A Method and System for Monitoring the Production Quality of Self-Adhesive Labels Based on Image Analysis

Abstract

This invention discloses a method and system for monitoring the production quality of self-adhesive labels based on image analysis, belonging to the field of label quality monitoring technology. The method includes: acquiring self-adhesive labels; performing image analysis on monitoring images of the face material layer to determine the label face material quality coefficient; conducting comprehensive adhesive performance testing on the adhesive layer to obtain the label adhesive quality coefficient; integrating bilateral characteristic detection of the backing paper layer according to the weighting of the backing paper bilateral characteristic detection to obtain the label backing paper quality coefficient; setting label quality expectation conditions based on the self-adhesive labels; performing expectation compliance judgment; and outputting the self-adhesive label quality detection result. This invention solves the technical problems of lack of accuracy and comprehensiveness in self-adhesive label production quality monitoring and low quality control efficiency in the prior art, achieving the technical effect of improving the accuracy and comprehensiveness of self-adhesive label production quality monitoring and enhancing quality control efficiency.

Description

Technical Field

[0001] This invention relates to the field of label quality monitoring technology, specifically to a method and system for monitoring the production quality of self-adhesive labels based on image analysis. Background Technology

[0002] In the field of self-adhesive label production, quality monitoring has always been a crucial step. With the increasing market demand for self-adhesive labels and the growing diversification of application scenarios, the quality requirements for self-adhesive labels are also rising. However, current technologies for quality monitoring in self-adhesive label production face several challenges. Firstly, traditional quality monitoring methods often focus on a single indicator or aspect, lacking comprehensive testing of all layers of the self-adhesive label (such as the face stock, adhesive layer, and backing paper). This one-sided approach results in inaccurate and incomplete quality monitoring results, failing to truly reflect the overall quality of the self-adhesive label. Secondly, existing quality monitoring methods also suffer from inefficiencies in quality control. Due to the lack of advanced and intelligent testing methods, the testing process is cumbersome and time-consuming, increasing production costs and hindering the timely detection and resolution of quality problems during production. This can lead to inconsistent quality in the produced self-adhesive labels.

[0003] Existing technologies suffer from technical problems such as a lack of accuracy and comprehensiveness in monitoring the quality of self-adhesive production, as well as low efficiency in quality control. Summary of the Invention

[0004] This application provides a method and system for monitoring the production quality of self-adhesive labels based on image analysis, which is used to address the technical problems of lack of accuracy and comprehensiveness in the quality monitoring of self-adhesive label production, as well as the low efficiency of quality control in the prior art.

[0005] In view of the above problems, this application provides a method and system for monitoring the production quality of self-adhesive labels based on image analysis.

[0006] The first aspect of this application provides a method for monitoring the production quality of self-adhesive labels based on image analysis, the method comprising:

[0007] A self-adhesive label is obtained, comprising a face layer, an adhesive layer, and a backing paper layer. Image analysis is performed on the face layer monitoring image to determine the face layer quality coefficient. The adhesive layer is subjected to comprehensive adhesion performance testing based on multi-dimensional adhesion performance detection factors, including initial tack, holding power, and adhesion migration, to obtain the label adhesive quality coefficient. The backing paper layer is subjected to bilateral characteristic detection integration based on the backing paper bilateral characteristic detection weights to obtain the label backing paper quality coefficient. Based on the self-adhesive label, label quality expectation conditions are set, including expected face layer quality coefficient, expected label adhesive quality coefficient, and expected label backing paper quality coefficient. Based on the label quality expectation conditions, the expected compliance of the face layer quality coefficient, the adhesive quality coefficient, and the backing paper quality coefficient is judged, and the self-adhesive label quality detection result is output.

