Intelligent developing active packaging film thickness uniformity detection method and system

Abstract

The invention discloses an intelligent developing active packaging film thickness uniformity detection method and system, and relates to the field of film thickness uniformity detection.The intelligent developing active packaging film thickness uniformity detection method comprises the steps that a terahertz time-domain spectrum detection device is arranged, and area scanning is conducted on a to-be-detected developing active packaging film; network edge nodes are set, and data acquisition, data preprocessing and film thickness uniformity judgment are processed step by step; based on a convolutional neural network algorithm, preliminarily confirming the film thickness of the developing active packaging film; environment data are collected, and the film thickness of the color developing active packaging film is corrected through an environment error compensation algorithm; analyzing the influence of different film thicknesses of the developing active packaging films on the data of the terahertz time-domain spectroscopy instrument, and performing feedback correction on the film thicknesses of the developing active packaging films; and carrying out uniformity evaluation on the to-be-detected developing active packaging film area. The method has the advantages that non-contact detection is adopted, the efficiency of detecting the film thickness uniformity of the developing active packaging film is effectively improved, and the error influence caused by the film thickness and the environment is avoided.

Description

Technical Field

[0001] The present invention relates to the field of film thickness uniformity detection, and in particular to a method and system for detecting the film thickness uniformity of an intelligent color-developing active packaging film. Background Art

[0002] Color-activated packaging film is a smart packaging material that can directly reflect the freshness of the packaged material, temperature changes or changes in the gas composition inside the package through color changes, thereby monitoring the quality of the packaged material in real time. The key quality indicator of film thickness uniformity directly affects its mechanical properties, barrier properties, optical properties and functional performance. Traditional methods for detecting the thickness uniformity of color-activated packaging film include non-contact optical detection and contact detection methods. Among them, non-contact optical detection mainly includes: laser triangulation reflection method, spectral confocal method, X-ray fluorescence method, infrared interferometry and terahertz time-domain spectroscopy, etc. Contact detection methods mainly include: Ruler / screw micrometer, profilometer / step meter and nanoindenter, but the contact detection method can easily cause scratches on the color-developing active packaging film, destroying the uniformity of the thickness of the color-developing active packaging film, resulting in damage to the color-developing active packaging film while detecting the thickness of the color-developing active packaging film, thereby affecting the quality of the color-developing active packaging film. Terahertz time-domain spectroscopy (THz-TDS) technology detection is a high-precision, non-contact, non-destructive detection method that can effectively obtain the thickness of the color-developing active packaging film without damaging the physical and chemical properties of the color-developing active packaging film, thereby effectively improving the accuracy and reliability of the color-developing active packaging film thickness uniformity detection.

[0003] The existing terahertz time-domain spectroscopy technology for detecting the thickness of color-developing active packaging films lacks analysis of environmental influence errors and film thickness errors. Different film thicknesses and environmental factors will cause the signal attenuation and refractive index of terahertz time-domain spectroscopy to change, thereby affecting the evaluation of the thickness uniformity of color-developing active packaging films and resulting in inaccurate evaluation results. Summary of the Invention

[0004] In order to solve the above technical problems, a method and system for detecting the thickness uniformity of an intelligent color-developing active packaging film is provided. This technical solution solves the problem of the lack of analysis of environmental influence errors and film thickness errors raised in the above background technology. Different film thicknesses and environmental factors will cause the signal attenuation and refractive index of the terahertz time-domain spectrum to change, thereby affecting the evaluation of the thickness uniformity of the color-developing active packaging film and resulting in inaccurate evaluation results.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0006] A method for detecting thickness uniformity of an intelligent color-developing active packaging film, comprising:

[0007] A terahertz time-domain spectroscopy detection device is set up and a parallel detection array is integrated to perform regional scanning of the color-developing active packaging film to be detected;

[0008] Based on the Internet of Things technology, network edge nodes are set up to process data collection, data preprocessing and film thickness uniformity judgment in steps;

[0009] Based on the spectrum generated by the terahertz time-domain spectroscopy instrument and the convolutional neural network algorithm, the spectrum was analyzed to preliminarily confirm the thickness of the color-developing active packaging film;

[0010] Collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm;

[0011] Based on big data, the impact of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data was analyzed, and feedback corrections were made to the color-active packaging film thickness;

[0012] According to the corrected thickness value of the color-developing active packaging film, the uniformity of the color-developing active packaging film area to be tested is evaluated.

