Aluminum plastic film quality on-line monitoring system and method based on comparative analysis

Abstract

The invention discloses an aluminum plastic film quality online monitoring system and method based on comparative analysis, and belongs to the technical field of aluminum plastic film detection. The system comprises a data acquisition module, a database, a defect identification module, a three-dimensional model analysis module, an intelligent calculation and evaluation module, an interactive display module and an intelligent monitoring module, the data acquisition module is used for acquiring image data and camera parameter data in an aluminum plastic film quality monitoring process; the database is used for storing image data of different aluminum plastic film defect types; the defect identification module is used for establishing a CNN model and training the CNN model; the three-dimensional model analysis module is used for establishing a three-dimensional space model; the intelligent calculation and evaluation module is used for calculating defect degrees corresponding to different aluminum plastic film defects; calculating the comprehensive quality value of the aluminum plastic film; the interactive display module is used for providing a man-machine interaction platform; and the intelligent monitoring module is used for monitoring the defect degree and the quality evaluation value.

Description

Technical Field

[0001] The invention relates to the technical field of aluminum-plastic film detection, and in particular to an aluminum-plastic film quality online monitoring system and method based on comparative analysis. Background Art

[0002] As a key material for lithium battery packaging, the quality stability and production efficiency of aluminum-plastic film have become the core elements of market competitiveness. Traditional aluminum-plastic film quality monitoring methods often rely on manual sampling and offline testing. This method is not only inefficient, but also difficult to fully cover every link in the production process, resulting in quality problems that are difficult to be discovered and corrected in a timely manner. Against this background, online monitoring technology for aluminum-plastic film quality has emerged. This technology integrates advanced image processing algorithms and intelligent analysis systems to achieve real-time and continuous monitoring of the entire production process of aluminum-plastic film. By comparing and analyzing with preset standards, it can accurately identify surface defects, and timely discover and warn potential quality problems. In the existing technology, the degree of aluminum-plastic film defects is often judged based on the grayscale of the collected image. When the quality of aluminum-plastic film is monitored online, the analysis results of some tiny scratches and voids that are difficult to be qualified are often unable to be verified, resulting in low system accuracy. In addition, the single image data makes it impossible for the system to conduct an effective and comprehensive assessment of the severity of aluminum-plastic film defects, affecting production efficiency and aluminum-plastic film quality control level. Summary of the invention

[0003] The object of the present invention is to provide an aluminum-plastic film quality online monitoring system and method based on comparative analysis to solve the problems raised in the above background technology.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions: an online monitoring method for aluminum-plastic film quality based on comparative analysis, the method comprising the following steps:

[0005] S10, establishing a three-dimensional space model, using an inspection instrument to monitor the quality of the aluminum-plastic film, and collecting image data and camera parameter data during the quality inspection of the aluminum-plastic film; the camera parameters include built-in parameters and external parameters of the CCD camera;

[0006] S20, based on the image data of aluminum-plastic film with defects stored in the database, establish a CNN model and train the CNN model; analyze and process the collected image data, input the processed image data into the trained CNN model, and identify the aluminum-plastic film defects and corresponding defect types in the image data;

[0007] S30, according to the aluminum-plastic film defects in the identified image data and the camera parameter data, based on the triangulation principle, respectively determining the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model, and obtaining the surface functions of different aluminum-plastic film defects in the three-dimensional space model;

[0008] S40, according to the surface functions of different aluminum-plastic film defects in the three-dimensional space model, respectively calculating the defect degrees corresponding to different aluminum-plastic film defects; according to the defect types corresponding to different aluminum-plastic film defects and the influence weights of different defect types on the quality of the aluminum-plastic film, online evaluating the quality of the aluminum-plastic film is performed to calculate the comprehensive quality value of the aluminum-plastic film;

[0009] S50, providing a human-computer interaction platform, and digitally displaying the defect degrees corresponding to the calculated different aluminum-plastic film defects and the quality evaluation values ​​of the aluminum-plastic film; wherein, determining the unqualified thresholds of the defect degrees under different aluminum-plastic film defect types and the unqualified thresholds of the comprehensive quality values ​​of the aluminum-plastic film; when the defect degrees and the comprehensive quality values ​​reach the unqualified thresholds, determining that the quality of the aluminum-plastic film is unqualified, and sending the unqualified information to the management personnel.

