A method and system for intelligently detecting defective PBAT degradation membrane materials

Abstract

The present invention discloses a method and system for intelligently detecting defective PBAT degradation film materials, the method comprising collecting images of each PBAT degradation film material to be tested and calculating the flatness of each image, screening out the PBAT degradation film materials to be tested corresponding to the images whose flatness is lower than the standard flatness threshold; performing light compensation processing on the images from several angles to obtain light compensation images of each PBAT degradation film material to be tested; screening the light compensation images at each angle by contrast learning to obtain the image corresponding to the light angle with the smallest loss of the contour characteristics of the PBAT degradation film particles, obtaining the crack information therein, and screening out the PBAT degradation film materials to be tested corresponding to each crack information. The present invention identifies defective products in the material by detecting the surface flatness and thickness; identifies the cracks of the PBAT degradation film in the image by edge detection, and eliminates the mechanical artifacts generated by the tiny mechanical vibration of the film blowing machine during the image acquisition process by calibration, thereby improving the accuracy of crack identification.

Description

Technical Field

[0001] The present invention relates to the field of thin film detection technology, and in particular to an intelligent detection method and system for defective PBAT degradation film materials. Background Art

[0002] As consumers' demand for environmentally friendly products increases and environmental awareness improves, the market demand for degradable plastics is also growing. As one of the representatives of biodegradable plastics, the market demand for PBAT degradable film will also grow accordingly.

[0003] PBAT degradation film is a kind of macromolecular compound synthesized from petroleum. It has excellent biodegradability, ductility, elongation at break, heat resistance and impact resistance. It is widely used in plastic packaging films, agricultural mulch films and other film materials, which can effectively reduce environmental pollution. However, due to unreasonable resin raw material formula, improper processing conditions and other reasons, the existing PBAT degradation film will have defects, thus becoming a defective product. In order to prevent defective products from entering the market, they need to be screened and eliminated. The existing technology often checks for defective products by manual selection, which has low elimination efficiency and poor accuracy of manual detection methods.

[0004] It can be seen that how to accurately identify defective PBAT plastic films has become a technical problem that technical personnel in this field need to solve urgently. Summary of the invention

[0005] The present invention provides an intelligent detection method for defective PBAT degradation film materials, so as to realize accurate screening of defective PBAT plastic film products.

[0006] In order to solve the above technical problems, an embodiment of the present invention provides an intelligent detection method for defective PBAT degradation film materials, comprising:

[0007] Surface acoustic wave images are collected for each PBAT degradation film material to be tested to obtain an acoustic wave image to be tested, and optical images are collected for each PBAT degradation film material to be tested to obtain an optical image to be tested.

[0008] The pixel matrix of each acoustic wave image to be detected is traversed and obtained, and a plurality of image segmentation centers in the acoustic wave image to be detected are obtained by a corner point detection algorithm; and each acoustic wave image to be detected is divided into a plurality of areas to be detected according to each of the image segmentation centers.

[0009] The regional flatness in each of the regions to be detected and the image flatness of the acoustic wave image to be detected are calculated, and the PBAT degradation film materials to be detected corresponding to the images to be detected whose regional flatness and image flatness are lower than the standard flatness threshold are screened out to obtain a first PBAT degradation film qualified product.

[0010] The optical image of each of the PBAT degradation film materials to be tested is subjected to light compensation processing from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be tested.

[0011] The first light compensation images at various angles are screened by contrast learning to obtain a second light compensation image corresponding to a light angle with the smallest loss of contour features of the PBAT degradation film particles; edge detection is performed on the second light compensation image to obtain texture information of the PBAT degradation film material in the image, each of the texture information is analyzed and processed to obtain crack information therein, and the PBAT degradation film materials to be tested corresponding to each crack information are screened out to obtain a second PBAT degradation film qualified product.

[0012] Furthermore, the first light compensation images at various angles are screened by contrast learning to obtain the second light compensation image corresponding to the light angle with the smallest loss of the contour characteristics of the PBAT degradation film particles, including:

[0013] Obtain the particle profile information of the standard PBAT degradation film, extract the particle profile features of the standard PBAT degradation film according to the particle profile information of the standard PBAT degradation film, extract the particle profile features of the first light compensation image at each angle, obtain the particle profile features of the PBAT degradation film to be tested, and calculate the absolute mean difference between each particle profile feature of the PBAT degradation film to be tested and the particle profile feature of the standard PBAT degradation film:

[0014] ;

[0015] in, is the expected value of the random light noise following the normal distribution, is the expected value that the particle profile characteristics of the PBAT degradation film to be tested obey the data distribution, This is the characteristic diagram of the particle profile of the standard PBAT degradation film. This is the contour characteristic diagram of the PBAT degradation film particles to be tested.

