A defect feature recognition method and system based on PBAT degradation film material

Abstract

The present invention discloses a defect feature recognition method and system based on PBAT degradation film materials, wherein the method includes collecting a blown film forming video to obtain grayscale images of each degradation film, obtaining key points for film surface defect analysis based on the grayscale difference characteristics between each pixel point in each degradation film grayscale image and the image as a whole, obtaining each film surface defect analysis area based on the texture difference characteristics and aggregation characteristics between all film surface defect analysis key points, constructing a defect morphology feature vector, obtaining a defect matching frame sequence based on the periodic characteristics of each film surface defect analysis area, calculating the non-even factor defect index of each film surface defect analysis area, and then obtaining the defect feature recognition result of each film surface defect analysis area. A defect feature recognition method and system based on PBAT degradation film materials provided by an embodiment of the present invention improves the accuracy of surface defect feature recognition of PBAT degradation film materials.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a defect feature recognition method and system based on PBAT degradation film materials. Background Art

[0002] PBAT biodegradable film material belongs to thermoplastic biodegradable plastics, which is a copolymer of butylene adipate and butylene terephthalate. It has the characteristics of both PBA and PBT, and has good ductility and elongation at break, as well as good heat resistance, impact resistance and biodegradability. With the improvement of people's environmental awareness and the increase in demand for biodegradable plastics, PBAT biodegradable film material, as an excellent biodegradable plastic, has broad application prospects and huge market potential.

[0003] Since the production process of PBAT degradable film materials includes complex and sophisticated process flows such as esterification reaction, polycondensation reaction, and viscosity enhancement treatment, various defects may appear on the surface of the PBAT degradable film materials during the production process of the PBAT degradable film materials, thereby reducing the performance of the PBAT degradable film materials. Therefore, defect identification of PBAT degradable film materials has extremely important practical significance.

[0004] However, since the defects on the surface of PBAT degradation film materials are usually subtle and similar to the characteristics of the PBAT degradation film materials themselves, the existing defect feature recognition technology has low accuracy in identifying the defect features on the surface of PBAT degradation film materials.

[0005] It can be seen that how to improve the accuracy of identifying surface defect characteristics of PBAT degradation film materials has become a technical problem that technical personnel in this field need to solve urgently. Summary of the invention

[0006] The present invention provides a defect feature recognition method and system based on PBAT degradation film materials, so as to solve the problem that the existing defect feature recognition technology has low accuracy in identifying surface defect features of PBAT degradation film materials, and improve the accuracy of identifying surface defect features of PBAT degradation film materials.

[0007] In order to solve the above technical problems, an embodiment of the present invention provides a defect feature recognition method based on PBAT degradation film materials.

[0008] In the process of blown film forming of the PBAT degradation film material, a blown film forming video is collected, and the blown film forming video is processed to obtain a grayscale image of each degradation film;

[0009] Acquire the key points for analyzing the film surface defects of each of the grayscale images of the degradation film based on the grayscale difference characteristics of each pixel point in each of the grayscale images of the degradation film and the image as a whole;

[0010] A clustering algorithm is used to cluster each of the key points for analyzing the defects on the film surface to obtain a film surface defect analysis region, wherein the clustering process is designed to use the texture difference characteristics and the aggregation characteristics between each of the key points for analyzing the defects on the film surface as the evaluation criteria of the clustering algorithm;

[0011] Constructing a defect morphology feature vector of each film surface defect analysis region based on the texture feature, grayscale difference feature and distribution feature of each film surface defect analysis key point;

[0012] Acquire a defect matching frame sequence for each of the film surface defect analysis regions based on the similarity between each of the defect morphology feature vectors and the overlapping area of ​​each of the film surface defect analysis regions;

[0013] The non-even factor defect index of each of the film surface defect analysis regions is calculated based on the equally spaced features of each of the defect matching frame sequences, and the defect feature recognition result of each of the film surface defect analysis regions is obtained based on the non-even factor defect index.

[0014] As one of the preferred solutions, the clustering algorithm is used to perform clustering processing on each of the key points of the membrane surface defect analysis to obtain a membrane surface defect analysis area, including:

[0015] Calculate the corresponding film surface defect homogeneity index based on the distance and texture difference characteristics between any two film surface defect analysis key points;

[0016] The normalized value of the film surface defect homogeneity index between all the film surface defect analysis key points in each of the degradation film grayscale images is used as a similarity measurement criterion of the clustering algorithm, and the clustering algorithm is used to divide all the film surface defect analysis key points in each of the degradation film grayscale images into a plurality of film surface defect clusters;

[0017] The minimum circumscribed rectangle of all the film surface defect analysis key points in each of the film surface defect clusters is used as the film surface defect analysis area.