[0010] An aluminum-plastic film quality online monitoring system based on comparative analysis, the system comprises a data acquisition module, a database, a defect recognition module, a three-dimensional model analysis module, an intelligent calculation evaluation module, an interactive display module and an intelligent monitoring module;

[0011] The data acquisition module is used to collect image data and camera parameter data during the quality monitoring process of the aluminum-plastic film; send the collected image data to the defect recognition module; and send the collected camera parameter data to the three-dimensional model analysis module;

[0012] The database is used to store image data of different aluminum-plastic film defect types;

[0013] The defect recognition module is used to establish a CNN model based on the image data of aluminum-plastic film with defects stored in the database, and train the CNN model; analyze and process the collected image data, input the processed image data into the trained CNN model, identify the aluminum-plastic film defects in the image data and the corresponding defect types; send the aluminum-plastic film defects in the identified image data to the three-dimensional model analysis module; send the aluminum-plastic film defects in the identified image data and the corresponding defect types to the intelligent computing evaluation module;

[0014] The three-dimensional model analysis module is used to establish a three-dimensional space model, and according to the aluminum-plastic film defects in the identified image data and the camera parameter data, based on the triangulation principle, respectively determine the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model, and obtain the surface functions of different aluminum-plastic film defects in the three-dimensional space model; the surface functions of different aluminum-plastic film defects in the three-dimensional space model are sent to the intelligent calculation evaluation module;

[0015] The intelligent calculation and evaluation module is used to calculate the defect degrees corresponding to different aluminum-plastic film defects according to the surface functions of different aluminum-plastic film defects in the three-dimensional space model; to evaluate the quality of the aluminum-plastic film online according to the defect types corresponding to different aluminum-plastic film defects and the influence weights of different defect types on the quality of the aluminum-plastic film, and to calculate the comprehensive quality value of the aluminum-plastic film; and to send the calculated defect degrees corresponding to the different aluminum-plastic film defects and the quality evaluation values ​​of the aluminum-plastic film to the interactive display module;

[0016] The interactive display module is used to provide a human-computer interaction platform to digitally display the defect degrees corresponding to different calculated aluminum-plastic film defects and the quality assessment values ​​of the aluminum-plastic film;

[0017] The intelligent monitoring module is used to determine the unqualified thresholds of the defect degree and the unqualified thresholds of the comprehensive quality value of the aluminum-plastic film under different aluminum-plastic film defect types; monitor the defect degree and the quality evaluation value of the aluminum-plastic film corresponding to different aluminum-plastic film defects displayed in the interactive display module. When the defect degree and the comprehensive quality value reach the unqualified threshold, the aluminum-plastic film quality is determined to be unqualified, and the unqualified information is sent to the management personnel.

[0018] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: by establishing a three-dimensional space model, based on the principle of triangulation, the surface functions of different aluminum-plastic film defects in the three-dimensional space model are determined respectively, and the defect degrees corresponding to different aluminum-plastic film defects and the comprehensive quality values ​​of the aluminum-plastic film are calculated, thereby more accurately reflecting the current quality problems of the aluminum-plastic film, improving the accuracy of the system's online monitoring and the quality control level of the aluminum-plastic film; by using two CCD cameras to respectively collect image data and perform image comparison and analysis, the fault tolerance of the system is improved, thereby improving the production efficiency of the aluminum-plastic film. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the steps of an online monitoring method for aluminum-plastic film quality based on comparative analysis of the present invention;

[0020] Figure 2 The present invention is a structural schematic diagram of an aluminum-plastic film quality online monitoring system based on comparative analysis. DETAILED DESCRIPTION

[0021] 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. 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.

[0022] See also Figure 1-2 , the present invention provides a technical solution:

[0023] See also Figure 1 In the first embodiment, a method for online monitoring of aluminum-plastic film quality based on comparative analysis is provided. The quality of the aluminum-plastic film is monitored online by an inspection instrument. Two industrial CCD cameras are set up on the left and right sides above the aluminum-plastic film. At the same time, a high-brightness LED linear focusing cold light source is used to backlight the aluminum-plastic film. The CCD camera and a high-speed image processing system are used to perform real-time online scanning to monitor online whether there are scratches and pits on the aluminum-plastic film, replacing the manual visual method, improving the detection efficiency and detecting extremely subtle defects that cannot be observed by the naked eye. The method comprises the following steps:

[0024] S10, establishing a three-dimensional space model, using an inspection instrument to monitor the quality of the aluminum-plastic film, and collecting image data and camera parameter data during the quality monitoring process of the aluminum-plastic film; the camera parameters include built-in parameters and external parameters of the CCD camera.