[0016] According to the absolute mean difference value, the image corresponding to the light angle with the smallest loss of the contour feature of the PBAT degradation film particles is selected as the second light compensation image.

[0017] Furthermore, the surface acoustic wave image acquisition of each PBAT degradation film material to be tested to obtain the acoustic wave image to be tested includes:

[0018] Using laser surface acoustic wave technology, the laser is focused on the surface of the PBAT degradation film material to be tested, the surface acoustic wave is excited, and several first acoustic wave images are generated based on the reflected acoustic wave signal.

[0019] The first acoustic wave images are de-noised by using a median filter to obtain a second acoustic wave image.

[0020] The second acoustic wave images are registered and fused according to the feature points of the second acoustic wave images to obtain an acoustic wave image to be detected.

[0021] Furthermore, the calculating of the area flatness in each of the areas to be detected and the image flatness of the acoustic wave image to be detected includes:

[0022] For the acoustic wave image to be detected and each of the regions to be detected, the image is grayed out, and a gray level co-occurrence matrix is ​​calculated by comparing the gray levels of adjacent pixels in the image.

[0023] Contrast features are extracted from each gray level co-occurrence matrix, and the flatness of the corresponding matrix is ​​calculated, so as to obtain the image flatness of the acoustic wave image to be detected and the regional flatness of each region to be detected.

[0024] Furthermore, the optical image of each of the PBAT degradation film materials to be tested is subjected to light compensation processing from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be tested, including:

[0025] The first optical image of each of the PBAT degradation film materials to be tested is segmented to obtain a background image in each of the light images.

[0026] The average brightness of each pixel in each of the background images is calculated, and the average brightness of the background image with the highest average brightness is selected as the reference white.

[0027] Light compensation is performed on each of the optical images so that the background image brightness of each of the optical images reaches the value of the reference white, thereby obtaining a plurality of first light compensation images of each of the PBAT degradation film materials to be tested.

[0028] Furthermore, the edge detection is performed on the second light compensation image to obtain the texture information of the PBAT degradation film material in the image, including:

[0029] For each second light compensation image of the PBAT degradation film material to be tested, a PBAT degradation film material image is identified and segmented therefrom.

[0030] The texture of the PBAT degradation film material is obtained by analyzing the gradient threshold of each pixel point in the PBAT degradation film material image.

[0031] A neighborhood analysis is performed on each of the textures, and the pixel points of the textures are corrected according to the analysis results to obtain the texture information of the PBAT degradation film material in the image.

[0032] Furthermore, the analyzing and processing of each of the texture information to obtain crack information therein includes:

[0033] The crack data set of PBAT degradation film is obtained, and the feature extraction of the crack data in the data set is performed to obtain the crack characteristics of the PBAT degradation film material.

[0034] Feature extraction is performed on the texture information of each of the PBAT degradation film materials to be tested to obtain texture features.

[0035] Each of the texture features is compared with the crack features one by one to obtain the crack information in each PBAT degradation film material to be tested.

[0036] Furthermore, after performing light compensation on each of the optical images so that the background image brightness of each of the optical images reaches the value of the reference white, thereby obtaining a plurality of first light compensation images of each of the PBAT degradation film materials to be tested, the method further includes:

[0037] Obtain the mechanical vibration frequency and mechanical vibration amplitude of the PBAT film blowing machine.

[0038] According to the mechanical vibration frequency and the mechanical vibration amplitude, motion calibration is performed on each of the first light compensation images to eliminate mechanical artifacts in each of the first light compensation images.

[0039] Further, the performing motion calibration on each of the first light compensation images according to the mechanical vibration frequency and the mechanical vibration amplitude to eliminate mechanical artifacts in each of the first light compensation images includes:

[0040] Calibration filter parameters are set according to the mechanical vibration frequency and the mechanical vibration amplitude.

[0041] Mechanical artifact filtering is performed on the first light compensation image according to the calibration filter parameters to eliminate mechanical artifacts generated by the PBAT film blowing machine.