[0008] A second aspect of this application provides an image analysis-based self-adhesive label production quality monitoring system, the system comprising:

[0009] The system includes: a self-adhesive label acquisition module for acquiring self-adhesive labels, wherein each self-adhesive label comprises a face material layer, an adhesive layer, and a backing paper layer; a face material quality coefficient determination module for determining the face material quality coefficient by performing image analysis based on monitoring images of the face material layer; an adhesive quality coefficient acquisition module for comprehensively testing the adhesive performance of the adhesive layer based on multi-dimensional adhesion performance detection factors, wherein the multi-dimensional adhesion performance detection factors include initial tack, holding tack, and adhesion migration; and a backing paper quality coefficient acquisition module. The acquisition module is used to integrate the bilateral characteristic detection of the backing paper layer according to the bilateral characteristic detection weight to obtain the label backing paper quality coefficient; the label quality expectation condition setting module sets the label quality expectation conditions based on the self-adhesive label, wherein the label quality expectation conditions include the expected face material quality coefficient, the expected label adhesive quality coefficient, and the expected label backing paper quality coefficient; the self-adhesive label quality detection result output module performs expectation compliance judgment on the label face material quality coefficient, the label adhesive quality coefficient, and the label backing paper quality coefficient based on the label quality expectation conditions, and outputs the self-adhesive label quality detection result.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] The process involves acquiring self-adhesive labels; performing image analysis on the monitoring images of the face material layer to determine the label face material quality coefficient; conducting comprehensive adhesion performance testing on the adhesive layer based on multi-dimensional adhesion performance detection factors to obtain the label adhesive quality coefficient; integrating bilateral characteristic detection on the backing paper layer to obtain the label backing paper quality coefficient; setting label quality expectation conditions based on the self-adhesive labels; and judging the expected compliance of the label face material quality coefficient, the label adhesive quality coefficient, and the label backing paper quality coefficient based on the label quality expectation conditions, outputting the self-adhesive label quality inspection result. This achieves the technical effect of improving the accuracy and comprehensiveness of self-adhesive label production quality monitoring and enhancing quality control efficiency. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A schematic diagram of the process for monitoring the production quality of self-adhesive labels based on image analysis, provided in an embodiment of this application.

[0014] Figure 2 A schematic diagram of the structure of an image analysis-based self-adhesive label production quality monitoring system provided in this application embodiment.

[0015] Figure labeling: 10 Self-adhesive label acquisition module, 20 Label face material quality coefficient determination module, 30 Label adhesive quality coefficient acquisition module, 40 Label backing paper quality coefficient acquisition module, 50 Label quality expectation condition setting module, 60 Self-adhesive label quality inspection result output module. Detailed Implementation

[0016] This application provides a method and system for monitoring the production quality of self-adhesive labels based on image analysis, which addresses the technical problems of the lack of accuracy and comprehensiveness in the quality monitoring of self-adhesive label production, as well as the low efficiency of quality control in the prior art.

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0018] Example 1, as Figure 1As shown, this application provides a method for monitoring the production quality of self-adhesive labels based on image analysis, the method comprising:

[0019] Step S100: Obtain a self-adhesive label, wherein the self-adhesive label includes a face layer, an adhesive layer, and a backing paper layer.

[0020] Specifically, self-adhesive labels are a common material used for identification and adhesion. They consist of three main parts: a face stock layer, an adhesive layer, and a backing paper layer. The face stock layer is the outermost layer of the self-adhesive label, directly in contact with the outside environment. It typically has a certain degree of printability and abrasion resistance to ensure clear printing of various identification information and maintain surface integrity during use. The face stock layer can be made of various materials, such as paper or plastic film, with different materials suitable for different application scenarios. The adhesive layer, located between the face stock layer and the backing paper layer, is the key part that allows the self-adhesive label to adhere to the surface of an object. The performance of the adhesive layer directly affects the label's adhesion and durability. Its characteristics include initial tack, holding power, and adhesion migration, which need to be evaluated in detail during subsequent testing. The backing paper layer is the bottom layer of the self-adhesive label. It supports and protects the adhesive layer. The backing paper layer needs to have sufficient strength to prevent damage during label production and use. Simultaneously, it needs appropriate anti-stick properties so that the label can be easily peeled off from the backing paper during use. The process of obtaining self-adhesive labels involves directly extracting samples from the production line or randomly selecting a certain number of samples from a batch of labels that have already been produced. These samples will be used as the objects of subsequent quality monitoring to ensure that the quality of the entire self-adhesive label production process is controllable.

[0021] Step S200: Perform image analysis based on the surface material monitoring image of the surface material layer to determine the label surface material quality coefficient.

