Defect detection method, device and storage medium for plastic film

Abstract

The present application relates to the technical field of image processing, and particularly relates to a defect detection method and device for plastic film and a storage medium, the method comprising: collecting a film image and synchronously collecting sensor data, establishing a vibration distortion transfer model according to the film image and the film vibration data, and correcting the film image for distortion by using the vibration distortion transfer model, collecting an infrared thermal imaging image of the plastic film, and extracting a film temperature gradient, then establishing a thermal stress offset coefficient according to the film temperature gradient, and determining distortion features by combining the thermal stress offset coefficient with a film coating defect state, collecting a polarized light image of the plastic film, and determining a stress state of a film substrate by using the polarized light image, enhancing the distortion features of the plastic film according to the stress state of the film substrate, and sending the enhanced distortion features to a user. The present application effectively improves the detection accuracy of the plastic film.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to a defect detection method and device for plastic film and a storage medium. BACKGROUND

[0002] Current plastic film defect detection mainly relies on visible light image analysis to identify surface defects such as scratches and bubbles through surface texture or grayscale changes. However, functional composite films used in the aerospace field need to meet stringent mechanical and thermal stability requirements, and traditional methods have significant limitations: first, mechanical vibration of the film in high-speed production lines can cause image distortion, and existing technologies lack effective vibration distortion correction models; second, deep defects such as substrate internal stress and thermal shrinkage distortion during the cooling stage cannot be captured through surface images, leading to missed detection of hidden defects and subsequent delamination failure of the film in extreme environments.

[0003] To address the above problems, although some research has attempted to introduce infrared thermal imaging or polarized light detection, multi-modal data has not yet formed a collaborative analysis system: temperature gradient analysis of infrared images has not been quantitatively associated with historical qualified data, making it difficult to distinguish between process-induced thermal distortion and material defects; the results of polarized light stress detection are independently output and not integrated with coating defects and thermal stress states, making it impossible to diagnose the root cause of the defects. In addition, existing methods lack sufficient compensation for vibration interference and lack a closed-loop feedback mechanism for process optimization, making it difficult to directly guide production line parameter adjustments based on the detection results. Therefore, there is an urgent need for a multi-modal collaborative solution that integrates vibration correction, thermal stress analysis, and substrate stress detection to ensure anti-interference capability while achieving accurate positioning and classification of defects in all dimensions from the surface to the deep layer. SUMMARY

[0004] The present application aims to provide a defect detection method and device for plastic film and a storage medium to solve at least one of the problems in the prior art.

[0005] To achieve the above-mentioned purpose, according to one aspect of the present application, a defect detection method for plastic film is provided, comprising:

[0006] Collecting film images and synchronously collecting sensor data;

[0007] Establishing a vibration distortion transfer model based on the film images and film vibration data, and correcting the distortion of the film images using the vibration distortion transfer model;

[0008] Collecting infrared thermal imaging images of the plastic film, extracting the film temperature gradient, and then establishing a thermal stress offset coefficient based on the film temperature gradient, and determining the distortion feature based on the thermal stress offset coefficient and the film coating defect state;

[0009] Collecting a polarized light image of the plastic film, and determining a stress state of the film substrate by the polarized light image;

[0010] Enhancing a distortion feature of the plastic film according to the stress state of the film substrate, and sending the enhanced distortion feature to a user.

[0011] Optionally, coupling the film amplitude, the vibration frequency, the vibration direction and the film moving direction to construct a film vibration offset vector α;

[0012] Establishing a vibration distortion transmission model according to the vibration offset vector α, and correcting the film image: correcting the position of each pixel point in the film image according to the vibration offset vector α, and reducing the film image by |α| times in the direction of the vibration offset vector α.

[0013] Optionally, performing regional gray value fluctuation analysis on the film image after distortion correction to determine the defect state of the film coating, and the process is as follows: dividing the film image after distortion correction into regions by a rectangle with the film image reduced by 20 times in proportion to obtain each film sub-image, calculating the coefficient of variation of the gray value of each pixel point in each film sub-image, and then comparing the coefficient of variation of the gray value of each pixel point with the set threshold interval, dividing the film sub-image belonging to the threshold interval into a defect sub-image; dividing the film sub-image not belonging to the threshold interval into a non-defect sub-image.