[0013] Preferably, the collecting of environmental data near the terahertz time-domain spectroscopy detection device and correcting the thickness of the color-developing active packaging film by using an environmental error compensation algorithm specifically includes:

[0014] An environmental monitoring device is provided to collect environmental data near the terahertz time-domain spectroscopy detection device, wherein the environmental data includes: temperature, humidity, and light intensity;

[0015] A synchronous acquisition method is used to keep the environmental data acquisition frequency consistent with the acquisition frequency of the terahertz time-domain spectroscopy detection device to ensure data consistency;

[0016] Using filtering and normalization algorithms, the collected environmental data and color-active packaging film data are filtered and normalized;

[0017] An environmental error compensation algorithm was established to correct the preliminary confirmation results of the thickness of the color-developing active packaging film.

[0018] Preferably, the analysis of the influence of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data based on big data and the feedback correction of the color-active packaging film thickness specifically includes:

[0019] Set up sample groups with different color-active packaging film thicknesses, sample them separately using a terahertz time-domain spectroscopy detection device, and obtain thickness assessment values ​​and environmentally corrected thickness assessment values;

[0020] Based on the big data, the target film thickness values ​​of the sample groups with different film thicknesses of the color-developing active packaging films were set and recorded as standard values;

[0021] Obtaining absolute errors of the thickness assessment value and the thickness assessment value after environmental correction according to absolute differences between the thickness assessment value and the thickness assessment value after environmental correction and the standard value;

[0022] According to the absolute error between the thickness evaluation value and the thickness evaluation value after environmental correction, compensation values ​​are set for the thickness of different color-developing active packaging films.

[0023] Furthermore, this solution proposes an intelligent color-developing active packaging film thickness uniformity detection system for implementing the above-mentioned intelligent color-developing active packaging film thickness uniformity detection method, comprising:

[0024] A hardware scanning module, which is used to set up a terahertz time-domain spectroscopy detection device and integrate a parallel detection array to perform regional scanning on the color-active packaging film to be detected;

[0025] An edge node module is used to set up a network edge node based on the Internet of Things technology, and to process data collection, data preprocessing, and film thickness uniformity judgment in steps;

[0026] A film thickness confirmation module is used to analyze the spectrum generated by the terahertz time-domain spectroscopy instrument based on a convolutional neural network algorithm to preliminarily confirm the thickness of the color-developing active packaging film;

[0027] An environmental error correction module, which is used to collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm;

[0028] A film thickness error correction module, which is used to analyze the impact of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data based on big data and provide feedback correction to the color-active packaging film thickness;

[0029] A uniformity evaluation module is used to perform uniformity evaluation on the color-developing active packaging film area to be detected according to the corrected film thickness value of the color-developing active packaging film.

[0030] Preferably, the film thickness confirmation module includes:

[0031] A preliminary confirmation unit is used to analyze the spectrum generated by the terahertz time-domain spectroscopy instrument based on a convolutional neural network algorithm to preliminarily confirm the thickness of the color-developing active packaging film;

[0032] An environmental error correction unit, which is used to collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm;

[0033] A film thickness error correction unit is used to analyze the influence of different color-developing active packaging film thicknesses on terahertz time-domain spectrometer data based on big data, and to perform feedback correction on the color-developing active packaging film thickness.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] This solution provides an intelligent color-active packaging film thickness uniformity detection solution. It reduces the environmental influence error and film thickness influence error generated by the terahertz time-domain spectroscopy instrument during the data acquisition process through an environmental error compensation algorithm and film thickness feedback correction, thereby improving the detection accuracy of the color-active packaging film thickness by terahertz time-domain spectroscopy technology, ensuring the reliability of the color-active packaging film thickness uniformity assessment, and improving the efficiency of the color-active packaging film thickness uniformity detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a flow chart of a method for detecting thickness uniformity of an intelligent color-developing active packaging film according to the present invention;