[0022] Specifically, the process involves acquiring monitoring images of the faceplate layer using specialized image acquisition equipment. This equipment accurately captures the surface features of the faceplate layer, including color, texture, flatness, and the presence of defects. Then, image analysis is performed on the acquired faceplate layer monitoring images, a process involving multiple technologies and algorithms. For example, color analysis algorithms are used to assess the color consistency of the faceplate layer. By comparing color values ​​in different areas of the image, color deviations are identified, which can affect the label's appearance quality and printing effect. Texture analysis is also crucial. By analyzing the uniformity and clarity of the texture, the material properties and surface treatment of the faceplate layer can be understood. Uneven or unclear textures indicate quality problems with the faceplate layer, such as uneven material mixing or imperfect surface processing. Flatness analysis determines the smoothness of the faceplate layer by detecting surface undulations in the image. Uneven faceplate layers lead to printing difficulties and affect the label's clarity and accuracy. Furthermore, the images are inspected for defects such as scratches, holes, and spots. These defects directly affect the label's appearance quality and performance. A comprehensive evaluation of the surface layer monitoring images was conducted to ultimately determine a label surface quality coefficient that can quantify the quality status of the surface layer. This coefficient serves as an important basis for judging whether the surface layer quality meets the requirements and provides crucial data support for subsequent self-adhesive label quality monitoring.

[0023] Step S300: Perform a comprehensive adhesive performance test on the adhesive layer based on multidimensional adhesive performance testing factors to obtain the label adhesive quality coefficient, wherein the multidimensional adhesive performance testing factors include initial tack, holding tack and adhesive migration.

[0024] Specifically, to obtain the label adhesive quality coefficient, a comprehensive adhesive performance test of the adhesive layer is required, based on multi-dimensional adhesive performance testing factors including initial tack, holding tack, and adhesion migration. For initial tack, a test instrument is activated based on relevant factors to perform multiple tests on the adhesive layer, acquiring and processing the data. The results are then used to build a deep learning model, outputting evaluation coefficients that are added to the test results. For holding tack, the test instrument is activated according to the command, and the results are used to train a neural network. When the loss coefficient condition is met, an evaluation channel is generated, and the coefficients are added to the test results. Regarding adhesion migration, multiple test scenarios are built using activated scenario factors to test the adhesive layer and obtain the amount of migrated adhesive. The coefficients are then evaluated and calculated and added to the test results. Finally, the multi-dimensional adhesive performance testing factors are comprehensively considered, weighted, and calculated. The weighted average of the comprehensive test results yields the label adhesive quality coefficient.

[0025] Step S400: Perform bilateral characteristic detection and integration on the backing paper layer according to the bilateral characteristic detection weight to obtain the label backing paper quality coefficient.

[0026] Specifically, the work is carried out based on the weighting of the bilateral characteristics of the backing paper. For the backing paper layer, anti-stick properties and strength properties are tested separately. For anti-stick properties testing, multiple backing paper samples (with the same production parameters as the backing paper layer) are obtained. Then, a peel force tester is used to test the peel force of each sample, obtaining multiple test peel forces. The peel force test results are calculated by summation, and then the pre-built anti-stick properties testing model is activated. The peel force test results are input into the model, and the anti-stick properties coefficient of the backing paper is output. For strength properties testing, the strength properties coefficient of the backing paper is obtained through appropriate testing methods. Then, the obtained anti-stick properties coefficient and strength properties coefficient of the backing paper are normalized to obtain the bilateral characteristics testing results of the backing paper. Finally, based on the weighting of the bilateral characteristics testing of the backing paper, this testing result is weighted and calculated to generate the label backing paper quality coefficient. This coefficient can reflect the quality status of the backing paper layer and provides an important basis for the overall quality assessment of self-adhesive labels.

[0027] Step S500: Based on the self-adhesive label, set the label quality expectation conditions, wherein the label quality expectation conditions include the expected face material quality coefficient, the expected label adhesive quality coefficient, and the expected label backing paper quality coefficient.

[0028] Specifically, based on the acquired self-adhesive labels, expected label quality conditions are set, encompassing the expected face material quality coefficient, the expected label adhesive quality coefficient, and the expected label backing paper quality coefficient. The expected face material quality coefficient is a pre-defined quantitative indicator of the ideal quality the face material layer should achieve, based on the application scenario and quality standards of the self-adhesive label. It comprehensively considers the requirements of the face material layer in terms of color, texture, flatness, and absence of defects. The expected label adhesive quality coefficient is a quantitative indicator set for the adhesive layer, considering the ideal standards it should meet in terms of multi-dimensional adhesive performance such as initial tack, holding power, and adhesion migration. The expected label backing paper quality coefficient is a quantitative indicator set based on the ideal state that the backing paper layer should achieve in terms of anti-stick properties and strength properties. By clarifying these expected conditions, a clear reference standard is provided for judging the quality inspection results of self-adhesive labels.