[0042] Another embodiment of the present invention provides a PBAT degradation film material defective intelligent detection system, comprising:

[0043] The image acquisition module is used to perform surface acoustic wave image acquisition on each PBAT degradation film material to be tested to obtain the acoustic wave image to be detected, and to perform optical image acquisition on each PBAT degradation film material to be tested to obtain the optical image to be detected.

[0044] The image segmentation module is used to traverse and obtain the pixel matrix of each acoustic wave image to be detected, obtain a plurality of image segmentation centers in the acoustic wave image to be detected by a corner point detection algorithm; and divide each acoustic wave image to be detected into a plurality of areas to be detected according to each of the image segmentation centers.

[0045] The flatness calculation module is used to calculate the regional flatness in each of the to-be-detected regions and the image flatness of the to-be-detected acoustic wave image, and screen out the to-be-detected PBAT degradation film materials corresponding to the to-be-detected images whose regional flatness and image flatness are lower than the standard flatness threshold, so as to obtain the first PBAT degradation film qualified product.

[0046] The light compensation module is used to perform light compensation processing on the optical image of each of the PBAT degradation film materials to be tested from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be tested.

[0047] The crack analysis module is used to screen the first light compensation images at various angles by contrast learning to obtain the second light compensation image corresponding to the light angle with the smallest loss of the contour characteristics of the PBAT degradation film particles; perform edge detection on the second light compensation image to obtain the texture information of the PBAT degradation film material in the image, analyze and process each of the texture information to obtain the crack information therein, screen out the PBAT degradation film materials to be tested corresponding to each crack information, and obtain the second PBAT degradation film qualified product.

[0048] Compared with the prior art, the embodiments of the present invention have the following advantages:

[0049] (1) The surface flatness of the PBAT degradation film is tested by acoustic wave detection to identify defective PBAT products with uneven surfaces. A PBAT simulation model is constructed based on the image. The thickness of the PBAT degradation film material is obtained based on the light reflection results by simulating light irradiation, and defective PBAT degradation films with thickness that does not meet the standard are screened out;

[0050] (2) Cracks in the PBAT degradation film in the image are identified through edge detection, and motion calibration is performed according to the mechanical vibration frequency and amplitude of the film blowing machine to eliminate mechanical artifacts caused by the tiny mechanical vibration of the film blowing machine during image acquisition, thereby improving the accuracy of crack identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A flowchart of the steps of the intelligent detection method for defective PBAT degradation film materials provided in an embodiment of the present invention;

[0052] Figure 2 This is a structural block diagram of the intelligent detection system for defective PBAT degradation membrane materials provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] 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.

[0054] 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.

[0055] 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.

[0056] 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.

[0057] An embodiment of the present invention provides a method for intelligently detecting defective PBAT degradation film materials. For details, see Figure 1 , Figure 1The flowchart of the step of the intelligent detection method of defective PBAT degradation membrane materials in one embodiment of the present invention is shown, including steps S11 to S15:

[0058] Step S11, performing surface acoustic wave image acquisition on each PBAT degradation film material to be tested to obtain an acoustic wave image to be detected, and performing optical image acquisition on each PBAT degradation film material to be tested to obtain an optical image to be detected.

[0059] As a biodegradable and environmentally friendly material, PBAT is increasingly used in the production of plastic wrap, plastic bags, surgical gowns, disposable gloves, etc., thanks to its excellent biodegradability and physical properties. However, due to problems with the film blowing process, defects in film blowing tools, resin raw material problems, or debris in the mold, the produced degradable film materials may have defects such as uneven surface, uneven thickness, and cracks on the film surface, resulting in defective PBAT degradable film materials.

[0060] If defective degradation film materials enter the market, it will cause certain troubles to users. For example, when the degradation film with uneven thickness is subjected to external force, the thinner part is easily damaged, which affects the service life. Similarly, since the toughness of PBAT degradation film materials may be slightly insufficient compared to some traditional plastics, materials with cracks on the surface of the degradation film will also be more likely to break during use. In addition, the uneven surface and uneven thickness may cause the degradation rate of the degradation film in the natural environment to be inconsistent, and some areas will degrade more slowly, prolonging the existence time of plastic waste. Therefore, it is necessary to conduct defect detection on PBAT degradation film materials to eliminate defective products.