[0018] As one of the preferred solutions, the method of calculating the corresponding film surface defect isotype index based on the distance and texture difference characteristics between any two film surface defect analysis key points includes:

[0019] Calculating the LBP feature descriptor of each pixel in each of the degradation film grayscale images, and recording the DTW distance between the LBP feature descriptors of two pixels as the corresponding neighborhood texture difference index;

[0020] Based on the linear relationship between the Euclidean distance between any two key points of film surface defect analysis and the neighborhood texture difference index, the corresponding film surface defect isotype index is calculated.

[0021] As one of the preferred solutions, the method of obtaining the key points of film surface defect analysis of each grayscale image of the degradation film based on the grayscale difference characteristics between each pixel point in each grayscale image of the degradation film and the image as a whole includes:

[0022] Based on the grayscale difference characteristics between each pixel point in each grayscale image of the degradation film and the image as a whole, the film surface grayscale difference factor of each pixel point in each grayscale image of the degradation film is calculated:

[0023]

[0024] in, is the film surface grayscale difference factor of the mth pixel in the i-th degradation film grayscale image, is the normalization function, is the gray value of the mth pixel in the gray image of the i-th degradation film, is the average gray value of all pixels in the gray image of the i-th degradation film;

[0025] A preset multiple of the average value of the film surface grayscale difference factors of all pixels in all the degradation film grayscale images is used as a grayscale difference threshold, and a pixel point in each of the degradation film grayscale images whose film surface grayscale difference factor is greater than the grayscale difference threshold is used as a key point for film surface defect analysis of each of the degradation film grayscale images.

[0026] As one of the preferred solutions, the defect morphology feature vector of each film surface defect analysis area is constructed based on the texture features, grayscale difference features and distribution features of each film surface defect analysis key point, including:

[0027] Acquire the defect key contour of each film surface defect analysis area based on the distribution characteristics of each film surface defect analysis key point;

[0028] Obtaining a chain code of each of the defect key contours, calculating the shape number of each of the defect key contours, and using the shape number of the defect key contour of each of the film surface defect analysis regions as a defect contour shape descriptor of each of the film surface defect analysis regions;

[0029] Constructing a gray level co-occurrence matrix of each of the film surface defect analysis regions, taking the entropy value of the gray level co-occurrence matrix of each of the film surface defect analysis regions as the texture complexity index of each of the film surface defect analysis regions, and taking the average value of all the film surface gray level difference factors of each of the film surface defect analysis regions as the film surface defect significance factor of each of the film surface defect analysis regions;

[0030] A defect morphology feature vector of each film surface defect analysis area is constructed based on the defect contour shape descriptor, the texture complexity index and the film surface defect significance factor.

[0031] As one of the preferred solutions, the defect key contour of each film surface defect analysis area is obtained based on the distribution characteristics of each film surface defect analysis key point, including:

[0032] For each of the membrane surface defect analysis key points in each of the membrane surface defect analysis regions, a line segment connecting each of the membrane surface defect analysis key points and a center point is used as an eccentric line of each of the membrane surface defect analysis key points, and an angle between the eccentric line of each of the membrane surface defect analysis key points and a horizontal direction is used as an eccentric angle of each of the membrane surface defect analysis key points, and all of the membrane surface defect analysis key points in each of the membrane surface defect analysis regions are arranged in ascending order according to the corresponding eccentric angles, to obtain a defect key point sequence for each of the membrane surface defect analysis regions;

[0033] All the membrane surface defect analysis key points in each membrane surface defect analysis area are connected in sequence according to the defect key point sequence, and the first membrane surface defect analysis key point and the last membrane surface defect analysis key point in the defect key point sequence are connected to obtain the defect key contour of each membrane surface defect analysis area.

[0034] As one of the preferred solutions, the defect matching frame sequence of each film surface defect analysis area is obtained based on the similarity between each defect morphology feature vector and the overlapping area of ​​each film surface defect analysis area, including:

[0035] All the grayscale images of the degraded film are numbered in the order of their acquisition time, each of the film surface defect analysis regions in each of the grayscale images of the degraded film is used as the film surface defect analysis region to be tested, and each of the film surface defect analysis regions in the remaining grayscale images of the degraded film is used in turn as the film surface defect analysis region to be matched;

[0036] The defect analysis area of ​​the film surface to be tested and the defect analysis area of ​​the film surface to be matched are graded to obtain corresponding defect morphology difference index and overlapping area analysis results; based on the linear relationship between the defect morphology difference index and the overlapping area analysis result, the corresponding non-shift defect matching index is calculated;

[0037] Performing template matching processing on the film surface defect analysis area to be tested and the film surface defect analysis area to be matched to obtain a matching result of the film surface defect analysis area to be tested and the film surface defect analysis area to be matched, wherein the template matching processing is designed to use the normalized processing result of the non-shifted defect matching index as a matching criterion of the template matching algorithm;

[0038] All the grayscale images of the degraded film in the film surface defect analysis region to be matched that are successfully matched with the film surface defect analysis template to be tested are numbered in ascending order as a defect matching frame sequence of the film surface defect analysis region to be tested.