[0014] Optionally, obtaining an infrared thermal imaging image of a historical qualified film, and extracting the film length corresponding to each preset temperature in the infrared thermal imaging image of the historical qualified film, and storing it as a temperature-film length table;

[0015] Dividing the infrared thermal imaging image into intervals according to the temperature-film length table to obtain each temperature interval image, and calculating the offset coefficient w(i) of each temperature interval image, and setting w(i) = [BL(i)-L(i)] / BL(i);

[0016] In the formula, i is a digital subscript representing the i-th preset temperature, BL(i) is the film length corresponding to the i-th preset temperature in the temperature-film length table, and L(i) is the film length corresponding to the i-th preset temperature in the infrared thermal imaging image.

[0017] Optionally, determining the film thermal fluctuation state of each temperature interval image, and the film thermal fluctuation state includes a normal state and an abnormal state;

[0018] In the case of normal thin film thermal fluctuation state, the offset coefficient of the i-th temperature interval image is used as the thermal stress offset coefficient of the temperature interval image, and in the case of abnormal thin film thermal fluctuation state, w(i) x exp{CV(i)} is used as the thermal stress offset coefficient of the temperature interval image, wherein CV(i) represents the coefficient of variation of the temperature of the i-th temperature interval image.

[0019] Optionally, the temperature interval image with a thermal stress offset coefficient greater than or equal to the thermal stress offset threshold is marked as a distortion region image, and the distortion region image is matched with the defect sub-image: if there is a defect sub-image belonging to the k-th distortion region image, it is determined that the distortion feature within the k-th distortion region image is a defect sub-image within the distortion region image, and the defect sub-image within the distortion region image is marked as a coating defect region; if there is no defect sub-image belonging to the distortion region image, it is determined that the distortion feature within the k-th distortion region image is the distortion region image, and the k-th distortion region image is marked as a material thermal distortion region.

[0020] The temperature interval image with a thermal stress offset coefficient less than the thermal stress offset threshold is marked as a non-distortion region image.

[0021] Optionally, based on the polarized light image, the reflection light distortion rate of the polarized light image is collected, and the polarized light image is divided into polarized light sub-images to obtain polarized light sub-images.

[0022] According to the reflection light distortion rate, the stress state of the film substrate is determined: in the case that the reflection light distortion rate of the j-th polarized light sub-image is greater than 5%, the polarized light sub-image is determined to be a substrate deformation stress region; in the case that the reflection light distortion rate of the j-th polarized light sub-image is less than or equal to 5%, the polarized light sub-image is determined to be a substrate uniform stress region.

[0023] Optionally, the thermal stress offset coefficient of the temperature interval image containing the substrate deformation stress region is updated to ; wherein B(i, j) represents the reflection light distortion rate of the j-th substrate deformation stress region contained in the i-th temperature interval image, and the distortion feature is sent to the user.

[0024] According to another aspect of the present application, a defect detection device for plastic film is provided, comprising:

[0025] A data acquisition module is configured to acquire a film image and synchronously acquire sensor data.

[0026] A distortion correction module is configured to establish a vibration distortion transfer model according to the film image and the film vibration data, and correct the distortion of the film image based on the vibration distortion transfer model.

[0027] The feature analysis module is configured to collect an infrared thermal imaging image of the plastic film, extract a temperature gradient of the film, and then establish a thermal stress offset coefficient based on the temperature gradient of the film, and determine the distortion feature based on the thermal stress offset coefficient and a defect state of a coating layer of the film.

[0028] The stress detection module is configured to collect a polarized light image of the plastic film, and determine a stress state of a base material of the film based on the polarized light image.

[0029] The feature enhancement module is configured to enhance the distortion feature of the plastic film based on the stress state of the base material of the film, and send the enhanced distortion feature to a user.

[0030] According to still another aspect of the present application, a computer readable storage medium is provided, which stores a computer program. The computer program is configured to control an electronic device in which the computer readable storage medium is located to execute the defect detection method for a plastic film when the computer program is run.