[0025] Furthermore, the inspection instrument includes two CCD cameras; wherein, the two CCD cameras are respectively mounted above the aluminum-plastic film to collect image data during the quality monitoring process of the aluminum-plastic film, and the two CCD cameras are respectively installed at different positions to form a viewing angle difference; a three-dimensional space model about the X-axis, Y-axis and Z-axis is established, and when the inspection instrument is used to monitor the quality of the aluminum-plastic film, the corresponding camera parameters of the two CCD cameras in the three-dimensional space model are determined respectively; wherein, the built-in parameters include focal length and principal point; and the external parameters include rotation matrix and translation vector.

[0026] It should be noted that the built-in parameters include the focal length and principal point coordinates of the camera in the three-dimensional space model; the external parameters include the rotation matrix and translation vector of the camera in the three-dimensional space model; image data are collected simultaneously by two CCD cameras, the collected image data are compared and analyzed, and input into the CNN model for aluminum-plastic film defect judgment and identification, thereby improving the fault tolerance of the system.

[0027] S20. Based on the image data of aluminum-plastic film with defects stored in the database, a CNN model is established and the CNN model is trained; the collected image data is analyzed and processed, and the processed image data is input into the trained CNN model to identify the aluminum-plastic film defects and the corresponding defect types in the image data.

[0028] Furthermore, the method for training the CNN model is: establishing a database, storing image data of different types of aluminum-plastic film defects in the database, and dividing the image data stored in the database into a training set and a validation set, establishing a CNN model, using the training set to train the CNN model so that the CNN model can identify aluminum-plastic film defects and corresponding defect types in the image data, and using the validation set to evaluate the performance of the trained CNN model, and adjusting the model parameters according to the recognition accuracy of the CNN model.

[0029] It should be noted that the CNN model is trained based on the defective aluminum-plastic film image data stored in the database, and the defective aluminum-plastic film image data in the database is a continuously updated process; in this implementation, when the quality of the aluminum-plastic film is determined to be unqualified, the image data corresponding to the aluminum-plastic film is labeled with the defect type and stored in the database, thereby improving the accuracy of the model.

[0030] S30, according to the aluminum-plastic film defects in the identified image data and the camera parameter data, based on the principle of triangulation, respectively determine the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model, and obtain the surface functions of different aluminum-plastic film defects in the three-dimensional space model.

[0031] Specifically, the method steps are:

[0032] S301, analyzing the collected camera parameter data, determining the internal parameter matrices K1 and K2 of the two CCD cameras according to the built-in parameters of the CCD cameras, and determining the external parameters [R1|P1] and [R2|P2] of the two CCD cameras according to the external parameters of the CCD cameras; wherein R1 and R2 represent the rotation matrices of the two CCD cameras respectively; and P1 and P2 represent the translation vectors of the two CCD cameras respectively;

[0033] S302, according to the aluminum-plastic film defects in the identified image data, respectively determine the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model, and obtain (X ij ,Y ij ,Z ij ), where X ij , Y ij , Z ij Satisfying the condition formula:

[0034]

[0035] Among them, H1 and H2 represent the distance from the principal point of the two CCD cameras to the coordinate point (X ij ,Y ij ,Z ij ) rays; and Respectively represent the inverse of the internal parameter matrix; [X ij ,Y ij ,X ij ,1] indicates (X ij ,Y ij ,Z ij ) is a homogeneous coordinate in a three-dimensional space model; i = 1, 2, ... N; j = 1, 2, ... M i ; N represents the number of defects in the aluminum-plastic film; M i represents the number of characteristic points in the i-th aluminum-plastic film defect;

[0036] It should be noted that the feature points correspond to the pixels in the image data; one pixel represents one feature point; by analyzing the pixel position information of the aluminum-plastic film defects in the image data and based on the triangulation principle, the position information of the pixel in the corresponding three-dimensional space model is determined, so as to analyze the subsequent degree of aluminum-plastic film defects, replacing the inaccuracy of the analysis results based on the image grayscale.