[0061] Specifically, it is necessary to first collect images of each PBAT degradation film material to be tested. For surface flatness detection, this embodiment uses acoustic wave detection to perform surface detection on each PBAT degradation film material to be tested. According to the acoustic wave detection results, a surface acoustic wave image of each PBAT degradation film material to be tested is generated as the acoustic wave image to be detected.

[0062] Specifically, the method of obtaining the acoustic wave image to be detected is:

[0063] Using laser surface acoustic wave technology, the laser is focused on the surface of the PBAT degradation film material to be tested, the surface acoustic wave is excited, and several first acoustic wave images are generated based on the reflected acoustic wave signal.

[0064] The first acoustic wave images are de-noised by using a median filter to obtain a second acoustic wave image.

[0065] The second acoustic wave images are registered and fused according to the feature points of the second acoustic wave images to obtain an acoustic wave image to be detected.

[0066] Due to the light transmittance of the PBAT degradation film, this embodiment uses light irradiation to measure its thickness. Therefore, a camera device is set on the film forming side of the film blowing machine to regularly capture and collect optical images of each PBAT degradation film material to be tested as the optical image to be detected according to a preset acquisition frequency.

[0067] Step S12, traverse and obtain the pixel matrix of each acoustic wave image to be detected, obtain several image segmentation centers in the acoustic wave image to be detected by a corner point detection algorithm; and divide each acoustic wave image to be detected into several areas to be detected according to each image segmentation center.

[0068] Each acoustic wave image to be detected needs to be divided into a number of regions to be detected. Specifically, in order to make the region division more helpful for the accuracy of detecting uneven points of the material, the method adopted in this embodiment is:

[0069] Traverse the pixels of each acoustic wave image to be detected to obtain a pixel matrix. Use the corner detection algorithm to determine several corner points by detecting places where the local pixel grayscale values ​​in the matrix change significantly. The determined corner points are used as image segmentation centers.

[0070] Each acoustic wave image to be detected is divided according to each image segmentation center to obtain a number of areas to be detected.

[0071] By determining the partition center through corner point detection, the key feature points in the image can be quickly found. Before further flatness detection, the image is roughly partitioned according to obvious features, which improves the efficiency of subsequent flatness calculation.

[0072] Step S13, calculating the regional flatness of each of the regions to be detected and the image flatness of the acoustic wave image to be detected, screening out the PBAT degradation film materials to be detected corresponding to the images to be detected whose regional flatness and image flatness are lower than the standard flatness threshold, and obtaining the first PBAT degradation film qualified product.

[0073] All the acoustic wave images to be detected and the corresponding pixel images in each area to be detected are grayed, and the gray level co-occurrence matrix is ​​calculated by comparing the gray levels of adjacent pixels in the image.

[0074] Contrast features are extracted from the gray level co-occurrence matrix of each image to be detected and the gray level co-occurrence matrix of each area to be detected, and the flatness of the corresponding matrix is ​​calculated to obtain the image flatness of each acoustic wave image to be detected and the area flatness of each area to be detected.

[0075] The flatness of each image is compared with the standard flatness threshold. If the flatness is lower than the standard flatness threshold, the PBAT degradation film material to be tested corresponding to the image flatness is screened out as a defective product. Then the flatness of each area is compared with the standard flatness threshold. If the flatness is lower than the standard flatness threshold, the PBAT degradation film material to be tested corresponding to the flatness of the area is also screened out as a defective product, thereby obtaining the first PBAT degradation film qualified product.

[0076] Preferably, after obtaining the first qualified PBAT degradation film, the thickness of the PBAT degradation film material to be tested needs to be tested, specifically:

[0077] A light irradiation simulation is performed on each optical image to be detected to obtain a simulated light reflection result of each PBAT degradation film material to be detected; the thickness of the PBAT degradation film material to be detected is calculated according to the simulated light reflection result, and defective PBAT degradation film materials whose thickness exceeds the standard thickness range are screened out from the first PBAT degradation film material.

[0078] The light irradiation simulation for each optical image to be detected can be realized by constructing a model. Specifically, the method adopted in this embodiment is:

[0079] (1) Generate a PBAT degradation film simulation model based on the optical images to be tested.