[0031] Compared with the prior art, the present application has the following beneficial effects: the present application significantly improves the comprehensive performance of defect detection of functional composite film through multi-modal collaborative detection and closed-loop optimization mechanism. The innovation lies in that the production environment vibration interference is overcome synchronously, the accurate positioning of surface and deep defects is realized, and a direct feedback link from detection results to process control is established. For the mechanical vibration problem in the high-speed production line, a special correction model is constructed to eliminate image acquisition distortion; infrared thermal imaging and polarized light technology are fused to synchronously capture coating surface defects, thermal stress distortion and base material stress distribution, breaking through the recognition bottleneck of hidden defects in traditional methods. Especially through thermodynamic analysis and stress state mapping, process deformation and material failure risk are accurately distinguished. Finally, based on the defect feature enhancement and classification results, production line parameter adjustment instructions are automatically generated, forming a closed-loop control from defect diagnosis to process improvement, which greatly reduces the comprehensive defect rate of high-end film applications in the field of aerospace. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0033] Figure 1 The flowchart of the defect detection method for the plastic film of the present embodiment.

[0034] Figure 2 The flowchart of the film image distortion correction method of the present embodiment.

[0035] Figure 3 A flowchart of the distortion feature analysis method of the present embodiment is shown in the following.

[0036] Figure 4 A structural diagram of the defect detection device for plastic film provided in the present embodiment is shown in the following. DETAILED DESCRIPTION

[0037] In order to more clearly illustrate the present application, the present application is further described below in conjunction with preferred embodiments and the accompanying drawings. Similar components are denoted by the same reference numerals in the drawings. It should be understood by those skilled in the art that the following specific description is illustrative rather than limiting, and should not limit the scope of protection of the present application.

[0038] It should be noted that although the terms first, second, third, etc. may be used in the embodiments of the present application, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of the present application, the first can also be referred to as the second, and similarly, the second can also be referred to as the first.

[0039] The acquisition, storage, use, processing, etc. of data in the technical solutions of the present application all comply with the relevant provisions of national laws and regulations.

[0040] Specifically, the defect detection method for plastic film described in the present application is applied to the defect detection of functional composite film, wherein the functional composite film is specifically a plastic film produced by a metal plating layer in a vacuum plating production line in the field of aerospace; in the present application, the defect detection is performed in combination with the analysis process of vibration interference and plating cavity thermal distortion.

[0041] In the above application scenarios, the present application provides a defect detection method for plastic film, and a flowchart thereof can be referred to Figure 1 as shown in the following, which specifically comprises:

[0042] Step S101, collecting a film image and synchronously collecting sensor data; the film image is an RGB image of a plastic film; the sensor data comprises film vibration data, and the film vibration data is a micro-vibration amplitude of the film, a moving direction of the film and a vibration direction.

[0043] It should be noted that the collection process of the film image and the sensor data in the present application is simultaneous collection, so as to correct the slight error in the film image collection process by the sensor data; in the present application, the film image is by default an image taken after accurate focusing.

[0044] In the present application, all kinds of images are collected simultaneously, and the collection area, the collection point and the size proportion of the film in the image are the same.

[0045] Please continue to refer to Figure 1 As shown in the method for defect detection of plastic film further comprises:

[0046] Step S102, according to the film image and the film vibration data to establish vibration distortion transfer model, and with vibration distortion transfer model to film image distortion correction.

[0047] Specifically, in order to realize the distortion correction of the film image in step S102, the present application further provides a film image distortion correction method as shown in the method for defect detection of plastic film further comprises: Figure 2 As shown in the method for defect detection of plastic film further comprises:

[0048] Step S201, based on the film image and the film vibration data to determine the film vibration offset vector, and with the film vibration offset vector to establish vibration distortion transfer model.

[0049] Specifically, in the step S201, the establishment process of vibration distortion transfer model is as follows:

[0050] The film amplitude, vibration direction and film moving direction are coupled to construct the film vibration offset vector α, and α=(zf+yf)×|zf+yf|×P is set;

[0051] In the formula, zf is the unit vector of the vibration direction, yf is the unit vector of the film moving direction, and P is the film amplitude;

[0052] The vibration distortion transfer model is established with the vibration offset vector α, and the film image is corrected: according to the vibration offset vector α, the position of each pixel point in the film image is corrected, and the film image is reduced by |α| times along the direction of the vibration offset vector α.

[0053] Specifically, the innovation integrates multiple parameters into vector α and realizes millisecond-level dynamic correction, which guarantees the real-time demand of online detection.

[0054] Please continue to refer to Figure 2 As shown in the method for defect detection of plastic film further comprises:

[0055] Step S202, according to the film image after distortion correction to determine the film coating defect state.