[0039] As one preferred solution, the graded treatment includes a first-level treatment and a second-level treatment;

[0040] The first-level processing is configured to perform a similarity analysis on the defect morphology feature vector of the defect analysis area of ​​the film surface to be tested and the defect morphology feature vector of the defect analysis area of ​​the film surface to be matched, so as to obtain a corresponding defect morphology dissimilarity index;

[0041] The second level processing is configured to analyze the defect analysis area of ​​the film surface to be tested and the defect analysis area of ​​the film surface to be matched to obtain corresponding overlapping area analysis results.

[0042] As one of the preferred solutions, the non-even factor defect index of each film surface defect analysis area is calculated based on the equally spaced characteristics of each defect matching frame sequence, including:

[0043] A first-order difference is performed on the defect matching frame sequence of each of the film surface defect analysis areas to obtain a matching frame difference sequence of each of the film surface defect analysis areas, and each of the matching frame difference sequences is processed to obtain a non-even factor defect index of each of the film surface defect analysis areas.

[0044] Yet another embodiment of the present invention provides a defect feature recognition system based on PBAT degradation film materials, the system comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, a defect feature recognition method based on PBAT degradation film materials as described above is implemented.

[0045] Compared with the prior art, the embodiments of the present invention have the following advantages:

[0046] (1) In the process of blown film forming of PBAT degradation film materials, a blown film forming video is collected, and the blown film forming video is processed to obtain grayscale images of each degradation film. The grayscale difference characteristics of the pixels in each degradation film grayscale image are first analyzed, and the film surface defect analysis key points of each degradation film grayscale image are obtained based on the grayscale difference characteristics of each pixel in each degradation film grayscale image and the image as a whole, so as to distinguish the pixel points in the defect area from the pixel points in the normal area. Subsequently, only the film surface defect analysis key points are analyzed, thereby improving the real-time defect feature recognition of the produced PBAT degradation film material in the process of blown film forming of PBAT degradation film materials;

[0047] (2) Since a variety of defects may appear on the surface of PBAT degradation film materials, and defects such as discoloration and wrinkles usually appear in patches, the film surface defect analysis areas of each degradation film grayscale image are obtained based on the texture difference characteristics between all film surface defect analysis key points in each degradation film grayscale image and the clustering characteristics of all film surface defect analysis key points, so that all pixels in the same defect area are analyzed as a whole, thereby improving the reliability of defect feature recognition of the produced PBAT degradation film materials;

[0048] (3) Considering that the defects on the surface of PBAT degradation film materials caused by accidental factors are often irregular, and the defects caused by non-accidental factors will appear similar defects at a specific interval length, the defect morphology feature vector of each film surface defect analysis area is first constructed based on the texture characteristics, grayscale difference characteristics and distribution characteristics of each film surface defect analysis key point in each film surface defect analysis area to describe the defect morphology characteristics of each film surface defect analysis area;

[0049] (4) The periodicity of the appearance of the film surface defect analysis area is analyzed again, and the defect matching frame sequence of each film surface defect analysis area is obtained based on the similarity between each defect morphology feature vector and the overlapping area of ​​each film surface defect analysis area. Then, the non-even factor defect index of each film surface defect analysis area is calculated based on the equal interval characteristics of each defect matching frame sequence, and then the defect feature recognition result of each film surface defect analysis area is obtained. The similarity and equal interval of the appearance of the film surface defect analysis area are comprehensively considered to improve the accuracy of the defect feature recognition of the surface of the PBAT degradation film material. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of a defect feature identification process based on PBAT degradation film material in one embodiment of the present invention;

[0051] Figure 2 is a schematic diagram of an eccentric line and an eccentric angle in one embodiment of the present invention;

[0052] Figure 3 is a schematic diagram of a critical defect profile in one embodiment of the present invention;

[0053] Figure 4 It is a schematic diagram of obtaining defect feature recognition results in one embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] In the description of this application, the terms "first", "second", "third", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", "third", etc. may explicitly or implicitly include one or more of the feature. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0056] In the description of the present application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. The terms "vertical", "horizontal", "left", "right", "upper", "lower" and similar expressions used herein are only for illustrative purposes, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0057] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those commonly understood by those skilled in the art. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by specific circumstances.

[0058] An embodiment of the present invention provides a defect feature recognition method based on PBAT degradation film material. For details, see Figure 1 , Figure 1 Shown is a schematic diagram of a defect feature identification process based on PBAT degradation film material in one embodiment of the present invention.

[0059] Step S1: During the film blowing process of the PBAT degradation film material, a film blowing video is collected, and the film blowing video is processed to obtain a grayscale image of each degradation film.