[0080] (2) Selecting several evenly distributed incident points in the PBAT degradation film simulation model;

[0081] (3) Using incident light from multiple angles to simulate the light irradiation of the optical image to be detected at each incident point, obtaining the light reflection information of each light path, and obtaining the light path reflection information group corresponding to the PBAT degradation film material to be tested according to each light reflection information as the simulated light reflection result.

[0082] Step S14, performing light compensation processing on the optical image of each of the PBAT degradation film materials to be tested from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be tested.

[0083] The first optical image of each PBAT degradation film material to be tested is segmented to obtain a background image excluding the PBAT degradation film material in each light image.

[0084] Traverse the pixels in each background image, calculate the average brightness of each background image, select the background image with the highest average brightness, and use its brightness as the reference white. The reference white is the light intensity reference coefficient for illumination compensation. Usually, the part with the highest pixel brightness in an image is used as the reference value, and other pixels are light compensated according to this value.

[0085] In this embodiment, since multiple images are collected at different angles for the same degradation film material, the method for selecting the reference white is improved, and the average pixel brightness of the brightest image of the background image is used as the reference white, so that the entire image of the other first optical images (including the background and the PBAT material) is light compensated, so that the background image brightness of all the first optical images of each PBAT degradation film material to be tested reaches the value of the reference white, thereby obtaining a plurality of first light compensation images of each PBAT degradation film material to be tested.

[0086] Step S15, using a comparative learning method to screen the first light compensation images at each angle, and obtain a second light compensation image corresponding to the light angle with the smallest loss of the contour characteristics of the PBAT degradation film particles; performing edge detection on the second light compensation image to obtain the texture information of the PBAT degradation film material in the image, analyzing and processing each of the texture information to obtain the crack information therein, and screening out the PBAT degradation film materials to be tested corresponding to each crack information, and obtaining a second PBAT degradation film qualified product.

[0087] In actual optical image acquisition, due to different light angles, the contour features of microscopic particles in the image are lost, so it is necessary to find the image acquired at the best light angle.

[0088] According to the PBAT degradation film data collected under the standard environment, the standard PBAT degradation film particle contour information is obtained, and the standard PBAT degradation film particle contour features are extracted according to the standard PBAT degradation film particle contour information. The particle contour features are extracted for the first light compensation image at each angle to obtain the particle contour features of the PBAT degradation film to be tested, and the absolute mean difference between each particle contour feature of the PBAT degradation film to be tested and the particle contour feature of the standard PBAT degradation film is calculated:

[0089] ;

[0090] in, is the expected value of the random light noise following the normal distribution, is the expected value that the particle profile characteristics of the PBAT degradation film to be tested obey the data distribution, This is the characteristic diagram of the particle profile of the standard PBAT degradation film. It is the contour characteristic diagram of the PBAT degradation film particles to be tested;

[0091] According to the absolute mean difference value, the image corresponding to the light angle with the smallest loss of the contour characteristics of the PBAT degradation film particles is selected as the second light compensation image.

[0092] In order to further improve the detection accuracy of defective products and screen out defective products in PBAT degradation film materials to the greatest extent, it is necessary to perform crack detection on each material. This embodiment uses image segmentation and texture analysis to achieve crack detection, specifically:

[0093] For each second light compensation image of the PBAT degradation film material to be tested, the PBAT degradation film material image is identified and segmented therefrom.

[0094] The texture of the PBAT degradation film material is obtained by analyzing the gradient threshold of each pixel point in the PBAT degradation film material image; specifically, the speed of pixel value change in the PBAT degradation film material image is analyzed from two aspects: gradient intensity and gradient direction, so as to determine the texture in the degradation film material according to the analysis results.

[0095] A neighborhood analysis is performed on each texture, and the pixel points of the texture are corrected according to the analysis results to obtain clearer and more accurate texture features as the texture information of the PBAT degradation film material in the image.

[0096] Among the acquired textures, some are normal textures presented in the image due to the volatility of the material itself, so abnormal crack information should be detected from them.

[0097] According to the historical defective product records, the PBAT degradation film crack data set is obtained, and the crack data in the data set is feature extracted to obtain the crack characteristics of the PBAT degradation film material. The texture information of each PBAT degradation film material to be tested is feature extracted to obtain the texture characteristics.

[0098] Each of the texture features is compared with the crack features one by one, and according to the data similarity between each texture feature and the crack feature, the PBAT degradation film material to be tested with cracks is screened out and excluded as a defective product.