[0056] Specifically, in the step S202, the determination process of the film coating defect state is as follows:

[0057] The region gray value fluctuation of the distortion-corrected film image is analyzed to determine the defect state of the film coating, and the process is as follows: the distortion-corrected film image is divided into film sub-images by a rectangular with a scale of 20 times smaller than the film image, and the coefficient of variation of the gray value of each pixel point in each film sub-image is calculated, and then the coefficient of variation of the gray value of each pixel point is compared with the set threshold interval, and the film sub-image belonging to the threshold interval is divided into a defect sub-image; the film sub-image not belonging to the threshold interval is divided into a non-external defect sub-image.

[0058] For example, the setting process of the threshold interval is not specifically limited in the present application, and those skilled in the art can set it according to the statistical results of the collected data. In the present application, the threshold interval can be set to [0.17, 1].

[0059] Specifically, the vibration offset vector α is constructed by coupling the film amplitude, vibration frequency, vibration direction and moving direction, and the image pixel position is dynamically corrected based on the exposure time, effectively solving the image blur and geometric distortion problems caused by the production line mechanical vibration, providing real and reliable surface topography data for subsequent gray scale analysis, and the model is compatible with multiple types of production line vibration spectrum.

[0060] Please continue to refer to Figure 1 As shown, the defect detection method for plastic film further comprises:

[0061] Step S103, collecting the infrared thermal imaging image of the plastic film, and extracting the film temperature gradient, and then establishing the thermal stress offset coefficient based on the film temperature gradient, and determining the distortion feature based on the thermal stress offset coefficient and the defect state of the film coating.

[0062] In order to realize the determination of the distortion feature in step S103 in the present application, the present application further provides a distortion feature analysis method as shown in Figure 3 As shown, the distortion feature analysis method comprises:

[0063] Step S301, collecting the infrared thermal imaging image of the plastic film; the collection of the infrared thermal imaging image in the present application can be collected by thermal imaging of the plastic film by an infrared thermal imager.

[0064] Please continue to refer to Figure 3 As shown, the distortion feature analysis method further comprises:

[0065] Step S302, extracting the film temperature gradient, and establishing the thermal stress offset coefficient of the infrared thermal imaging image based on the film temperature gradient as the judgment standard.

[0066] Specifically, in the step S302, the thermal stress offset coefficient of the infrared thermal imaging image is established as follows:

[0067] acquire the infrared thermal imaging image of the historical qualified film, and extract the film length corresponding to each preset temperature in the infrared thermal imaging image of the historical qualified film, and store it as a temperature-film length table;

[0068] divide the infrared thermal imaging image by the temperature-film length table to obtain each temperature interval image, and calculate the offset coefficient w(i) of each temperature interval image, and set w(i) = [BL(i)-L(i)] / BL(i);

[0069] In the formula, i is a digital subscript representing the i-th preset temperature, BL(i) is the film length corresponding to the i-th preset temperature in the temperature-film length table, and L(i) is the film length corresponding to the i-th preset temperature in the infrared thermal imaging image.

[0070] Determine the film thermal fluctuation state of each temperature interval image in the above-mentioned manner of determining the film coating defect state of the film image, and the film thermal fluctuation state includes a normal state and an abnormal state.

[0071] When the film thermal fluctuation state is a normal state, the offset coefficient of the i-th temperature interval image is used as the thermal stress offset coefficient of the temperature interval image, and when the film thermal fluctuation state is an abnormal state, w(i) x exp{CV(i)} is used as the thermal stress offset coefficient of the temperature interval image, wherein CV(i) represents the coefficient of variation of the temperature of the i-th temperature interval image.

[0072] For example, the specific process of "determining the film thermal fluctuation state of each temperature interval image in the above-mentioned manner of determining the film coating defect state of the film image, and the film thermal fluctuation state includes a normal state and an abnormal state" in this application is to calculate the coefficient of variation of the temperature of each coordinate point in each temperature interval image, and to judge the film thermal fluctuation state by a threshold interval, the process is the same as the above-mentioned judgment process of the film coating defect state, which will not be described herein.

[0073] Specifically, the value of "each preset temperature" in the part "and extract the film length corresponding to each preset temperature in the infrared thermal imaging image of the historical qualified film" in this application is not specifically set in this application, and those skilled in the art can freely set it, as long as it meets the requirements that each set temperature is an arithmetic sequence and belongs to the normal temperature range of the film.

[0074] It is worth noting that since the defect detection of the plastic film in this application is real-time detection of the processed plastic film, the plastic film is in the cooling process during the detection process of the plastic film, therefore, the acquisition of the infrared thermal imaging image of the plastic film can reflect the cooling process of the plastic film, and further analyze the distortion characteristics of the plastic film.