[0037] Figure 2 This is a flow chart for analyzing the spectrum generated by the terahertz time-domain spectroscopy instrument based on the convolutional neural network algorithm to preliminarily determine the thickness of the color-active packaging film;

[0038] Figure 3 This is a flow chart for collecting environmental data near the terahertz time-domain spectroscopy detection device of the present invention and correcting the thickness of the color-active packaging film using an environmental error compensation algorithm;

[0039] Figure 4 This is a flowchart of the present invention that analyzes the impact of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data based on big data and provides feedback correction for the color-active packaging film thickness;

[0040] Figure 5 This is a flow chart of uniformity evaluation of a color-developing active packaging film area to be tested based on the corrected thickness value of the color-developing active packaging film according to the present invention. DETAILED DESCRIPTION

[0041] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0042] Reference Figure 1 As shown, a method for detecting thickness uniformity of an intelligent color-developing active packaging film comprises:

[0043] A terahertz time-domain spectroscopy detection device is set up and a parallel detection array is integrated to perform regional scanning of the color-developing active packaging film to be detected;

[0044] Based on the Internet of Things technology, network edge nodes are set up to process data collection, data preprocessing and film thickness uniformity judgment in steps;

[0045] Based on the spectrum generated by the terahertz time-domain spectroscopy instrument and the convolutional neural network algorithm, the spectrum was analyzed to preliminarily confirm the thickness of the color-developing active packaging film;

[0046] Collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm;

[0047] Based on big data, the impact of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data was analyzed, and feedback corrections were made to the color-active packaging film thickness;

[0048] According to the corrected thickness value of the color-developing active packaging film, the uniformity of the color-developing active packaging film area to be tested is evaluated.

[0049] It can be explained that terahertz time-domain spectroscopy (THz-TDS) technology detection is a high-precision, non-contact, non-destructive detection method, which can obtain the thickness of the color-developing active packaging film without damaging the physical and chemical properties of the color-developing active packaging film. However, when the terahertz time-domain spectroscopy technology is used for large-scale and high-operation detection, it will be affected by the detection environment and thickness of the color-developing active packaging film, resulting in signal attenuation and refractive index changes, thereby affecting the thickness uniformity evaluation of the color-developing active packaging film. Therefore, this scheme reduces the environmental influence error and film thickness influence error through the environmental error compensation algorithm and film thickness feedback correction, thereby improving the detection accuracy of the color-developing active packaging film thickness by the terahertz time-domain spectroscopy technology, ensuring the reliability of the thickness uniformity evaluation of the color-developing active packaging film, and improving the efficiency of the thickness uniformity detection of the color-developing active packaging film.

[0050] The terahertz time-domain spectroscopy detection device is provided and integrated with a parallel detection array to perform regional scanning on the color-developing active packaging film to be detected, specifically comprising:

[0051] According to the width of the color-active packaging film to be detected, the array spacing of the parallel detection array of the terahertz time-domain spectroscopy detection device is set, and each array probe is packaged as an independent module;

[0052] The color-active packaging film is tested by vertically incident linearly polarized terahertz waves to reduce surface scattering of the color-active packaging film.

[0053] A synchronous triggering method is used to perform a comprehensive area scan on the color-active packaging film area to be detected.

[0054] It can be explained that when the thickness of the color-developing active packaging film to be detected is detected by terahertz time-domain spectroscopy technology, in order to improve the detection efficiency and detection accuracy, the terahertz time-domain spectroscopy detection equipment needs to be adjusted. This solution adopts a parallel detection array method combined with a synchronous trigger method to realize the collection of large-scale synchronous data on the thickness of the color-developing active packaging film to be detected, thereby improving the detection efficiency while ensuring data consistency.

[0055] The method of setting up network edge nodes based on the Internet of Things technology and performing step-by-step processing of data collection, data preprocessing and film thickness uniformity judgment specifically includes:

[0056] Based on the Internet of Things technology, the process of terahertz time-domain spectroscopy detection of color-active packaging films is divided into several network edge nodes;

[0057] Among them, the network edge nodes include: data acquisition, data preprocessing and film thickness uniformity judgment;

[0058] Different data transmission channels are divided according to the type of network edge nodes. The overall efficiency of colorimetric active packaging film detection is improved through independent data transmission channels, and data transmission delay is reduced.