[0037] In this embodiment, the internal parameter matrix K represents the internal characteristics of the CCD camera and is a 3x3 matrix, which is expressed as: Wherein, f1 and f2 represent the focal length of the CCD camera on the X-axis and Y-axis in the three-dimensional space model; C1 and C2 represent the principal point coordinates of the CCD camera in the three-dimensional space model; the external parameter [R|P] represents the position and direction of the CCD camera in the three-dimensional space model, which is a 3x4 matrix; it is expressed as: Among them, R represents the rotation matrix, which is a 3x3 rotation matrix used to rotate the point in the three-dimensional space model coordinate system to the camera coordinate system; P represents the translation vector, which is a 3x1 translation vector used to displace the camera coordinate system point in the three-dimensional space model coordinate system point; through comparative analysis, the corresponding identical feature points in the image data collected by the two CCD cameras are determined respectively, the identical feature points are matched, and converted into spatial coordinates in the same three-dimensional space model; among them, H1 and H2 are the rays from the principal point coordinate point of the two CCD cameras in the three-dimensional space model to the feature point, and the homogeneous coordinates [X ij ,Y ij ,Z ij ,1], so as to obtain the position information of the characteristic points corresponding to each aluminum-plastic film defect in the three-dimensional space model.

[0038] S303, according to (X ij ,Y ij ,Y ij ), determine the characteristic point set corresponding to different aluminum-plastic film defects, and obtain the surface functions F1(X,Y,Z), F2(X,Y,Z), ..., F N (X,Y,Z).

[0039] It should be noted that the surface functions F1(X,Y,Z), F2(X,Y,Z), ..., F N X, Y, and Z in (X, Y, Z) have corresponding definition domains in different aluminum-plastic film defects. By obtaining the feature point sets corresponding to different aluminum-plastic film defects, the surface functions of different aluminum-plastic film defects in the three-dimensional space model are determined respectively according to the obtained feature point sets, so as to further analyze these tiny scratches and voids that are difficult to be qualified, thereby improving the accuracy of system analysis. In this implementation, the shape of aluminum-plastic film defects is determined through surface functions, and the aluminum-plastic film defects and defect types are verified, thereby assisting the CNN model in judging the aluminum-plastic film defects.

[0040] S40. According to the surface functions of different aluminum-plastic film defects in the three-dimensional space model, the defect degrees corresponding to different aluminum-plastic film defects are calculated respectively; according to the defect types corresponding to different aluminum-plastic film defects and the influence weights of different defect types on the quality of the aluminum-plastic film, the quality of the aluminum-plastic film is evaluated online to calculate the comprehensive quality value of the aluminum-plastic film.

[0041] Specifically, the method steps are:

[0042] S401, determine the plane function F(X,Y,Z0) of the aluminum-plastic film in the three-dimensional space model, and according to F1(X,Y,Z), F2(X,Y,Z), ..., F N (X, Y, Z), respectively determine the defect degree corresponding to different aluminum-plastic film defects, according to the calculation formula:

[0043]

[0044] Among them, S i Indicates the defect degree corresponding to the i-th aluminum-plastic film defect; F i (X, Y, ε) represents the surface function of the i-th aluminum-plastic film defect in the three-dimensional space model; ε represents the integral variable of the height in the three-dimensional space model; G i represents the defect area corresponding to the i-th aluminum-plastic film defect; Z0 represents the height of the aluminum-plastic film plane in the three-dimensional space model, and Z0 is a constant;

[0045] In this embodiment, the Z axis on the three-dimensional space model is the horizontal height, the inspection instrument is installed at a fixed position, the aluminum-plastic film moves through the conveyor belt on the inspection instrument, and the conveyor belt is at the same horizontal height. By determining the horizontal height of the conveyor belt and the thickness of the aluminum-plastic film, the height Z0 of the aluminum-plastic film plane in the three-dimensional space model is determined; by calculating the defect degree corresponding to different aluminum-plastic film defects, under the condition that the defect area of ​​the aluminum-plastic film remains unchanged, the calculated The larger it is, the greater the degree of defects in the aluminum-plastic film, that is, the worse the quality.

[0046] S402, respectively determine the weights w1, w2, ...w of the influence on the quality of the aluminum-plastic film under different types of aluminum-plastic film defects L , and according to the defect types corresponding to different aluminum-plastic film defects, the quality of the aluminum-plastic film is evaluated online to obtain the comprehensive quality value K of the aluminum-plastic film:

[0047]

[0048] in, It represents the influence weight of the defect type corresponding to the i-th aluminum-plastic film defect on the quality of the aluminum-plastic film; L represents the number of aluminum-plastic film defect types.