[0075] Specifically, the offset coefficient is generated based on historical data, and the calculation logic is adaptively adjusted based on the thin film thermal fluctuation state to convert the abstract temperature gradient into a quantifiable parameter, thereby providing core data support for distortion classification.

[0076] Please continue to refer to Figure 3 As shown in the figure, the distortion feature analysis method further includes:

[0077] Step S303, according to the thermal stress offset coefficient of the infrared thermal imaging image and the defect state of the thin film coating layer, the distortion feature is determined.

[0078] Specifically, in the step S303, the determination process of the distortion feature is as follows:

[0079] The temperature interval image of the thermal stress offset coefficient greater than or equal to the thermal stress offset threshold value is marked as a distortion area image, and the distortion area image is matched with the defect sub-image: if there is a defect sub-image belonging to the kth distortion area image, it is determined that the distortion feature in the kth distortion area image is the defect sub-image in the distortion area image, and the defect sub-image in the distortion area image is marked as a coating defect area; if there is no defect sub-image belonging to the distortion area image, it is determined that the distortion feature in the kth distortion area image is the distortion area image, and the kth distortion area image is marked as a material thermal distortion area.

[0080] The temperature interval image of the thermal stress offset coefficient less than the thermal stress offset threshold value is marked as a non-distortion area image, and no distortion feature is set.

[0081] For example, the value of the thermal stress offset threshold value is not specifically limited in this application, and those skilled in the art can freely set it as long as it meets the value requirement of the thermal stress offset threshold value. The best value of the thermal stress offset threshold value in this application is set to 1.1.

[0082] Specifically, the thermal stress offset coefficient and the coating defect state are analyzed together: when the thermal stress offset coefficient is greater than or equal to the threshold value, it is marked as a distortion area image, and matched with the defect sub-image: the overlapping area is determined as "coating defect", and only the thermal stress abnormal area is determined as "material thermal distortion". The double verification mechanism clearly identifies the defect source and specifically guides the optimization of the vacuum coating process parameters to reduce the scrap rate.

[0083] Please continue to refer to Figure 1 As shown in the figure, the defect detection method for plastic film further includes:

[0084] Step S104, acquiring a polarized light image of the plastic film, and determining the stress state of the film substrate based on the polarized light image. The polarized light image of the plastic film in this application is an image obtained by shooting the plastic film with a polarized camera.

[0085] Specifically, in the step S104, the determination process of the stress state of the film substrate is as follows:

[0086] Based on the polarized light image, the reflection light distortion rate of the polarized light image is collected, and the polarized light image is divided by each film sub-image to obtain a polarized light sub-image;

[0087] According to the reflection light distortion rate, the stress state of the film substrate is determined: when the reflection light distortion rate of the jth polarized light sub-image is greater than 5%, it is determined that the polarized light sub-image is a substrate deformation stress area; when the reflection light distortion rate of the jth polarized light sub-image is less than or equal to 5%, it is determined that the polarized light sub-image is a substrate uniform stress area.

[0088] Specifically, the acquisition method of the "reflection light distortion rate" in the present application is: the pixel points of the polarized light image are set to the pixel points of the film image, and the reflection light angle of the pixel points of the polarized light image is detected, the reflection light angle of the pixel points of the polarized light image is compared with the incident angle of the polarized light image, and the pixel points of the polarized light image whose offset proportion of the reflection light angle of the pixel points of the polarized light image and the incident angle of the polarized light image is greater than 50% are taken as the distortion pixel points, and finally the proportion of the distortion pixel points in the polarized light image is taken as the reflection light distortion rate.

[0089] Specifically, the reflection light distortion rate of the polarized light image is used to analyze the internal stress distribution of the film, and the stress state and the physical position are accurately mapped in units of sub-images, the apparent limitation of traditional detection is broken through, and the risk of hidden failure such as delamination and cracking caused by micro-strain of the substrate is warned in advance.

[0090] Please continue to refer to Figure 1 As shown in the figure, the defect detection method for the plastic film further comprises:

[0091] In step S105, the distortion characteristics of the plastic film are enhanced according to the stress state of the film substrate, and the enhanced distortion characteristics are sent to the user.