[0060] The PBAT degradation film material may cause various defects due to equipment abnormalities during the blown film molding process, which in turn affects the overall quality and production efficiency of the PBAT degradation film material. In order to ensure product quality and production efficiency, it is necessary to identify the defect characteristics of the produced PBAT degradation film material during the blown film molding process of the PBAT degradation film material.

[0061] Specifically, a CCD industrial camera is placed directly above the film blowing machine, and the produced PBAT degradation film material is photographed during the film blowing molding process of the PBAT degradation film material to obtain a film blowing molding video. In order to facilitate a detailed analysis of the film blowing molding video, the film blowing molding video is framed to obtain a color image of each degradation film, and then each degradation film color image is grayed to obtain a grayscale image of each degradation film.

[0062] Step S2: obtaining the key points for analyzing the film surface defects of each degradation film grayscale image based on the grayscale difference characteristics between each pixel point in each degradation film grayscale image and the entire image.

[0063] When defects such as crystal points, discoloration, and wrinkles appear on the surface of the degradation film material, the grayscale characteristics of the defective area and the normal defect-free area are different. Therefore, the grayscale difference characteristics of the pixels in each degradation film grayscale image can be analyzed to distinguish the pixels in the defective area from those in the normal area.

[0064] Specifically, based on the grayscale difference characteristics between each pixel point in each grayscale image of the degradation film and the entire image, the film surface grayscale difference factor of each pixel point in each grayscale image of the degradation film is calculated:

[0065]

[0066] in, is the film surface grayscale difference factor of the mth pixel in the i-th degradation film grayscale image, is the normalization function, is the gray value of the mth pixel in the gray image of the i-th degradation film, is the average grayscale value of all pixels in the grayscale image of the i-th degradation film.

[0067] It should be noted that the film surface grayscale difference factor of each pixel in each degradation film grayscale image is used to measure the relative difference in the grayscale value of each pixel in each degradation film grayscale image and the overall image. When the absolute value of the difference between the grayscale value of a certain pixel and the remaining pixels is larger than the grayscale mean of the overall image, it means that the relative difference in the grayscale value of the pixel and the overall image is greater, and the pixel is more likely to be located in the defect area.

[0068] Furthermore, a preset multiple of the average value of the film surface grayscale difference factors of all pixels in all degraded film grayscale images is used as a grayscale difference threshold, and the pixel points in each degraded film grayscale image whose film surface grayscale difference factor is greater than the grayscale difference threshold are used as key points for film surface defect analysis.

[0069] It should be noted that the preset multiple is a manually preset value, and the range of the preset multiple is 0 to 1. In this embodiment, the preset multiple is 0.7. As other implementation methods, the implementer can select it at will, and this application does not impose any special restrictions on this. The grayscale difference threshold is set according to the film surface grayscale difference factor of all pixels in the grayscale image of all degradation films, and the pixels with a large difference in grayscale value from the overall grayscale image of the degradation film are selected as the key points for film surface defect analysis. Only the key points for film surface defect analysis are analyzed later, which improves the real-time defect feature identification of the produced PBAT degradation film material during the film blowing molding process of the PBAT degradation film material.

[0070] Step S3: acquiring each film surface defect analysis region of each degraded film grayscale image based on the texture difference characteristics between all film surface defect analysis key points in each degraded film grayscale image and the aggregation characteristics of all film surface defect analysis key points.

[0071] Considering that there may be multiple defects on the surface of PBAT degradation film materials, and defects such as discoloration and wrinkles usually appear in patches, when identifying defects on the surface of PBAT degradation film materials, all pixels in the same defect area need to be analyzed as a whole. By clustering the key points of film surface defect analysis in each degradation film grayscale image, it is possible to divide the defects in different areas of the PBAT degradation film material surface.

[0072] Specifically, the LBP feature descriptor of each pixel in each degradation film grayscale image is calculated, and the DTW distance between the LBP feature descriptors of two pixels is recorded as the neighborhood texture difference index between the two pixels.

[0073] It should be noted that the LBP feature descriptor is an operator used to describe local texture features of an image, and the DTW distance is used to describe the similarity between two sequences. The calculation of the LBP feature descriptor and the DTW distance are both well-known technologies, and this application will not elaborate on them here.

[0074] Furthermore, based on the linear relationship between the Euclidean distance between any two film surface defect analysis key points in each degradation film grayscale image and the neighborhood texture difference index, the degradation film defect isotype index between any two film surface defect analysis key points is calculated.

[0075] It should be noted that when the Euclidean distance between the two key points of film surface defect analysis is smaller and the neighborhood texture difference index is smaller, it means that the distance between the two key points of film surface defect analysis is closer, and the texture features of the surrounding areas are more similar, the more likely they belong to the same defect area, and the larger the degradation film defect isotype index value.

