreme points will also be generated. Fig. 9 The first bright spot in the area viewed from right to left is the edge extreme point of the edge band we are actually looking for. Then, by (3), the distance threshold of adjacent extreme points is set to dis. If the distance between multiple extreme points in the “extreme points of the edge of the board to be selected” is within dis, the rightmost extreme point is retained and the other extreme points are cleared. Similarly, if the distance between the extreme points in the “extreme points of the edge of the board to be selected” is within dis, the leftmost extreme point is retained and the other extreme points are cleared. This step can effectively remove the extreme point interference caused by most adjacent textures, but there may still be some extreme points of irregular textures, and some textures may not be removed. Therefore, the background area pixel mean is introduced as a reference.

[0059] Specifically, among the extreme value points on the edge of the board to be selected, the right side of the extreme value point on the edge of the board that is actually debonding should be the background area, where the pixel mean is low. Assuming that the pixel mean is lower than a certain threshold in a certain area to the right of an extreme value point encountered for the first time, it is considered to be a valid extreme value point on the edge, and the valid extreme value points of each small ROI in the debonding area are collected to form a "list of extreme value points on the edge of the board". The extreme value points encountered for the first time are used mainly to prevent recognition on the edge banding; and the edge banding does not have various patterns, and the basic color remains consistent. Therefore, among the extreme value points on the edge of the edge banding to be selected, the left side of the edge banding should be the background area, where the pixel mean is low. Assuming that the pixel mean is lower than a certain value in a certain area to the left of an extreme value point encountered for the first time, the extreme value point is considered to be a valid extreme value point on the edge of the edge banding, and the positions of the valid extreme value points of each small ROI in the debonding area are collected to form a "list of extreme value points on the edge of the edge banding".

[0060] According to an embodiment of the present invention, the method further includes filtering the board edge abnormal points from the board edge extreme value point list, specifically including: A straight line fitting algorithm is used for the list of extreme point values ​​on the edge of the plank using a random sampling consistency algorithm, wherein when the number of fitted straight line point sets exceeds a preset proportion of the total number of point sets, the straight line is considered valid; The fixed straight line angle range is 60-120 degrees in the vertical direction; The point whose distance from the fixed extreme edge point to the fitting straight line is within the preset pixel is regarded as a valid point.

[0061] It should be noted that in this embodiment, the "board edge extreme point list" is the first extreme point encountered in each small ROI. However, when there is a "chip" or damage in the edge area of ​​the board, the first extreme point encountered at this time may be the chip area inside the board, such as Fig.10 As shown, in order to prevent the interference of "collapse", the idea of ​​filtering the extreme points of the edge of the wooden board by straight line is proposed, that is, the edge points at the edge of the wooden board can be fitted into a straight line. Specifically, the "extreme point list of the edge of the wooden board" is fitted into a straight line algorithm by a random sampling consistency algorithm. When the number of fitted straight line point sets exceeds the preset proportion of the total number of point sets, the straight line is considered valid, and the straight line angle range is fixed to 60-120 degrees in the vertical direction, and the point whose distance from the extreme edge point to the fitting straight line is within the preset pixel is a valid point. After the operation in this embodiment, the abnormal points on the edge of the wooden board can be filtered out.

[0062] Finally, if Fig.11 As shown, after completing the filtering of some textures and non-background areas, determine whether the ROI has reached the end. If so, filter the extreme points of the wooden board in a straight line, and traverse each valid group of edge point pairs in the list of extreme points of the wooden board edge after filtering the abnormal points on the wooden board edge to find the maximum distance as the maximum distance for debonding.

[0063] Further, it should be noted that there are arched debonding, unilateral warping debonding, and separation debonding types. In the processing process, there are also interferences of wood stripes and chipping. After actual measurement by the measurement method of the present application, different test results can be clearly obtained. Among them, Fig. 12A A schematic diagram showing the display of arched debonding is shown. Fig. 12B The schematic diagram of the display of one-side warping and debonding is shown. Fig. 12C A schematic diagram showing a separation type debonding is shown. Fig.12D A schematic diagram showing the display of plank stripe interference is shown. Fig.12E A display schematic diagram showing chipping interference.

[0064] A third aspect of the present invention provides a computer-readable storage medium, which includes a program for a method for measuring delamination of a wooden board edge banding. When the program for measuring delamination of a wooden board edge banding is executed by a processor, the steps of a method for measuring delamination of a wooden board edge banding as described in any one of the above items are implemented.

[0065] The present invention discloses a method, system and storage medium for measuring debonding of edge bands of wooden boards, which can effectively solve the influence of various interference problems existing in the industrial field on debonding measurement. The idea of ​​equally dividing the debonding area is adopted to improve the versatility of the image algorithm. When calculating the extreme points of the edges of wooden boards and edge bands, the idea of ​​multiple filtering is adopted for screening to improve the robustness of the edges of wooden boards and edge bands.