[0092] Specifically, in the step S105, the process of enhancing the distortion characteristics according to the stress state of the film substrate is as follows:

[0093] The thermal stress offset coefficient of the temperature interval image containing the substrate deformation stress area is updated to ; wherein B(i,j) represents the reflection light distortion rate of the jth substrate deformation stress area contained in the ith temperature interval image, and the distortion characteristics are sent to the user.

[0094] Specifically, the distortion features are dynamically enhanced based on the stress state of the substrate: stress concentration signals are strengthened in the film defect area, and temperature-stress coupling features are amplified in the material thermal distortion area; finally, the enhanced results are pushed to the user end in real time, directly related to the process adjustment of the production line, realizing closed-loop control from detection to optimization.

[0095] Referring to Figure 4 As shown in the figure, the defect detection device for plastic film provided by the present application comprises:

[0096] The data acquisition module is configured to acquire the film image and synchronously acquire sensor data.

[0097] The distortion correction module is configured to establish a vibration distortion transmission model according to the film image and the film vibration data, and correct the distortion of the film image based on the vibration distortion transmission model.

[0098] The feature analysis module is configured to acquire an infrared thermal imaging image of the plastic film, extract a temperature gradient of the film, establish a thermal stress offset coefficient based on the temperature gradient of the film, and determine the distortion features based on the thermal stress offset coefficient and the film coating defect state.

[0099] The stress detection module is configured to acquire a polarized light image of the plastic film, and determine the stress state of the film substrate based on the polarized light image.

[0100] The feature enhancement module is configured to enhance the distortion features of the plastic film based on the stress state of the film substrate, and send the enhanced distortion features to the user.

[0101] The defect detection device for plastic film provided by the present application can execute the defect detection method for plastic film provided by any embodiment of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0102] The present application also provides a computer readable storage medium, which is a tangible physical storage medium that can store the above computer program and various types of data used in the program; the physical storage medium includes but is not limited to random access memory, read-only memory, optical disc, hard disk, etc. existing physical storage medium or combination of media.

[0103] As will be appreciated by one of ordinary skill in the art, all or some of the steps, systems, etc. in the above-disclosed methods can be embodied in software, firmware, hardware, and any suitable combination thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application- specific integrated circuit. Such software can be distributed on computer readable media, which can comprise computer storage media (or non-transitory media), and communication media (or transitory media). As will be appreciated by one of ordinary skill in the art, the term computer storage media includes all tangible and non-tangible, removable and non-removable media implemented in any method or technology for storage of information such as computer readable programs, data structures, program modules or other data. Further, as will be appreciated by one of ordinary skill in the art, communication media typically embodies computer readable programs, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.

[0104] So far, the technical solutions of the present application have been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the present application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the present application, and the technical solutions after these changes or replacements will all fall within the protection scope of the present application.

Claims

1. A method for defect detection of a plastic film, characterized by, The method comprises the following steps: Collecting a film image and synchronously collecting sensor data; Establishing a vibration distortion transmission model according to the film image and the film vibration data, and correcting the film image by using the vibration distortion transmission model, and determining a film coating defect state according to the corrected film image; Collecting an infrared thermal imaging image of the plastic film, extracting a film temperature gradient, and then establishing a thermal stress offset coefficient according to the film temperature gradient, and determining a distortion feature according to the thermal stress offset coefficient and the film coating defect state; the distortion feature is an image data type, which is used to determine a material thermal distortion region and a film defect region in the plastic film according to the thermal stress offset coefficient and the film coating defect state; Collecting a polarized light image of the plastic film, and determining a stress state of the film substrate according to the polarized light image; Enhancing the thermal stress offset coefficient of the plastic film according to the stress state of the film substrate, and sending the enhanced thermal stress offset coefficient to a user.

2. The method for defect detection of a plastic film according to claim 1, wherein Coupling the film amplitude, the vibration frequency, the vibration direction and the film moving direction to construct a film vibration offset vector α; Establishing a vibration distortion transmission model according to the vibration offset vector α, and correcting the film image: correcting the positions of each pixel point in the film image according to the vibration offset vector α, and reducing the film image by |α| times in the direction of the vibration offset vector α.

3. The method for defect detection of a plastic film according to claim 2, wherein Performing regional gray value fluctuation analysis on the corrected film image to determine the defect state of the film coating, and the process is as follows: dividing the corrected film image into regions by using a rectangle which is 20 times smaller than the film image to obtain film sub-images, calculating the variation coefficients of the gray values of each pixel point in the film sub-images, and then comparing the variation coefficients of the gray values of each pixel point with a set threshold interval, and dividing the film sub-images belonging to the threshold interval into defect sub-images; dividing the film sub-images not belonging to the threshold interval into non-external defect sub-images.