[0029] Step S600: Based on the expected label quality conditions, determine whether the label face material quality coefficient, the label adhesive quality coefficient, and the label backing paper quality coefficient meet the expected requirements, and output the self-adhesive label quality inspection results.

[0030] Specifically, based on the set label quality expectations, the label face material quality coefficient is judged to meet the expected conditions. The actual measured label face material quality coefficient is carefully compared with the expected face material quality coefficient. If the two values ​​are equal or the difference is within a reasonable error range, then the label face material quality can be determined to meet the expected conditions. This means that the face material layer has reached the expected quality standards in terms of color, texture, flatness, and lack of defects. Next, a similar judgment is made on the label adhesive quality coefficient. The actual label adhesive quality coefficient is compared with the expected label adhesive quality coefficient. If the actual coefficient meets the requirements set for the expected label adhesive quality coefficient, including standards for multi-dimensional adhesive performance such as initial tack, holding power, and adhesion migration, then the adhesive layer quality is determined to meet the expectations. Then, the label backing paper quality coefficient is judged to meet the expected conditions. The actual label backing paper quality coefficient is compared with the expected label backing paper quality coefficient. When the actual coefficient matches the expected coefficient, considering that the backing paper layer has reached the expected state in terms of anti-stick properties and strength properties, the backing paper layer quality is determined to meet the expectations. Finally, the quality inspection result of the self-adhesive label is output by combining the judgment results of the above three aspects. If the label face material quality coefficient, label adhesive quality coefficient, and label backing paper quality coefficient all meet their respective expected conditions, then the overall quality inspection result of the self-adhesive label is qualified. Conversely, if any coefficient does not meet the expected conditions, the quality inspection result of the self-adhesive label is judged as unqualified, thus providing an accurate assessment conclusion for the production quality of self-adhesive labels.

[0031] In one possible implementation, step S300 further includes:

[0032] Step S310: Perform adhesive performance testing on the adhesive layer according to the multidimensional adhesive performance testing factor to obtain the adhesive performance testing results.

[0033] Step S320: Initialize the weight settings according to the multidimensional adhesion performance detection factor, and determine the first weight condition for adhesion performance detection.

[0034] Step S330: Calculate the proportion based on the adhesion performance test results to obtain the second weighting condition for adhesion performance test.

[0035] Step S340: Calculate the concentrated value of each dimension's weight based on the first weight condition and the second weight condition for adhesion performance detection to obtain the third weight condition for adhesion performance detection.

[0036] Step S350: Perform a weighted calculation on the adhesion performance test results according to the third weighting condition of the adhesion performance test, and output the label adhesive quality coefficient.

[0037] Specifically, the adhesive performance testing of an adhesive layer is a process that comprehensively considers multiple adhesive performance testing factors to obtain adhesive performance test results including initial tack evaluation coefficients, holding power evaluation coefficients, and adhesion migration evaluation coefficients. First, initial tack is tested. Based on the initial tack requirement in the multi-dimensional adhesive performance testing factors, an initial tack test command is generated. This command activates an initial tack tester to perform multiple initial tack tests on the adhesive layer, obtaining multiple initial tack test data. The initial tack test result is obtained by ensemble value calculation on these data. Then, a deep learning model is built using the initial tack test result record set and the initial tack test evaluation record set. Finally, based on the initial tack test results, the model outputs the initial tack test evaluation coefficient, which is added to the adhesive performance test results. Next, holding power is tested. According to the holding power test command, a holding power tester is activated to perform a holding power confidence test on the adhesive layer, obtaining the holding power test result. Using the tackiness test result record set as input and the tackiness detection and evaluation record set as output, a residual neural network is trained. At each predetermined training iteration, a tackiness detection and evaluation loss coefficient is obtained. When this loss coefficient is less than the tackiness detection and evaluation loss threshold, a tackiness detection and evaluation channel is generated. Based on the tackiness test results, the tackiness detection and evaluation coefficient is obtained through this channel and added to the adhesion performance test results. Finally, adhesion migration is detected. Based on the adhesion migration requirements in the multi-dimensional adhesion performance test factors, adhesion migration detection instructions are generated, activating adhesion migration detection scenario factors, including adhesion target, adhesion duration, test environment, and applied pressure. Multiple adhesion migration detection scenarios are constructed based on these scenario factors. Adhesion migration tests are performed on the adhesive layer according to these scenarios, obtaining multiple migration adhesive amounts. Based on these scenarios and migration adhesive amounts, adhesion migration evaluation is performed, obtaining multiple adhesion migration evaluation coefficients. These coefficients are then quantified to generate adhesion migration detection evaluation coefficients, which are added to the adhesion performance test results.