[0059] It can be explained that when detecting the thickness of large quantities of color-developing active packaging films, the integrated data collection and data processing method may cause data transmission congestion, resulting in delays in the detection of color-developing active packaging films, which is not conducive to real-time detection. Therefore, this solution sets up network edge nodes based on Internet of Things technology, and processes data collection, data preprocessing and film thickness uniformity judgment in steps. Different data transmission channels are divided according to the type of network edge nodes. The overall efficiency of color-developing active packaging film detection is improved through independent data transmission channels, and data transmission delays are reduced.

[0060] Reference Figure 2 As shown, the spectrum generated by the terahertz time-domain spectroscopy instrument is analyzed based on the convolutional neural network algorithm, and the thickness of the color-developing active packaging film is preliminarily confirmed to include:

[0061] According to the time delay and amplitude change of the terahertz pulse after passing through the color-active packaging film, a time domain waveform diagram of terahertz time domain spectroscopy detection is generated;

[0062] The time domain waveform of terahertz time domain spectroscopy detection is converted into a frequency domain distribution diagram through the Fourier transform formula;

[0063] The main component features of the time domain waveform and frequency domain distribution diagram of terahertz time domain spectroscopy detection are extracted respectively. The main component features of the time domain waveform include: peak time, half-maximum width, and amplitude attenuation coefficient of the pulse; the main component features of the frequency domain distribution diagram include: absorption peak, dispersion curve, and dielectric constant.

[0064] According to the principal component features of the time domain waveform and the frequency domain distribution graph, a fusion image of the principal component features of the time domain waveform and the frequency domain distribution graph is established;

[0065] The Gaussian filtering algorithm is used to perform filtering and noise reduction on the fusion image of the main component features of the acquired time domain waveform and frequency domain distribution diagram;

[0066] Based on the convolutional neural network algorithm, the fusion image of the principal component characteristics of the time domain waveform and the frequency domain distribution diagram was analyzed to preliminarily confirm the thickness value of the color-developing active packaging film.

[0067] It can be explained that the time domain waveform of the color-developing active packaging film can be obtained through the terahertz time-domain spectroscopy device, and the refractive index, absorption coefficient and thickness information of the color-developing active packaging film can be obtained through the peak time, half-height width and amplitude attenuation coefficient of the pulse of the time domain waveform. Among them, the pulse time delay is proportional to the refractive index of the sample, and the amplitude attenuation coefficient reflects the sample's absorption capacity of terahertz waves. In the multi-layer structure, the thickness of each layer can be calculated by the "echo" time difference of interface reflection, and the accuracy can reach micron level. Secondly, the time domain waveform can be converted into a frequency domain distribution diagram through Fourier transform. The absorption peak, dispersion curve and dielectric constant in the frequency domain distribution diagram reflect the multi-layer structure of the color-developing active packaging film, wherein the multi-layer structure includes: a protective layer, a color-developing layer and a base material layer. Finally, through the principal component characteristics of the time domain waveform and the frequency domain distribution diagram, the convolutional neural network algorithm is used to perform image analysis to preliminarily confirm the thickness value of the color-developing active packaging film.

[0068] Reference Figure 3 As shown, the collecting of environmental data near the terahertz time-domain spectroscopy detection device and the correction of the thickness of the color-developing active packaging film by using an environmental error compensation algorithm specifically include:

[0069] An environmental monitoring device is provided to collect environmental data near the terahertz time-domain spectroscopy detection device, wherein the environmental data includes: temperature, humidity, and light intensity;

[0070] A synchronous acquisition method is used to keep the environmental data acquisition frequency consistent with the acquisition frequency of the terahertz time-domain spectroscopy detection device to ensure data consistency;

[0071] Using filtering and normalization algorithms, the collected environmental data and color-active packaging film data are filtered and normalized;

[0072] An environmental error compensation algorithm was established to correct the preliminary confirmation results of the thickness of the color-developing active packaging film.