[0076] As an embodiment of the present application, the product of the Euclidean distance and the neighborhood texture difference index between any two key points of film surface defect analysis in each degradation film grayscale image is used as the degradation film defect heterogeneity factor between the any two key points of film surface defect analysis, and the calculation result of an exponential function with a natural constant as the base and the opposite number of the degradation film defect heterogeneity factor as the exponent is used as the degradation film defect homogeneity index between the any two key points of film surface defect analysis.

[0077] Furthermore, the normalized value of the degradation film defect homogeneity index between all the film surface defect analysis key points in each degradation film grayscale image is used as the similarity measurement criterion of the clustering algorithm, and the clustering algorithm is used to divide all the film surface defect analysis key points in each degradation film grayscale image into multiple degradation film defect clusters.

[0078] It should be noted that clustering algorithm is a well-known technology. In this embodiment, the k-means clustering algorithm is selected to divide the degradation film defect clustering clusters for all the key points of film surface defect analysis in each degradation film grayscale image. The selection of clustering algorithm is other implementation methods that the implementer can choose at his own discretion, and this application does not impose any special restrictions on this.

[0079] Furthermore, the minimum circumscribed rectangle of all the film surface defect analysis key points within each of the degradation film defect clusters is used as the film surface defect analysis region.

[0080] Step S4: constructing a defect morphology feature vector of each film surface defect analysis region based on the texture features, grayscale difference features and distribution features of each film surface defect analysis key point in each film surface defect analysis region.

[0081] It should be noted that since the defects on the surface of PBAT degradation film materials may be caused by accidental factors or non-accidental factors, the sensitivity of defect detection and defect identification should be improved for subtle defects on the surface of PBAT degradation film materials caused by non-accidental factors such as equipment failure, so as to conduct a more in-depth analysis of the cause of the failure later.

[0082] Considering that the defects on the surface of PBAT degradation film materials caused by accidental factors are often irregular, and the defects caused by non-accidental factors will appear similar defects at specific intervals, therefore, by analyzing the periodicity of the defect analysis area on the membrane surface, the surface defect characteristics of PBAT degradation film materials can be identified.

[0083] In order to analyze the similarity of the film surface defect analysis areas, it is necessary to first analyze the texture features, grayscale difference features of each film surface defect analysis area and the distribution features of each film surface defect analysis key point in each film surface defect analysis area.

[0084] Specifically, for each membrane surface defect analysis key point in each membrane surface defect analysis area, the line segment connecting each membrane surface defect analysis key point and the center point is used as the centrifugal line of each membrane surface defect analysis key point, and the angle between the centrifugal line of each membrane surface defect analysis key point and the horizontal direction is used as the centrifugal angle of each membrane surface defect analysis key point. All membrane surface defect analysis key points in each membrane surface defect analysis area are arranged in ascending order according to the corresponding centrifugal angles to obtain a defect key point sequence for each membrane surface defect analysis area.

[0085] In one embodiment of the present invention, the schematic diagram of the centrifugal line and the centrifugal angle is as follows Figure 2 shown. Figure 2 In the figure, s is the center point of the membrane surface defect analysis area, line segment a is the eccentric line of the key point t of the membrane surface defect analysis, and angle p is the eccentric angle of the key point t of the membrane surface defect analysis.

[0086] Furthermore, all membrane surface defect analysis key points in each membrane surface defect analysis area are connected in sequence according to the defect key point sequence, and the first membrane surface defect analysis key point and the last membrane surface defect analysis key point in the defect key point sequence of each membrane surface defect analysis area are connected to obtain the defect key contour of each membrane surface defect analysis area.

[0087] In one embodiment of the present invention, a schematic diagram of a critical defect profile is shown as follows: Figure 3shown.

[0088] Furthermore, the chain code of each defect key contour is obtained, the shape number of each defect key contour is calculated, and the shape number of the defect key contour of each membrane surface defect analysis area is used as the defect contour shape descriptor of each membrane surface defect analysis area.

[0089] The gray level co-occurrence matrix of each membrane surface defect analysis area is constructed, and the entropy value of the gray level co-occurrence matrix of each membrane surface defect analysis area is used as the texture complexity index of each membrane surface defect analysis area. The average value of the membrane surface gray level difference factors of all membrane surface defect analysis key points in each membrane surface defect analysis area is used as the membrane surface defect significance factor of each membrane surface defect analysis area.

[0090] It should be noted that the chain code is a method of describing a curve or boundary using the coordinates of the starting point of the curve and the direction code of the boundary point. The shape number is a boundary descriptor based on the chain code. The grayscale co-occurrence matrix reveals the texture information of the image by counting the grayscale value relationship between different pixels in the image. The entropy value of the grayscale co-occurrence matrix indicates the complexity of the grayscale distribution of the image. The acquisition of the chain code and the calculation of the shape number, the construction of the grayscale co-occurrence matrix and the calculation of the entropy value of the grayscale co-occurrence matrix are all well-known technologies, and this embodiment will not be described in detail here.