[0099] In addition, since the optical image to be detected is acquired in real time by a camera device arranged on the film forming side of the film blowing machine, in order to further improve the accuracy of defective product detection, it is necessary to eliminate the mechanical artifacts generated by the micro-vibration during the operation of the machine. The specific implementation method is as follows:

[0100] After acquiring the first light compensation image, mechanical operation data such as the mechanical vibration frequency and mechanical vibration amplitude of the PBAT film blowing machine are acquired, and calibration filter parameters are set according to the mechanical operation data.

[0101] Mechanical artifact filtering is performed on each first light compensation image according to the calibration filter parameters, thereby achieving motion calibration and eliminating mechanical artifacts in each first light compensation image caused by internal micro-vibration of the PBAT film blowing machine.

[0102] The intelligent detection method for defective PBAT degradation film materials of the present invention detects the surface flatness of the PBAT degradation film through acoustic wave detection, identifies defective PBAT products with uneven surfaces, constructs a PBAT simulation model according to an image, obtains the thickness of the PBAT degradation film material according to the light reflection result by simulating light irradiation, and screens out defective PBAT degradation films with thickness that does not meet the standard; identifies cracks of the PBAT degradation film in the image through edge detection, and performs motion calibration according to the mechanical vibration frequency and mechanical vibration amplitude of the film blowing machine, thereby eliminating mechanical artifacts generated by the tiny mechanical vibration of the film blowing machine during the image acquisition process, and improving the accuracy of crack identification.

[0103] The embodiment of the present invention also provides a PBAT degradation film material defective intelligent detection system, which is used to execute the PBAT degradation film material defective intelligent detection method as described above. Figure 2 This is a structural block diagram of a PBAT degradation film material defective intelligent detection system according to an embodiment of the present invention, wherein the device comprises:

[0104] The image acquisition module 21 is used to perform surface acoustic wave image acquisition on each PBAT degradation film material to be tested to obtain an acoustic wave image to be detected, and to perform optical image acquisition on each PBAT degradation film material to be tested to obtain an optical image to be detected.

[0105] The image segmentation module 22 is used to traverse and obtain the pixel matrix of each acoustic wave image to be detected, obtain a plurality of image segmentation centers in the acoustic wave image to be detected by a corner point detection algorithm; and divide each acoustic wave image to be detected into a plurality of areas to be detected according to each image segmentation center.

[0106] The flatness calculation module 23 is used to calculate the regional flatness in each of the to-be-detected regions and the image flatness of the to-be-detected acoustic wave image, and screen out the to-be-detected PBAT degradation film materials corresponding to the to-be-detected images whose regional flatness and image flatness are lower than the standard flatness threshold, so as to obtain the first PBAT degradation film qualified product.

[0107] The light compensation module 24 is used to perform light compensation processing on the optical image of each of the PBAT degradation film materials to be tested from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be tested.

[0108] The crack analysis module 25 is used to screen the first light compensation images at various angles by contrast learning to obtain a second light compensation image corresponding to a light angle at which the loss of the contour characteristics of the PBAT degradation film particles is minimal; perform edge detection on the second light compensation image to obtain texture information of the PBAT degradation film material in the image, analyze and process each of the texture information to obtain crack information therein, screen out the PBAT degradation film materials to be tested corresponding to each crack information, and obtain a second PBAT degradation film qualified product.