[0049] S50, providing a human-computer interaction platform, and digitally displaying the defect degrees corresponding to the calculated different aluminum-plastic film defects and the quality evaluation values ​​of the aluminum-plastic film; wherein, determining the unqualified thresholds of the defect degrees under different aluminum-plastic film defect types and the unqualified thresholds of the comprehensive quality values ​​of the aluminum-plastic film; when the defect degrees and the comprehensive quality values ​​reach the unqualified thresholds, determining that the quality of the aluminum-plastic film is unqualified, and sending the unqualified information to the management personnel.

[0050] In this embodiment, the unqualified threshold values ​​of the defect levels under different aluminum-plastic film defect types are determined respectively. The unqualified threshold K of the comprehensive value of aluminum-plastic film quality max , when there is S i Or when K reaches the unqualified threshold, the aluminum-plastic film is judged to be unqualified, and the unqualified information is sent to the management personnel, who can view the defect degree corresponding to different aluminum-plastic film defects and the quality assessment value of the aluminum-plastic film through the human-computer interaction platform.

[0051] See also Figure 2 , in this embodiment 2: an online monitoring system for aluminum-plastic film quality based on comparative analysis is provided, the system comprising a data acquisition module, a database, a defect recognition module, a three-dimensional model analysis module, an intelligent calculation and evaluation module, an interactive display module and an intelligent monitoring module;

[0052] The data acquisition module is used to collect image data and camera parameter data during the quality monitoring process of the aluminum-plastic film; send the collected image data to the defect recognition module; and send the collected camera parameter data to the three-dimensional model analysis module;

[0053] The database is used to store image data of different aluminum-plastic film defect types;

[0054] The defect recognition module is used to establish a CNN model based on the image data of aluminum-plastic film with defects stored in the database, and train the CNN model; analyze and process the collected image data, input the processed image data into the trained CNN model, identify the aluminum-plastic film defects in the image data and the corresponding defect types; send the aluminum-plastic film defects in the identified image data to the three-dimensional model analysis module; send the aluminum-plastic film defects in the identified image data and the corresponding defect types to the intelligent computing evaluation module;

[0055] The three-dimensional model analysis module is used to determine the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model based on the principle of triangulation according to the aluminum-plastic film defects in the identified image data and the camera parameter data, and obtain the surface functions of different aluminum-plastic film defects in the three-dimensional space model; send the surface functions of different aluminum-plastic film defects in the three-dimensional space model to the surface functions of different aluminum-plastic film defects in the three-dimensional space model;

[0056] The intelligent calculation and evaluation module is used to calculate the defect degrees corresponding to different aluminum-plastic film defects according to the surface functions of different aluminum-plastic film defects in the three-dimensional space model; to evaluate the quality of the aluminum-plastic film online according to the defect types corresponding to different aluminum-plastic film defects and the influence weights of different defect types on the quality of the aluminum-plastic film, and to calculate the comprehensive quality value of the aluminum-plastic film; and to send the calculated defect degrees corresponding to the different aluminum-plastic film defects and the quality evaluation values ​​of the aluminum-plastic film to the interactive display module;

[0057] The interactive display module is used to provide a human-computer interaction platform to digitally display the defect degrees corresponding to different calculated aluminum-plastic film defects and the quality assessment values ​​of the aluminum-plastic film;

[0058] The intelligent monitoring module is used to determine the unqualified thresholds of the defect degree and the unqualified thresholds of the comprehensive quality value of the aluminum-plastic film under different aluminum-plastic film defect types; monitor the defect degree and the quality evaluation value of the aluminum-plastic film corresponding to different aluminum-plastic film defects displayed in the interactive display module. When the defect degree and the comprehensive quality value reach the unqualified threshold, the aluminum-plastic film quality is determined to be unqualified, and the unqualified information is sent to the management personnel.

[0059] In this embodiment:

[0060] When the degree of defects under the defect type of the aluminum-plastic film reaches the unqualified threshold, the intelligent monitoring module determines that the quality of the aluminum-plastic film is unqualified and sends the unqualified quality information of the aluminum-plastic film to the interactive display module. The interactive display module records the unqualified aluminum-plastic film defect type and displays it digitally for the convenience of management personnel to view.