[0073] It can be explained that when using terahertz time-domain spectroscopy detection technology to detect the thickness of color-developing active packaging film, the environment will affect the detection results, among which temperature, humidity and light intensity are the main influencing factors. Therefore, in order to ensure the accuracy of the thickness measurement of the color-developing active packaging film, this solution uses a linear regression method based on the compensation results of each environmental influencing factor to obtain the compensation value of the environmental error, and then corrects the preliminary confirmed value of the thickness of the color-developing active packaging film by the environmental error compensation value to ensure the accuracy and reliability of the data; the expression for the error compensation value of the preliminary confirmed result of the thickness of the color-developing active packaging film due to temperature is:

[0074]

[0075] Where A is the error compensation value of the preliminary confirmation result of the thickness of the color-developing active packaging film due to temperature, R is the refractive index at the standard temperature, β and θ are the linear temperature correction coefficient and the quadratic temperature correction coefficient respectively. Multiple sets of data can be obtained through test experiments and the linear regression equation group can be used to calculate the linear temperature correction coefficient. The unit of the linear temperature correction coefficient is The unit of secondary temperature correction coefficient is T is the collected temperature value, is the set standard temperature value, γ is the thickness temperature correction coefficient, and d is the set color-developing active packaging film thickness value;

[0076] The error compensation value of the preliminary confirmation result of humidity on the thickness of the color-developing active packaging film is expressed as follows:

[0077]

[0078] Where B is the error compensation value of humidity on the preliminary confirmation result of the thickness of the color-developing active packaging film, k is the humidity influence coefficient, which can be calibrated through experiments, and RH is the relative humidity value;

[0079] The error compensation value of the preliminary confirmation result of the light intensity on the thickness of the color-developing active packaging film is expressed as follows:

[0080] C=μlog(I+1)

[0081] Where C is the error compensation value of the preliminary confirmation result of the thickness of the color-developing active packaging film by light intensity, μ is the light intensity influence coefficient, and I is the light intensity value;

[0082] The environmental error compensation algorithm expression is:

[0083] ΔG=α1A+α2B+α3C+δ

[0084] Where ΔG is the compensation value for the error in the preliminary confirmation result of the thickness of the color-developing active packaging film due to environmental influences; α1, α2, and α3 are the weights of the compensation values ​​for the error in the preliminary confirmation result of the thickness of the color-developing active packaging film due to temperature, humidity, and light intensity, respectively; A, B, and C are the compensation values ​​for the error in the preliminary confirmation result of the thickness of the color-developing active packaging film due to temperature, humidity, and light intensity, respectively; and δ is the correction constant term.

[0085] Reference Figure 4 As shown, the analysis of the influence of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data based on big data and the feedback correction of the color-active packaging film thickness specifically include:

[0086] Set up sample groups with different color-active packaging film thicknesses, sample them separately using a terahertz time-domain spectroscopy detection device, and obtain thickness assessment values ​​and environmentally corrected thickness assessment values;

[0087] Based on the big data, the target film thickness values ​​of the sample groups with different film thicknesses of the color-developing active packaging films were set and recorded as standard values;

[0088] Obtaining absolute errors of the thickness assessment value and the thickness assessment value after environmental correction according to absolute differences between the thickness assessment value and the thickness assessment value after environmental correction and the standard value;

[0089] According to the absolute error between the thickness evaluation value and the thickness evaluation value after environmental correction, compensation values ​​are set for the thickness of different color-developing active packaging films.

[0090] It can be explained that the film layer has an absorption and scattering effect on terahertz waves. When the THz wave passes through the film layer, part of the energy will be absorbed by the film layer, and part of the energy will deviate from the original propagation direction due to scattering, resulting in a weakening of the signal intensity. Secondly, in a multi-layer film structure, the terahertz waves between the layers of the film may interfere with each other, which may cause the resonance frequency to shift or the signal intensity to change, affecting the final detection result. Therefore, this scheme sets up sample groups with different color-developing active packaging film thicknesses, and marks the target film thickness value of each sample group with different color-developing active packaging film thicknesses. Through the detection of terahertz time-domain spectroscopy, the detection thickness value and the target film thickness value are obtained. The functional relationship between the detection thickness value and the target film thickness value is fitted based on the PID algorithm, thereby determining the influence of different film thicknesses on terahertz time-domain spectroscopy detection, and then setting compensation values ​​for different color-developing active packaging film thicknesses through the functional relationship to ensure the accuracy of film thickness detection;