[0109] The technical features and technical effects of the device proposed in the embodiment of the present invention are the same as those of the method proposed in the embodiment of the present invention, and are not described in detail here. Each module in the above device can be implemented in whole or in part by software, hardware, and a combination thereof. Each of the above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0110] 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 method for intelligently detecting defective PBAT degradation film materials, characterized in that: include: Performing surface acoustic wave image acquisition on each PBAT degradation film material to be tested to obtain an acoustic wave image to be detected, and performing optical image acquisition on each PBAT degradation film material to be tested to obtain an optical image to be detected; Traversing and acquiring a pixel matrix of each of the acoustic wave images to be detected, and acquiring a plurality of image segmentation centers in the acoustic wave images to be detected by a corner point detection algorithm; and dividing each of the acoustic wave images to be detected into a plurality of areas to be detected according to the image segmentation centers; Calculating the regional flatness of each of the to-be-detected regions and the image flatness of the to-be-detected acoustic wave image, and screening out the to-be-detected PBAT degradation film materials corresponding to the to-be-detected images whose regional flatness and image flatness are lower than the standard flatness threshold, to obtain a first PBAT degradation film qualified product; Performing light compensation processing on the optical image to be detected of each of the PBAT degradation film materials to be detected from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be detected; The first light compensation images at various angles are screened by contrast learning to obtain a second light compensation image corresponding to a light angle at which the loss of the contour characteristics of the PBAT degradation film particles is minimal; edge detection is performed on the second light compensation image to obtain texture information of the PBAT degradation film material in the image, and each of the texture information is analyzed and processed to obtain corresponding crack information, and the PBAT degradation film material to be tested corresponding to each crack information is screened out, and the corresponding defective product is determined based on the second PBAT degradation film qualified product; The method of using contrast learning to screen the first light compensation images at various angles to obtain a second light compensation image corresponding to a light angle with the smallest loss of contour features of PBAT degradation film particles includes: Obtain the particle profile information of the standard PBAT degradation film, extract the particle profile features of the standard PBAT degradation film according to the particle profile information of the standard PBAT degradation film, extract the particle profile features of the first light compensation image at each angle, obtain the particle profile features of the PBAT degradation film to be tested, and calculate the absolute mean difference between each particle profile feature of the PBAT degradation film to be tested and the particle profile feature of the standard PBAT degradation film: ； in, is the expected value of the random light noise following the normal distribution, is the expected value that the particle profile characteristics of the PBAT degradation film to be tested obey the data distribution, This is the characteristic diagram of the particle profile of the standard PBAT degradation film. It is the contour characteristic diagram of the PBAT degradation film particles to be tested; According to the absolute mean difference value, the image corresponding to the light angle with the smallest loss of the contour feature of the PBAT degradation film particles is selected as the second light compensation image.

2. The intelligent detection method for defective PBAT degradation film materials according to claim 1, characterized in that: The surface acoustic wave image acquisition of each PBAT degradation film material to be tested to obtain the acoustic wave image to be tested includes: Using laser surface acoustic wave technology, the laser is focused on the surface of the PBAT degradation film material to be tested, the surface acoustic wave is excited, and a number of first acoustic wave images are generated according to the reflected acoustic wave signal; Performing denoising on each of the first acoustic wave images by using a median filter to obtain a second acoustic wave image; The second acoustic wave images are registered and fused according to the feature points of the second acoustic wave images to obtain an acoustic wave image to be detected.

3. The intelligent detection method for defective PBAT degradation film materials according to claim 1, characterized in that: The calculating of the area flatness in each of the to-be-detected areas and the image flatness of the to-be-detected acoustic wave image comprises: Graying the acoustic wave image to be detected and the image corresponding to each of the areas to be detected, and calculating a gray level co-occurrence matrix by comparing the gray levels of adjacent pixels in the image; Contrast features are extracted from each gray level co-occurrence matrix, and the flatness of the corresponding matrix is ​​calculated to obtain the image flatness of the acoustic wave image to be detected and the regional flatness of each region to be detected.

4. The intelligent detection method for defective PBAT degradation film materials according to claim 1, characterized in that: The light compensation processing is performed on the optical image to be detected of each of the PBAT degradation film materials to be detected from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be detected, including: Performing image segmentation on each of the optical images to be detected of the PBAT degradation film material to be detected to obtain a background image in each of the optical images to be detected; Calculate the average brightness of each pixel in each background image, and select the average brightness of the background image with the highest average brightness as the reference white; Light compensation is performed on each of the optical images to be detected, so that the background image brightness of each of the optical images to be detected reaches the value of the reference white, so as to obtain a plurality of first light compensation images of each of the PBAT degradation film materials to be detected.

5. The intelligent detection method for defective PBAT degradation film materials according to claim 4, characterized in that: The performing edge detection on the second light compensation image to obtain texture information of the PBAT degradation film material in the image includes: Identifying the PBAT degradation film material image segmented from each second light compensation image of each of the PBAT degradation film materials to be tested; Obtaining the texture of the PBAT degradation film material by performing gradient threshold analysis on each pixel point in the image of the PBAT degradation film material; A neighborhood analysis is performed on each of the textures, and the pixel points of the textures are corrected according to the analysis results to obtain the texture information of the PBAT degradation film material in the image.