[0038] The first weight condition for adhesion performance testing is determined by setting initial weights based on multidimensional adhesion performance testing factors. This process requires comprehensive consideration of all aspects included in the multidimensional adhesion performance testing factors. For factors such as initial tack, holding tack, and adhesion migration, weights are assigned based on their relative importance in the overall performance of the adhesive and their impact on the final quality of the self-adhesive label. For example, if initial tack has a significant impact on the initial effect of label adhesion in practical applications, then a relatively high initial weight will be given to initial tack; similarly, if holding tack is crucial for the adhesion stability of the label during long-term use, an appropriate weight will be assigned to it; and for adhesion migration, if it has a significant impact on the adhesion quality of the label in different environments, a corresponding weight will also be determined. The weight condition determined by comprehensively considering all factors in this way is the first weight condition for adhesion performance testing, which will serve as the basis for subsequent calculations and adjustments of weights.

[0039] The adhesion performance test results are analyzed, including the initial tack evaluation coefficient, holding tack evaluation coefficient, and adhesion migration evaluation coefficient. The proportion of each coefficient in the overall adhesion performance test results is calculated. For example, the initial tack evaluation coefficient is divided by the sum of all coefficients to obtain the proportion of initial tack. In this way, the proportions of initial tack, holding tack, and adhesion migration in the adhesion performance test results are calculated separately. These proportions constitute the second weighting condition for adhesion performance testing, reflecting the relative importance of each adhesion performance factor in the actual test results, and providing a basis for further precise adjustment of the weights.

[0040] Based on the first and second weighting conditions for adhesion performance testing obtained earlier, the concentrated values ​​of the weights for each dimension are calculated to obtain the third weighting condition for adhesion performance testing. First, it is clarified that the first weighting condition for adhesion performance testing is obtained by initializing the weights based on the multi-dimensional adhesion performance testing factors, reflecting the theoretically expected weight allocation of each adhesion performance factor (such as initial tack, holding tack, and adhesion migration). The second weighting condition for adhesion performance testing is calculated based on the proportion of adhesion performance testing results, reflecting the weight of each factor in actual testing. Then, these two weighting conditions are considered comprehensively. For each adhesion performance factor, the weight values ​​in the first and second weighting conditions are integrated and calculated using, for example, a weighted average method. For example, for the initial tack factor, the weight value in the first weighting condition and the weight value corresponding to the proportion of initial tack in the second weighting condition are calculated to obtain a weight value that comprehensively considers both theory and practice. Finally, by performing this calculation on each adhesion performance factor, a new set of weight values ​​is obtained. This set of weight values ​​is the third weight condition for adhesion performance testing. It combines theory and practice and can more accurately reflect the weight of each adhesion performance factor in the final adhesion performance evaluation. This provides a more reasonable basis for subsequent weighted calculation of adhesion performance test results based on this weight condition.

[0041] The third weighting condition for adhesion performance testing is a comprehensive weighting value obtained by considering both the theoretical initial weighting (first weighting condition) and the weighting of actual test results (second weighting condition). This value accurately reflects the weight allocation of each adhesion performance factor (such as initial tack, holding tack, and adhesion migration) in the final adhesion performance evaluation. Then, the adhesion performance test results are weighted and calculated, including the initial tack evaluation coefficient, holding tack evaluation coefficient, and adhesion migration evaluation coefficient. For the initial tack evaluation coefficient, it is multiplied by the weight value corresponding to initial tack in the third weighting condition; for the holding tack evaluation coefficient, it is multiplied by the weight value corresponding to holding tack; and for the adhesion migration evaluation coefficient, it is multiplied by the weight value corresponding to adhesion migration. Finally, the results of the above weighted calculations are summed, and the total is the label adhesive quality coefficient. This coefficient comprehensively considers the various adhesion performance factors of the adhesive layer and their weights, accurately reflecting the quality status of the adhesive layer and providing an important basis for the overall quality evaluation of self-adhesive labels.