[0091] The fitting function relationship expression between the detected thickness value and the target film thickness value is:

[0092]

[0093] Where, Δu is the compensation value of the thickness of the color-active packaging film, K p , K i , K d are proportional, integral, and differential coefficients respectively, and e(t) is the difference between the detected thickness value and the target film thickness value. The expression is: Where x i is the preliminary detection value of the thickness of the i-th color-developing active packaging film, y i is the thickness detection value of the color-developing active packaging film after correction of the i-th environmental error, is the target film thickness value;

[0094] By judging the size of the error between the film thickness detection value after environmental error correction and the target film thickness value, as well as the error between the uncorrected film thickness detection value and the target film thickness value, the minimum error is intelligently selected as the correction value, thereby effectively avoiding the influence of the numerical deviation caused by the detection error of the environmental sensor itself under stable environment on the final result, and further improving the accuracy and reliability of terahertz time-domain spectroscopy for color-active packaging film thickness detection.

[0095] Reference Figure 5 As shown, the uniformity evaluation of the color-developing active packaging film area to be tested based on the corrected color-developing active packaging film thickness value specifically includes:

[0096] Obtaining all corrected thickness values ​​of the color-developing active packaging films in the color-developing active packaging film area to be detected;

[0097] Calculate the average of all corrected thickness values ​​of the color-developing active packaging films in the area;

[0098] Calculate the variance of all corrected film thickness values ​​of color-developing active packaging films in the region;

[0099] Determining the variation value of all the modified color-developing active packaging film thickness values ​​within the calculation area according to the average value and variance of all the modified color-developing active packaging film film thickness values ​​within the calculation area;

[0100] Based on big data, a mutation threshold is set to determine whether the mutation value of the thickness of the color-developing active packaging film exceeds the mutation threshold. If so, it means that the uniformity quality of the color-developing active packaging film to be tested is poor. If not, it means that the uniformity of the color-developing active packaging film to be tested meets the requirements.

[0101] It can be explained that this scheme obtains the film thickness value in the area of ​​the color-developing active packaging film to be tested, calculates the average and variance of all the corrected film thickness values ​​of the color-developing active packaging film, obtains the variation value of the film thickness value of all the corrected color-developing active packaging film in the area, and judges whether the uniformity quality of the color-developing active packaging film to be tested meets the standard through the variation value, so the uniformity of the color-developing active packaging film can be evaluated efficiently and accurately.

[0102] Furthermore, based on the same inventive concept as the above-mentioned intelligent color-developing active packaging film thickness uniformity detection method, this solution proposes an intelligent color-developing active packaging film thickness uniformity detection system, comprising:

[0103] A hardware scanning module, which is used to set up a terahertz time-domain spectroscopy detection device and integrate a parallel detection array to perform regional scanning on the color-active packaging film to be detected;

[0104] An edge node module is used to set up a network edge node based on the Internet of Things technology, and to process data collection, data preprocessing, and film thickness uniformity judgment in steps;

[0105] A film thickness confirmation module is used to analyze the spectrum generated by the terahertz time-domain spectroscopy instrument based on a convolutional neural network algorithm to preliminarily confirm the thickness of the color-developing active packaging film;

[0106] An environmental error correction module, which is used to collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm;

[0107] A film thickness error correction module, which is used to analyze the impact of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data based on big data and provide feedback correction to the color-active packaging film thickness;

[0108] a uniformity evaluation module, configured to evaluate the uniformity of the color-developing active packaging film region to be inspected based on the corrected thickness value of the color-developing active packaging film;

[0109] The film thickness confirmation module includes:

[0110] A preliminary confirmation unit is used to analyze the spectrum generated by the terahertz time-domain spectroscopy instrument based on a convolutional neural network algorithm to preliminarily confirm the thickness of the color-developing active packaging film;

[0111] An environmental error correction unit, which is used to collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm;

[0112] A film thickness error correction unit is used to analyze the influence of different color-developing active packaging film thicknesses on terahertz time-domain spectrometer data based on big data, and to perform feedback correction on the color-developing active packaging film thickness.