[0061] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for online monitoring of aluminum-plastic film quality based on comparative analysis, characterized in that: The method comprises the following steps: S10, establishing a three-dimensional space model, using an inspection instrument to monitor the quality of the aluminum-plastic film, and collecting image data and camera parameter data during the quality monitoring process of the aluminum-plastic film; the camera parameters include built-in parameters and external parameters of the CCD camera; S20, based on the image data of aluminum-plastic film with defects stored in the database, establish a CNN model and train the CNN model; analyze and process the collected image data, input the processed image data into the trained CNN model, and identify the aluminum-plastic film defects and corresponding defect types in the image data; S30, according to the aluminum-plastic film defects in the identified image data and the camera parameter data, based on the triangulation principle, respectively determining the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model, and obtaining the surface functions of different aluminum-plastic film defects in the three-dimensional space model; S40, according to the surface functions of different aluminum-plastic film defects in the three-dimensional space model, respectively calculating the defect degrees corresponding to different aluminum-plastic film defects; according to the defect types corresponding to different aluminum-plastic film defects and the influence weights of different defect types on the quality of the aluminum-plastic film, online evaluating the quality of the aluminum-plastic film is performed to calculate the comprehensive quality value of the aluminum-plastic film; S50, providing a human-computer interaction platform, and digitally displaying the defect degrees corresponding to the calculated different aluminum-plastic film defects and the quality evaluation values ​​of the aluminum-plastic film; wherein, determining the unqualified thresholds of the defect degrees under different aluminum-plastic film defect types and the unqualified thresholds of the comprehensive quality values ​​of the aluminum-plastic film; when the defect degrees and the comprehensive quality values ​​reach the unqualified thresholds, determining that the quality of the aluminum-plastic film is unqualified, and sending the unqualified information to the management personnel.

2. The method for online monitoring of aluminum-plastic film quality based on comparative analysis according to claim 1, characterized in that: The inspection instrument comprises two CCD cameras; wherein, the two CCD cameras are respectively mounted above the aluminum-plastic film to collect image data in the process of monitoring the quality of the aluminum-plastic film, and the two CCD cameras are respectively installed at different positions to form a viewing angle difference; a three-dimensional space model about the X-axis, the Y-axis and the Z-axis is established, and when the inspection instrument is used to monitor the quality of the aluminum-plastic film, the corresponding camera parameters of the two CCD cameras in the three-dimensional space model are respectively determined; wherein, the built-in parameters include the camera focal length and the camera principal point; and the external parameters include the camera rotation matrix and the translation vector.

3. The method for online monitoring of aluminum-plastic film quality based on comparative analysis according to claim 1, characterized in that: The method for training the CNN model is as follows: establishing a database, storing image data of different types of aluminum-plastic film defects in the database, dividing the image data stored in the database into a training set and a validation set, establishing a CNN model, using the training set to train the CNN model so that the CNN model can identify aluminum-plastic film defects and corresponding defect types in the image data, and using the validation set to evaluate the performance of the trained CNN model, and adjusting the model parameters according to the recognition accuracy of the CNN model.

4. The method for online monitoring of aluminum-plastic film quality based on comparative analysis according to claim 2, characterized in that: The method steps of step S30 are: S301, analyzing the collected camera parameter data, determining the internal parameter matrices K1 and K2 of the two CCD cameras according to the built-in parameters of the CCD cameras, and determining the external parameters [R1|P1] and [R2|P2] of the two CCD cameras according to the external parameters of the CCD cameras; wherein R1 and R2 represent the rotation matrices of the two CCD cameras respectively; and P1 and P2 represent the translation vectors of the two CCD cameras respectively; S302, according to the aluminum-plastic film defects in the identified image data, respectively determine the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model, and obtain (X ij ,Y ij ,Z ij ), where X ij , Y ij , Z ij Satisfying the condition formula: Among them, H1 and H2 represent the distance from the principal point of the two CCD cameras to the coordinate point (X ij ,Y ij ,Z ij ) rays; and Respectively represent the inverse of the internal parameter matrix; [X ij ,Y ij ,Z ij ,1] indicates (X ij ,Y ij ,Z ij ) is a homogeneous coordinate in a three-dimensional space model; i = 1, 2, ... N; j = 1, 2, ... M i ; N represents the number of defects in the aluminum-plastic film; M i represents the number of characteristic points in the i-th aluminum-plastic film defect; S303, according to (X ij ,Y ij ,Y ij ), determine the characteristic point set corresponding to different aluminum-plastic film defects, and obtain the surface functions F1(X,Y,Z), F2(X,Y,Z), ..., F N (X,Y,Z).