[0042] In one possible implementation, step S310 further includes:

[0043] Step S311: Based on the multidimensional adhesion performance detection factor, generate an initial tack test command, and based on the initial tack test command, activate the initial tack tester.

[0044] Step S312: Based on the adhesive layer, perform multiple initial tack tests using the initial tack tester to obtain multiple initial tack test data.

[0045] Step S313: Calculate the aggregate value based on the multiple initial tack test data to obtain the initial tack test result.

[0046] Step S314: Perform deep learning based on the initial tack test result record set and the initial tack test evaluation record set to build an initial tack test evaluation model.

[0047] Step S315: Based on the initial tack test results, output the initial tack test evaluation coefficient according to the initial tack test evaluation model, and add the initial tack test evaluation coefficient to the adhesion performance test results.

[0048] Specifically, the initial step in testing the initial tack of an adhesive layer based on multidimensional adhesion performance testing factors is to generate an initial tack test instruction. The multidimensional adhesion performance testing factors include relevant requirements and parameter settings for initial tack testing. Based on these settings, an instruction specifically for initial tack testing is generated. This instruction is the key signal to initiate the initial tack testing process; it activates the initial tack tester, preparing it for subsequent testing operations.

[0049] To obtain more accurate initial tack information for the adhesive layer, multiple initial tack tests were conducted on the adhesive layer using an initial tack tester. During this process, multiple samples with the same production parameters as the adhesive layer were specifically used. This is because the initial tack of the adhesive layer can be affected by various factors, and using samples with identical production parameters minimizes interference from other factors, ensuring that the test results accurately reflect the initial tack characteristics of the adhesive layer. Each test of a sample using the initial tack tester yielded one initial tack data point. Due to multiple tests, multiple initial tack data points were ultimately obtained, which will serve as important evidence for subsequent analysis and evaluation of the adhesive layer's initial tack.

[0050] The purpose of ensemble value calculation for multiple initial tack test data is to extract a result that can represent the overall initial tack characteristics from a large number of data points. The ensemble value calculation involves adding up the multiple initial tack test data points and then dividing by the total number of data points. The resulting average reflects the overall level of these data points. The advantage of this approach is that it comprehensively considers all test data and reduces the impact of individual outliers on the results. The initial tack test results obtained through this ensemble value calculation can serve as an important basis for subsequent evaluation and analysis, providing strong support for accurately determining the initial tack performance of the adhesive layer.

[0051] The initial viscosity detection result set is used as input data, and the initial viscosity detection evaluation set is used as the expected output data. The data is preprocessed, such as by normalization, to ensure it falls within a suitable range for better training. Then, a suitable BPNN structure is designed, determining the number of input layer nodes to match the number of features in the initial viscosity detection result set, and the number of output layer nodes to match the number of categories in the initial viscosity evaluation or the specific evaluation value. The number of hidden layers is determined based on experience and experimentation, generally starting with a smaller number of nodes and gradually increasing to find the optimal structure. Next, the BPNN is trained using the training data. During training, data enters the network from the input layer, undergoes computation and propagation through each layer, and generates prediction results at the output layer. The prediction results are compared with the expected output, the error is calculated, and then the error is propagated back from the output layer to the input layer using a backpropagation algorithm. The weights and biases of each layer in the network are adjusted based on the error. This process is repeated until the error reaches an acceptable range or the preset number of training iterations is reached. After training, the resulting BPNN model can quickly and accurately output the initial tack detection evaluation coefficient based on the new initial tack detection results. This model can automatically learn the complex relationship between the initial tack detection results and the evaluation, providing an effective method for the evaluation of initial tack.