[0091] The defect morphology feature vector of each film surface defect analysis area is constructed based on the defect contour shape descriptor, texture complexity index and film surface defect significance factor of each film surface defect analysis area:

[0092]

[0093] in, is the defect morphology feature vector of the jth film surface defect analysis area in the i-th degradation film grayscale image, is the defect contour shape descriptor of the jth film surface defect analysis area in the i-th degradation film grayscale image, is the texture complexity index of the jth film surface defect analysis area in the i-th degradation film grayscale image, is the film surface defect significance factor of the jth film surface defect analysis area in the i-th degradation film grayscale image.

[0094] Step S5: acquiring a defect matching frame sequence of each film surface defect analysis region based on the similarity between each defect morphology feature vector and the overlapping area of ​​each film surface defect analysis region.

[0095] It should be noted that when the production equipment of the PBAT degradation film material fails, the PBAT degradation film material produced will have similar defects at intervals of specific lengths. Therefore, before the periodic analysis of the film surface defect analysis area, it is necessary to first analyze the similarity between each film surface defect analysis area in each degradation film grayscale image and all film surface defect analysis areas in the remaining degradation film grayscale images.

[0096] Specifically, all the grayscale images of the degraded film are numbered in the order of their acquisition time, each film surface defect analysis area in each grayscale image of the degraded film is used as the film surface defect analysis area to be tested, and each film surface defect analysis area in the remaining grayscale images of the degraded film is used in turn as the film surface defect analysis area to be matched.

[0097] Then, a similarity analysis is performed on the defect morphology feature vector of the defect analysis region of the film surface to be tested and the defect morphology feature vector of the defect analysis region of the film surface to be matched to obtain the corresponding defect morphology dissimilarity index.

[0098] As an embodiment of the present application, the distance between the defect morphology feature vectors of the film surface defect analysis region to be tested and the film surface defect analysis region to be matched is used as the defect morphology difference index between the film surface defect analysis region to be tested and the film surface defect analysis region to be matched.

[0099] Then, the defect analysis area of ​​the film surface to be tested and the defect analysis area of ​​the film surface to be matched are analyzed to obtain the corresponding overlapping area analysis results. Based on the linear relationship between the defect morphology difference index and the overlapping area analysis results, the corresponding non-shift defect matching index is calculated.

[0100] It should be noted that, when the overlapping area of ​​the film surface defect analysis region to be tested and the film surface defect analysis region to be matched is larger and the defect morphology difference index between the film surface defect analysis region to be tested and the film surface defect analysis region to be matched is smaller, it means that the film surface defect analysis region to be tested and the film surface defect analysis region to be matched are more likely to be located in the same area of ​​different degraded film grayscale images, and the morphological features of the film surface defect analysis region to be tested and the film surface defect analysis region to be matched are similar, and the non-shift defect matching index value is larger.

[0101] As an embodiment of the present application, the ratio of the overlapping area of ​​the film surface defect analysis region to be tested and the film surface defect analysis region to be matched to the defect morphology difference index is used as the non-shift defect matching index between the film surface defect analysis region to be tested and the film surface defect analysis region to be matched.

[0102] Furthermore, the film surface defect analysis area to be tested and the film surface defect analysis area to be matched are subjected to template matching processing to obtain matching results of the film surface defect analysis area to be tested and the film surface defect analysis area to be matched, and the template matching processing is designed to use the normalized processing result of the non-shifted defect matching index as the matching criterion of the template matching algorithm.

[0103] As an embodiment of the present application, all non-shifted defect matching indices are normalized to obtain a non-shifted normalized defect matching index between the film surface defect analysis region to be tested and the film surface defect analysis region to be matched. When the non-shifted normalized defect matching index is greater than a preset matching threshold, it is determined that the film surface defect analysis region to be tested and the film surface defect analysis region to be matched are successfully matched; otherwise, it is determined that the film surface defect analysis region to be tested and the film surface defect analysis region to be matched are not successfully matched.

[0104] All the grayscale images of the degraded film where the defect analysis region of the film surface to be matched is located and which successfully matches the defect analysis region of the film surface to be tested are numbered in ascending order as a defect matching frame sequence of the defect analysis region of the film surface to be tested.

[0105] It should be noted that the preset matching threshold is a manually preset value. In this embodiment, the preset matching threshold is taken as 0.8. As for the value of the preset matching threshold, as other implementation methods, the implementer can select it at will, and this application does not impose any special restrictions on this.

[0106] Step S6: calculating the non-even factor defect index of each film surface defect analysis area based on the equally spaced features of each defect matching frame sequence, and obtaining the defect feature recognition result of each film surface defect analysis area based on the non-even factor defect index.