5. The method for online monitoring of aluminum-plastic film quality based on comparative analysis according to claim 4, characterized in that: The method steps of step S40 are: S401, determine the plane function F(X,Y,Z0) of the aluminum-plastic film in the three-dimensional space model, and according to F1(X,Y,Z), F2(X,Y,Z), ..., F N (X, Y, Z), respectively determine the defect degree corresponding to different aluminum-plastic film defects, according to the calculation formula: Among them, S i Indicates the defect degree corresponding to the i-th aluminum-plastic film defect; F i (X, Y, ε) represents the surface function of the i-th aluminum-plastic film defect in the three-dimensional space model; ε represents the integral variable of the height in the three-dimensional space model; G i represents the defect area corresponding to the i-th aluminum-plastic film defect; Z0 represents the height of the aluminum-plastic film plane in the three-dimensional space model, and Z0 is a constant; S402, respectively determine the weights w1, w2, ...w of the influence on the quality of the aluminum-plastic film under different types of aluminum-plastic film defects L , and according to the defect types corresponding to different aluminum-plastic film defects, the quality of the aluminum-plastic film is evaluated online to obtain the comprehensive quality value K of the aluminum-plastic film: in, It represents the influence weight of the defect type corresponding to the i-th aluminum-plastic film defect on the quality of the aluminum-plastic film; L represents the number of aluminum-plastic film defect types.

6. An online monitoring system for aluminum-plastic film quality based on comparative analysis, characterized in that: The system includes a data acquisition module, a database, a defect recognition module, a three-dimensional model analysis module, an intelligent calculation and evaluation module, an interactive display module and an intelligent monitoring module; The data acquisition module is used to collect image data and camera parameter data during the quality monitoring process of the aluminum-plastic film; send the collected image data to the defect recognition module; and send the collected camera parameter data to the three-dimensional model analysis module; The database is used to store image data of different aluminum-plastic film defect types; The defect recognition module is used to establish a CNN model based on the image data of aluminum-plastic film with defects stored in the database, and train the CNN model; analyze and process the collected image data, input the processed image data into the trained CNN model, identify the aluminum-plastic film defects in the image data and the corresponding defect types; send the aluminum-plastic film defects in the identified image data to the three-dimensional model analysis module; send the aluminum-plastic film defects in the identified image data and the corresponding defect types to the intelligent computing evaluation module; The three-dimensional model analysis module is used to establish a three-dimensional space model, and according to the aluminum-plastic film defects in the identified image data and the camera parameter data, based on the triangulation principle, respectively determine the position information of the feature points corresponding to each aluminum-plastic film defect in the three-dimensional space model, and obtain the surface functions of different aluminum-plastic film defects in the three-dimensional space model; the surface functions of different aluminum-plastic film defects in the three-dimensional space model are sent to the intelligent calculation evaluation module; The intelligent calculation and evaluation module is used to calculate the defect degrees corresponding to different aluminum-plastic film defects according to the surface functions of different aluminum-plastic film defects in the three-dimensional space model; According to the defect types corresponding to different aluminum-plastic film defects and the weights of influence on the quality of the aluminum-plastic film under different defect types, the quality of the aluminum-plastic film is evaluated online, and the comprehensive quality value of the aluminum-plastic film is calculated; the defect degrees corresponding to the calculated different aluminum-plastic film defects and the quality evaluation values ​​of the aluminum-plastic film are sent to the interactive display module; The interactive display module is used to provide a human-computer interaction platform to digitally display the defect degrees corresponding to different calculated aluminum-plastic film defects and the quality assessment values ​​of the aluminum-plastic film; The intelligent monitoring module is used to determine the unqualified threshold of the degree of defects under different aluminum-plastic film defect types and the unqualified threshold of the comprehensive value of the aluminum-plastic film quality; The defect degrees corresponding to the different aluminum-plastic film defects displayed in the interactive display module and the quality assessment values ​​of the aluminum-plastic film are monitored. When the defect degree and the comprehensive quality value reach the unqualified threshold, the aluminum-plastic film is judged to be unqualified and the unqualified information is sent to the management personnel.

Images