6. The intelligent detection method for defective PBAT degradation film materials according to claim 5, characterized in that: The analyzing and processing of each of the texture information to obtain crack information therein includes: Obtain a PBAT degradation film crack data set, perform feature extraction on the crack data in the data set, and obtain the crack features of the PBAT degradation film material; Extracting features of the texture information of each of the PBAT degradation film materials to be tested to obtain texture features; Each of the texture features is compared with the crack features one by one to obtain the crack information in each PBAT degradation film material to be tested.

7. The intelligent detection method for defective PBAT degradation film materials according to claim 4, characterized in that: After performing light compensation on each of the optical images to be detected so that the background image brightness of each of the optical images to be detected reaches the value of the reference white, so as to obtain a plurality of first light compensation images of each of the PBAT degradation film materials to be detected, the method further includes: Obtain the mechanical vibration frequency and mechanical vibration amplitude of the PBAT film blowing machine; According to the mechanical vibration frequency and the mechanical vibration amplitude, motion calibration is performed on each of the first light compensation images to eliminate mechanical artifacts in each of the first light compensation images.

8. The intelligent detection method for defective PBAT degradation film materials according to claim 7, characterized in that: The performing motion calibration on each of the first light compensation images according to the mechanical vibration frequency and the mechanical vibration amplitude to eliminate mechanical artifacts in each of the first light compensation images includes: Setting calibration filter parameters according to the mechanical vibration frequency and the mechanical vibration amplitude; Mechanical artifact filtering is performed on the first light compensation image according to the calibration filter parameters to eliminate mechanical artifacts generated by the PBAT film blowing machine.

9. An intelligent detection system for defective PBAT degradation film materials, characterized in that: include: An image acquisition module is used to acquire surface acoustic wave images of each PBAT degradation film material to be tested, to obtain an acoustic wave image to be tested, and to acquire optical images of each PBAT degradation film material to be tested, to obtain an optical image to be tested; An image segmentation module is used to traverse and obtain a pixel matrix of each of the acoustic wave images to be detected, and obtain a plurality of image segmentation centers in the acoustic wave images to be detected by a corner point detection algorithm; and divide each of the acoustic wave images to be detected into a plurality of areas to be detected according to each of the image segmentation centers; A flatness calculation module, used to calculate the regional flatness of each of the to-be-detected regions and the image flatness of the to-be-detected acoustic wave image, and to screen out the to-be-detected PBAT degradation film materials corresponding to the to-be-detected images whose regional flatness and image flatness are lower than a standard flatness threshold, to obtain a first PBAT degradation film qualified product; A light compensation module, used for performing light compensation processing on the optical image to be detected of each of the PBAT degradation film materials to be detected from several angles to obtain a first light compensation image of each of the PBAT degradation film materials to be detected; A crack analysis module, for screening the first light compensation images at various angles by contrast learning to obtain a second light compensation image corresponding to a light angle at which the loss of the contour characteristics of the PBAT degradation film particles is minimal; performing edge detection on the second light compensation image to obtain texture information of the PBAT degradation film material in the image, analyzing and processing each of the texture information to obtain corresponding crack information, screening out the PBAT degradation film material to be tested corresponding to each crack information, and determining the corresponding defective product based on the obtained second PBAT degradation film qualified product; The method of using contrast learning to screen the first light compensation images at various angles to obtain a second light compensation image corresponding to a light angle with the smallest loss of contour features of PBAT degradation film particles includes: Obtain the particle profile information of the standard PBAT degradation film, extract the particle profile features of the standard PBAT degradation film according to the particle profile information of the standard PBAT degradation film, extract the particle profile features of the first light compensation image at each angle, obtain the particle profile features of the PBAT degradation film to be tested, and calculate the absolute mean difference between each particle profile feature of the PBAT degradation film to be tested and the particle profile feature of the standard PBAT degradation film: ； in, is the expected value of the random light noise following the normal distribution, is the expected value that the particle profile characteristics of the PBAT degradation film to be tested obey the data distribution, This is the characteristic diagram of the particle profile of the standard PBAT degradation film. It is the contour characteristic diagram of the PBAT degradation film particles to be tested; According to the absolute mean difference value, the image corresponding to the light angle with the smallest loss of the contour feature of the PBAT degradation film particles is selected as the second light compensation image.

Images