[0113] In summary, the advantages of the present invention are: the use of non-contact detection effectively improves the efficiency of the film thickness uniformity detection of the color-developing active packaging film, and avoids the error influence caused by the film thickness and the environment.

[0114] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for detecting thickness uniformity of an intelligent color-developing active packaging film, characterized in that: include: A terahertz time-domain spectroscopy detection device is set up and a parallel detection array is integrated to perform regional scanning of the color-developing active packaging film to be detected; Based on the Internet of Things technology, network edge nodes are set up to process data collection, data preprocessing and film thickness uniformity judgment in steps; Based on the spectrum generated by the terahertz time-domain spectroscopy instrument and the convolutional neural network algorithm, the spectrum was analyzed to preliminarily confirm the thickness of the color-developing active packaging film; Collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm; Based on big data, the impact of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data was analyzed, and feedback corrections were made to the color-active packaging film thickness; According to the corrected thickness value of the color-developing active packaging film, the uniformity of the color-developing active packaging film area to be tested is evaluated.

2. The method for detecting thickness uniformity of an intelligent color-developing active packaging film according to claim 1, characterized in that: The terahertz time-domain spectroscopy detection device is provided and integrated with a parallel detection array to perform regional scanning on the color-developing active packaging film to be detected, specifically comprising: According to the width of the color-active packaging film to be detected, the array spacing of the parallel detection array of the terahertz time-domain spectroscopy detection device is set, and each array probe is packaged as an independent module; The color-active packaging film is tested by vertically incident linearly polarized terahertz waves to reduce surface scattering of the color-active packaging film. A synchronous triggering method is used to perform a comprehensive area scan on the color-active packaging film area to be detected.

3. The method for detecting thickness uniformity of an intelligent color-developing active packaging film according to claim 2, characterized in that: The method of setting up network edge nodes based on the Internet of Things technology and performing step-by-step processing of data collection, data preprocessing and film thickness uniformity judgment specifically includes: Based on the Internet of Things technology, the process of terahertz time-domain spectroscopy detection of color-active packaging films is divided into several network edge nodes; Among them, the network edge nodes include: data acquisition, data preprocessing and film thickness uniformity judgment; Different data transmission channels are divided according to the type of network edge nodes. The overall efficiency of colorimetric active packaging film detection is improved through independent data transmission channels, and data transmission delay is reduced.

4. The method for detecting thickness uniformity of an intelligent color-developing active packaging film according to claim 3, characterized in that: The spectrum generated by the terahertz time-domain spectroscopy instrument is analyzed based on a convolutional neural network algorithm to preliminarily confirm the thickness of the color-developing active packaging film. Specifically, the following information is provided: According to the time delay and amplitude change of the terahertz pulse after passing through the color-active packaging film, a time domain waveform diagram of terahertz time domain spectroscopy detection is generated; The time domain waveform of terahertz time domain spectroscopy detection is converted into a frequency domain distribution diagram through the Fourier transform formula; The main component features of the time domain waveform and frequency domain distribution diagram of terahertz time domain spectroscopy detection are extracted respectively. The main component features of the time domain waveform include: peak time, half-maximum width, and amplitude attenuation coefficient of the pulse; the main component features of the frequency domain distribution diagram include: absorption peak, dispersion curve, and dielectric constant. According to the principal component features of the time domain waveform and the frequency domain distribution graph, a fusion image of the principal component features of the time domain waveform and the frequency domain distribution graph is established; The Gaussian filtering algorithm is used to perform filtering and noise reduction on the fusion image of the main component features of the acquired time domain waveform and frequency domain distribution diagram; Based on the convolutional neural network algorithm, the fusion image of the principal component characteristics of the time domain waveform and the frequency domain distribution diagram was analyzed to preliminarily confirm the thickness value of the color-developing active packaging film.