4. The method for defect detection of a plastic film according to claim 3, wherein Obtaining an infrared thermal imaging image of a historical qualified film, extracting the film length corresponding to each preset temperature in the infrared thermal imaging image of the historical qualified film, and storing the film length as a temperature-film length table; Dividing the infrared thermal imaging image into intervals according to the temperature-film length table to obtain temperature interval images, and calculating the offset coefficient w(i) of each temperature interval image, wherein w(i)=[BL(i)-L(i)] / BL(i); In the formula, i is a digital index representing the i-th preset temperature, BL(i) is the film length corresponding to the i-th preset temperature in the temperature-film length table, and L(i) is the film length corresponding to the i-th preset temperature in the infrared thermal imaging image.

5. The method for defect detection of a plastic film according to claim 4, wherein Determining the film thermal fluctuation state of each temperature interval image, wherein the film thermal fluctuation state includes a normal state and an abnormal state; When the film thermal fluctuation state is the normal state, the offset coefficient of the i-th temperature interval image is used as the thermal stress offset coefficient of the temperature interval image, and when the film thermal fluctuation state is the abnormal state, w(i)×exp{CV(i)} is used as the thermal stress offset coefficient of the temperature interval image, wherein CV(i) represents the variation coefficient of the temperature of the i-th temperature interval image.

6. The method for defect detection of a plastic film according to claim 5, wherein marking the temperature interval image of the thermal stress offset coefficient greater than or equal to the thermal stress offset threshold as a distortion area image, and matching the distortion area image with the defect sub-image: if there is a defect sub-image belonging to the kth distortion area image, determining the distortion feature in the kth distortion area image as the defect sub-image in the distortion area image, and marking the defect sub-image in the distortion area image as a coating defect area; if there is no defect sub-image belonging to the distortion area image, determining the distortion feature in the kth distortion area image as the distortion area image, and marking the kth distortion area image as a material thermal distortion area; marking the temperature interval image of the thermal stress offset coefficient less than the thermal stress offset threshold as a non-distortion area image.

7. The method for defect detection of a plastic film according to claim 6, wherein Based on the polarized light image, the reflection light distortion rate of the polarized light image is collected, and each thin film sub-image is divided to obtain a polarized light sub-image; According to the reflection light distortion rate, the stress state of the thin film substrate is determined: when the reflection light distortion rate of the jth polarized light sub-image is greater than 5%, the polarized light sub-image is determined as a substrate deformation stress area; when the reflection light distortion rate of the jth polarized light sub-image is less than or equal to 5%, the polarized light sub-image is determined as a substrate uniform stress area.

8. The method for defect detection of a plastic film according to claim 7, wherein updating the thermal stress offset coefficient of the temperature interval image including the substrate deformation stress region to ; wherein B(i,j) represents the reflected light distortion rate of the jth substrate deformation stress region included in the ith temperature interval image, and the distortion feature is sent to the user.

9. An apparatus for defect detection of a plastic film, applied to the method for defect detection of a plastic film according to any one of claims 1 to 8, characterized in that, Comprise: The data acquisition module is used for collecting the film image and synchronously collecting the sensor data; The distortion correction module is used for establishing a vibration distortion transmission model according to the film image and the film vibration data, and correcting the distortion of the film image by the vibration distortion transmission model; The feature analysis module is used for collecting the infrared thermal imaging image of the plastic film, extracting the film temperature gradient, and then establishing the thermal stress offset coefficient according to the film temperature gradient, and determining the distortion feature by combining the thermal stress offset coefficient with the film coating defect state; The stress detection module is used for collecting the polarized light image of the plastic film, and determining the stress state of the film substrate according to the polarized light image; The feature enhancement module is used for enhancing the thermal stress offset coefficient of the plastic film according to the stress state of the film substrate, and sending the enhanced thermal stress offset coefficient to the user.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, wherein the computer program is used to control the electronic device where the computer readable storage medium is located to execute the defect detection method for the plastic film in any one of claims 1-8 when running. The computer readable storage medium stores a computer program, wherein the computer program is used to control the electronic device where the computer readable storage medium is located to execute the defect detection method for the plastic film in any one of claims 1-8 when running.

Images