[0107] Specifically, a first-order difference is performed on the defect matching frame sequence of each membrane surface defect analysis area to obtain a matching frame difference sequence of each membrane surface defect analysis area, and the matching frame difference sequence of each membrane surface defect analysis area is processed to obtain a non-even factor defect index of each membrane surface defect analysis area.

[0108] It should be noted that when the values ​​of each element in the matching frame differential sequence of the membrane surface defect analysis area are close to 0, it means that the time intervals of the occurrence of the membrane surface defect analysis area are more likely to be equal, the membrane surface defect analysis area is more likely to have defects caused by equipment failure, and the non-even factor defect index value is larger.

[0109] As an embodiment of the present application, the average value of all elements in the matching frame differential sequence of each film surface defect analysis area is used as the defect non-equidistant emergence factor of each film surface defect analysis area, and the calculation result of an exponential function with a natural constant as the base and the defect non-equidistant emergence factor as the exponent is used as the non-even factor defect index of each film surface defect analysis area.

[0110] It should be noted that the defect feature identification results include defects caused by accidental factors and defects caused by non-accidental factors.

[0111] In one embodiment of the present invention, a schematic diagram of obtaining defect feature recognition results is shown in FIG. Figure 4 shown.

[0112] Yet another embodiment of the present invention provides a defect feature recognition system based on PBAT degradation film materials, the system comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, a defect feature recognition method based on PBAT degradation film materials as described above is implemented.

[0113] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A defect feature identification method based on PBAT degradation film material, characterized in that: include: In the process of blown film forming of the PBAT degradation film material, a blown film forming video is collected, and the blown film forming video is processed to obtain a grayscale image of each degradation film; Acquire the key points for analyzing the film surface defects of each of the grayscale images of the degradation film based on the grayscale difference characteristics of each pixel point in each of the grayscale images of the degradation film and the image as a whole; A clustering algorithm is used to cluster each of the key points for analyzing the defects on the film surface to obtain a film surface defect analysis region, wherein the clustering process is designed to use the texture difference characteristics and the aggregation characteristics between each of the key points for analyzing the defects on the film surface as the evaluation criteria of the clustering algorithm; Constructing a defect morphology feature vector of each film surface defect analysis region based on the texture feature, grayscale difference feature and distribution feature of each film surface defect analysis key point; Acquire a defect matching frame sequence for each of the film surface defect analysis regions based on the similarity between each of the defect morphology feature vectors and the overlapping area of ​​each of the film surface defect analysis regions; Calculating a non-even factor defect index for each of the film surface defect analysis regions based on the equally spaced features of each of the defect matching frame sequences, and obtaining a defect feature recognition result for each of the film surface defect analysis regions based on the non-even factor defect index; The step of acquiring a defect matching frame sequence for each of the film surface defect analysis regions based on the similarity between the defect morphology feature vectors and the overlapping area of ​​each of the film surface defect analysis regions comprises: All the grayscale images of the degraded film are numbered in the order of their acquisition time, each of the film surface defect analysis regions in each of the grayscale images of the degraded film is used as the film surface defect analysis region to be tested, and each of the film surface defect analysis regions in the remaining grayscale images of the degraded film is used in turn as the film surface defect analysis region to be matched; The defect analysis area of ​​the film surface to be tested and the defect analysis area of ​​the film surface to be matched are graded to obtain corresponding defect morphology difference index and overlapping area analysis results; based on the linear relationship between the defect morphology difference index and the overlapping area analysis result, the corresponding non-shift defect matching index is calculated; Performing template matching processing on the film surface defect analysis area to be tested and the film surface defect analysis area to be matched to obtain a matching result of the film surface defect analysis area to be tested and the film surface defect analysis area to be matched, wherein the template matching processing is designed to use the normalized processing result of the non-shifted defect matching index as a matching criterion of the template matching algorithm; Arrange the numbers of all the degraded film grayscale images of the film surface defect analysis region to be matched that have successfully matched the template of the film surface defect analysis region to be tested in ascending order as the defect matching frame sequence of the film surface defect analysis region to be tested; The hierarchical treatment includes a first-level treatment and a second-level treatment; The first-level processing is configured to perform a similarity analysis on the defect morphology feature vector of the defect analysis area of ​​the film surface to be tested and the defect morphology feature vector of the defect analysis area of ​​the film surface to be matched, so as to obtain a corresponding defect morphology dissimilarity index; The second level processing is configured to analyze the defect analysis area of ​​the film surface to be tested and the defect analysis area of ​​the film surface to be matched to obtain corresponding overlapping area analysis results.

2. A defect feature identification method based on PBAT degradation film material according to claim 1, characterized in that: The clustering algorithm is used to perform clustering processing on each of the key points of the membrane surface defect analysis to obtain a membrane surface defect analysis area, including: Calculate the corresponding film surface defect homogeneity index based on the distance and texture difference characteristics between any two film surface defect analysis key points; The normalized value of the film surface defect homogeneity index between all the film surface defect analysis key points in each of the degradation film grayscale images is used as a similarity measurement criterion of the clustering algorithm, and the clustering algorithm is used to divide all the film surface defect analysis key points in each of the degradation film grayscale images into a plurality of film surface defect clusters; The minimum circumscribed rectangle of all the film surface defect analysis key points in each of the film surface defect clusters is used as the film surface defect analysis area.