5. The method for detecting thickness uniformity of an intelligent color-developing active packaging film according to claim 4, characterized in that: The collecting of environmental data near the terahertz time-domain spectroscopy detection device and the correction of the thickness of the color-developing active packaging film by using an environmental error compensation algorithm specifically include: An environmental monitoring device is provided to collect environmental data near the terahertz time-domain spectroscopy detection device, wherein the environmental data includes: temperature, humidity, and light intensity; A synchronous acquisition method is used to keep the environmental data acquisition frequency consistent with the acquisition frequency of the terahertz time-domain spectroscopy detection device to ensure data consistency; Using filtering and normalization algorithms, the collected environmental data and color-active packaging film data are filtered and normalized; An environmental error compensation algorithm was established to correct the preliminary confirmation results of the thickness of the color-developing active packaging film.

6. The method for detecting thickness uniformity of an intelligent color-developing active packaging film according to claim 5, characterized in that: The analysis of the influence of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data based on big data and the feedback correction of the color-active packaging film thickness specifically include: Set up sample groups with different color-active packaging film thicknesses, sample them separately using a terahertz time-domain spectroscopy detection device, and obtain thickness assessment values ​​and environmentally corrected thickness assessment values; Based on the big data, the target film thickness values ​​of the sample groups with different film thicknesses of the color-developing active packaging films were set and recorded as standard values; Obtaining absolute errors of the thickness assessment value and the thickness assessment value after environmental correction according to absolute differences between the thickness assessment value and the thickness assessment value after environmental correction and the standard value; According to the absolute error between the thickness evaluation value and the thickness evaluation value after environmental correction, compensation values ​​are set for the thickness of different color-developing active packaging films.

7. The method for detecting thickness uniformity of an intelligent color-developing active packaging film according to claim 6, characterized in that: The uniformity evaluation of the color-developing active packaging film area to be tested based on the corrected color-developing active packaging film thickness value specifically includes: Obtaining all corrected thickness values ​​of the color-developing active packaging films in the color-developing active packaging film area to be detected; Calculate the average of all corrected thickness values ​​of the color-developing active packaging films in the area; Calculate the variance of all corrected film thickness values ​​of color-developing active packaging films in the region; Determining the variation value of all the modified color-developing active packaging film thickness values ​​within the calculation area according to the average value and variance of all the modified color-developing active packaging film film thickness values ​​within the calculation area; Based on big data, a mutation threshold is set to determine whether the mutation value of the thickness of the color-developing active packaging film exceeds the mutation threshold. If so, it means that the uniformity quality of the color-developing active packaging film to be tested is poor. If not, it means that the uniformity of the color-developing active packaging film to be tested meets the requirements.

8. An intelligent color-developing active packaging film thickness uniformity detection system, characterized in that: A method for detecting thickness uniformity of an intelligent color-developing active packaging film according to any one of claims 1 to 7, comprising: A hardware scanning module, which is used to set up a terahertz time-domain spectroscopy detection device and integrate a parallel detection array to perform regional scanning on the color-active packaging film to be detected; An edge node module is used to set up a network edge node based on the Internet of Things technology, and to process data collection, data preprocessing, and film thickness uniformity judgment in steps; A film thickness confirmation module is used to analyze the spectrum generated by the terahertz time-domain spectroscopy instrument based on a convolutional neural network algorithm to preliminarily confirm the thickness of the color-developing active packaging film; An environmental error correction module, which is used to collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm; A film thickness error correction module, which is used to analyze the impact of different color-active packaging film thicknesses on terahertz time-domain spectroscopy instrument data based on big data and provide feedback correction to the color-active packaging film thickness; A uniformity evaluation module is used to perform uniformity evaluation on the color-developing active packaging film area to be detected according to the corrected film thickness value of the color-developing active packaging film.

9. The intelligent color-developing active packaging film thickness uniformity detection system according to claim 8, characterized in that: The film thickness confirmation module includes: A preliminary confirmation unit is used to analyze the spectrum generated by the terahertz time-domain spectroscopy instrument based on a convolutional neural network algorithm to preliminarily confirm the thickness of the color-developing active packaging film; An environmental error correction unit, which is used to collect environmental data near the terahertz time-domain spectroscopy detection device and correct the thickness of the color-developing active packaging film using an environmental error compensation algorithm; A film thickness error correction unit is used to analyze the influence of different color-developing active packaging film thicknesses on terahertz time-domain spectrometer data based on big data, and to perform feedback correction on the color-developing active packaging film thickness.

Images