[0066] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0067] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0068] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0069] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0070] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

Claims

1. A method for measuring the debonding of a wood board edge band, characterized in that: The following steps are involved: Acquire a wooden board image, and perform preprocessing based on the wooden board image to obtain a processed image; Based on the processed image, the debonding area is segmented at equal intervals to obtain an area of ​​interest; Acquiring the scope of the debonding connected domain, which includes local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain; After obtaining the range of the debonding connected domain, local extreme points and edge point pairs are used to perform debonding measurement calculations.

2. The method for measuring the debonding of a wood board edge band according to claim 1, characterized in that: The obtaining of the wooden board image and performing preprocessing based on the wooden board image to obtain a processed image specifically includes: The wooden board image is normalized, grayscale scaled, and contrast adjusted to obtain an adjusted image; Gaussian filtering and noise removal are performed on the adjusted image to obtain the processed image.

3. The method for measuring debonding of edge banding of a wood board according to claim 2, characterized in that: Obtaining the scope of the debonding connected domain, including local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain, specifically includes: Performing a segmentation operation on the grayscale image of the region of interest, using a local mean to perform local adaptive segmentation and performing a closing operation to connect each edge region; Performing vertical projection on the grayscale image of the region of interest and calculating the mean value to simplify it into a one-dimensional vector; A connected domain analysis is performed on the one-dimensional vector to obtain a range of each connected domain, wherein the connected domain range corresponds to a range of edge extreme value points.

4. A method for measuring debonding of wood board edge banding according to claim 3, characterized in that: Calculating extreme points includes: Taking a first-order derivative of the one-dimensional vector to obtain positive and negative extreme point vectors; A minimum threshold of extreme points is set to filter the positive and negative extreme point vectors, and target extreme points are screened in combination with the connected domain range, wherein the filtering also includes filtering the wood board grain interference and non-background areas to obtain a list of extreme points on the edge of the wood board.

5. The method for measuring debonding of edge banding of a wood board according to claim 4, characterized in that: The method further includes filtering the board edge abnormal points from the board edge extreme value point list, specifically including: A straight line fitting algorithm is used for the list of extreme point values ​​on the edge of the plank using a random sampling consistency algorithm, wherein when the number of fitted straight line point sets exceeds a preset proportion of the total number of point sets, the straight line is considered valid; The fixed straight line angle range is 60-120 degrees in the vertical direction; The point whose distance from the fixed extreme edge point to the fitting straight line is within the preset pixel is regarded as a valid point.

6. A method for measuring debonding of wood board edge banding according to claim 5, characterized in that: After filtering outliers on the edge of the board, each valid pair of edge points in the extreme value point list of the edge of the board is traversed to find the maximum distance as the maximum distance for debonding.

7. A system for measuring the debonding of edge banding of wooden boards, characterized in that: The method comprises a memory and a processor, wherein the memory comprises a method for measuring the debonding of a wood board edge band, and when the method for measuring the debonding of a wood board edge band is executed by the processor, the following steps are implemented: Acquire a wooden board image, and perform preprocessing based on the wooden board image to obtain a processed image; Based on the processed image, the debonding area is divided at equal intervals to obtain an area of ​​interest, wherein the area of ​​interest is ten pixels high and the width is the width of the processed image; Acquiring the scope of the debonding connected domain, which includes local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain; After obtaining the range of the debonding connected domain, local extreme points and edge point pairs are used to perform debonding measurement calculations.

8. The wood board edge banding debonding measurement system according to claim 7, characterized in that: The obtaining of the wooden board image and performing preprocessing based on the wooden board image to obtain a processed image specifically includes: The wooden board image is normalized, grayscale scaled, and contrast adjusted to obtain an adjusted image; Gaussian filtering and noise removal are performed on the adjusted image to obtain the processed image.

9. The wood board edge banding debonding measurement system according to claim 8, characterized in that: Obtaining the scope of the debonding connected domain, including local adaptive threshold segmentation and vertical projection to obtain the scope of the debonding connected domain, specifically includes: Performing a segmentation operation on the grayscale image of the region of interest, using a local mean to perform local adaptive segmentation and performing a closing operation to connect each edge region; Performing vertical projection on the grayscale image of the region of interest and calculating the mean value to simplify it into a one-dimensional vector; A connected domain analysis is performed on the one-dimensional vector to obtain a range of each connected domain, wherein the connected domain range corresponds to a range of edge extreme value points.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a method program for measuring debonding of a wooden board edge band. When the method program for measuring debonding of a wooden board edge band is executed by a processor, the steps of the method for measuring debonding of a wooden board edge band as described in any one of claims 1 to 6 are implemented.

Images