3. A defect feature identification method based on PBAT degradation film material according to claim 2, characterized in that: The method of calculating the corresponding film surface defect isotype index based on the distance between any two film surface defect analysis key points and the texture difference characteristics includes: Calculating the LBP feature descriptor of each pixel in each of the degradation film grayscale images, and recording the DTW distance between the LBP feature descriptors of two pixels as the corresponding neighborhood texture difference index; Based on the linear relationship between the Euclidean distance between any two key points of film surface defect analysis and the neighborhood texture difference index, the corresponding film surface defect isotype index is calculated.

4. A defect feature identification method based on PBAT degradation film material according to claim 1, characterized in that: The step of obtaining the key points of the film surface defect analysis of each of the grayscale images of the degradation film based on the grayscale difference characteristics of each pixel point in each of the grayscale images of the degradation film and the image as a whole comprises: Based on the grayscale difference characteristics between each pixel point in each grayscale image of the degradation film and the image as a whole, the film surface grayscale difference factor of each pixel point in each grayscale image of the degradation film is calculated: ； in, is the film surface grayscale difference factor of the mth pixel in the i-th degradation film grayscale image, is the normalization function, is the gray value of the mth pixel in the gray image of the i-th degradation film, is the average gray value of all pixels in the gray image of the i-th degradation film; A preset multiple of the average value of the film surface grayscale difference factors of all pixels in all the degradation film grayscale images is used as a grayscale difference threshold, and a pixel point in each of the degradation film grayscale images whose film surface grayscale difference factor is greater than the grayscale difference threshold is used as a key point for film surface defect analysis of each of the degradation film grayscale images.

5. A defect feature identification method based on PBAT degradation film material according to claim 4, characterized in that: The defect morphology feature vector of each film surface defect analysis area is constructed based on the texture feature, grayscale difference feature and distribution feature of each film surface defect analysis key point, including: Acquire the defect key contour of each of the film surface defect analysis areas based on the distribution characteristics of each of the film surface defect analysis key points; Obtaining a chain code of each of the defect key contours, calculating the shape number of each of the defect key contours, and using the shape number of the defect key contour of each of the film surface defect analysis regions as a defect contour shape descriptor of each of the film surface defect analysis regions; Constructing a gray level co-occurrence matrix of each of the film surface defect analysis regions, taking the entropy value of the gray level co-occurrence matrix of each of the film surface defect analysis regions as the texture complexity index of each of the film surface defect analysis regions, and taking the average value of all the film surface gray level difference factors of each of the film surface defect analysis regions as the film surface defect significance factor of each of the film surface defect analysis regions; A defect morphology feature vector of each film surface defect analysis area is constructed based on the defect contour shape descriptor, the texture complexity index and the film surface defect significance factor.

6. A defect feature identification method based on PBAT degradation film material according to claim 5, characterized in that: The step of obtaining the defect key contour of each film surface defect analysis area based on the distribution characteristics of each film surface defect analysis key point comprises: For each of the membrane surface defect analysis key points in each of the membrane surface defect analysis regions, a line segment connecting each of the membrane surface defect analysis key points and a center point is used as an eccentric line of each of the membrane surface defect analysis key points, and an angle between the eccentric line of each of the membrane surface defect analysis key points and a horizontal direction is used as an eccentric angle of each of the membrane surface defect analysis key points, and all of the membrane surface defect analysis key points in each of the membrane surface defect analysis regions are arranged in ascending order according to the corresponding eccentric angles, to obtain a defect key point sequence for each of the membrane surface defect analysis regions; All the membrane surface defect analysis key points in each membrane surface defect analysis area are connected in sequence according to the defect key point sequence, and the first membrane surface defect analysis key point and the last membrane surface defect analysis key point in the defect key point sequence are connected to obtain the defect key contour of each membrane surface defect analysis area.

7. A defect feature identification method based on PBAT degradation film material according to claim 1, characterized in that: The calculating of the non-even factor defect index of each film surface defect analysis area based on the equally spaced features of each defect matching frame sequence comprises: A first-order difference is performed on the defect matching frame sequence of each of the film surface defect analysis areas to obtain a matching frame difference sequence of each of the film surface defect analysis areas, and each of the matching frame difference sequences is processed to obtain a non-even factor defect index of each of the film surface defect analysis areas.

8. A defect feature recognition system based on PBAT degradation film material, characterized in that: The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, a defect feature identification method based on PBAT degradation film material as described in any one of claims 1 to 7 is implemented.

Images