[0052] Based on the obtained initial tack test results, they are input into the initial tack test evaluation model. This model is built using deep learning on the initial tack test result record set and the initial tack test evaluation record set, enabling accurate evaluation based on the input initial tack test results. After receiving the initial tack test results, the model performs internal calculations and analysis, outputting an initial tack test evaluation coefficient. This coefficient is a quantitative evaluation of the initial tack test results, reflecting the performance of the adhesive layer in terms of initial tack. For example, a higher evaluation coefficient indicates good initial tack, while a lower coefficient indicates problems with initial tack. Finally, this initial tack test evaluation coefficient is added to the adhesion performance test results. The adhesion performance test results are a comprehensive consideration of the overall adhesion performance of the adhesive layer, including multiple aspects such as initial tack, holding tack, and adhesion migration. Adding the initial tack test evaluation coefficient makes the adhesion performance test results more comprehensive and accurate, providing richer information for subsequent comprehensive analysis and judgment of the adhesive layer's adhesion performance.

[0053] In one possible implementation, step S310 further includes:

[0054] Step S316: Based on the multidimensional adhesion performance detection factor, generate a holding power detection command, and activate the holding power tester based on the holding power detection command; based on the adhesive layer, perform a holding power confidence test according to the holding power tester to obtain the holding power test result; use the holding power test result record set as input information and the holding power detection evaluation record set as output information to train the residual neural network, and obtain the holding power detection evaluation loss coefficient after each predetermined number of training iterations; when the holding power detection evaluation loss coefficient is less than the holding power detection evaluation loss threshold, generate a holding power detection evaluation channel; based on the holding power test result, obtain the holding power detection evaluation coefficient according to the holding power detection evaluation channel, and add the holding power detection evaluation coefficient to the adhesion performance detection result.

[0055] Specifically, the adhesive layer of the self-adhesive label is tested for tackiness. Based on the tackiness-related requirements in the multidimensional adhesion performance testing factors, a tackiness test command is generated. The multidimensional adhesion performance testing factors comprehensively consider multiple aspects such as initial tack, tackiness, and adhesion migration, among which tackiness is crucial for evaluating the adhesive performance of the adhesive layer during long-term use. The generated tackiness test command acts as a specific signal to activate the tackiness tester, a device specifically designed to measure the tackiness of adhesive layers. Upon receiving the tackiness test command, the tackiness tester begins operation, preparing to perform a tackiness test on the adhesive layer of the self-adhesive label.

[0056] Multiple samples with the same production parameters as the adhesive layer were tested. This is because the performance of the adhesive layer is affected by various factors in actual production, and using samples with the same production parameters minimizes the interference of other factors, ensuring that the test results accurately reflect the tack properties of the adhesive layer. Multiple tests are conducted to improve the accuracy and reliability of the results. Since tack properties fluctuate, a single test result cannot fully reflect the tack performance of the adhesive layer. Multiple tests obtain more data, allowing for a more accurate assessment of the adhesive layer's tack. After multiple tests, the mean value is taken as the tack test result. The mean value is used to reduce the impact of individual outliers, making the test results more stable and reliable. Through this testing process, a test result that accurately reflects the tack of the adhesive layer is obtained, providing an important basis for subsequent evaluation and analysis.

[0057] The tackiness test result record set is used as input information. This record set contains various data obtained from previous tackiness tests, such as tackiness test results under different conditions and results obtained from tests using different samples. This data reflects the actual performance of the adhesive layer in terms of tackiness. Simultaneously, the tackiness detection and evaluation record set is used as output information. This record set contains information after evaluating the tackiness test results. Then, a residual neural network is used for training. Residual neural networks have strong learning and generalization abilities and can effectively handle complex nonlinear relationships. During training, the network continuously adjusts its weights and parameters to make the output result as close as possible to the expected output in the tackiness detection and evaluation record set. At each predetermined number of training iterations, the tackiness detection and evaluation loss coefficient is obtained. This loss coefficient measures the difference between the current network output and the expected output. The smaller the loss coefficient, the closer the network's prediction result is to the actual evaluation result. By continuously monitoring the loss coefficient, the training progress and performance of the network can be understood. If the loss coefficient gradually decreases and reaches a stable low value, it indicates that the network has learned the relationship between the tackiness test results and the evaluation well. Through this training process, a model can be obtained that can accurately assess the tackiness of the adhesive layer, providing strong support for the quality monitoring of self-adhesive labels.

[0058] When the holding power detection and evaluation loss coefficient is less than the holding power detection and evaluation loss threshold, it means that after training the residual neural network, the gap between the network's evaluation result of holding power and the actual expected evaluation result is small enough to be acceptable. At this point, a holding power detection and evaluation channel is generated. This channel can be understood as a path or mechanism specifically designed for accurate holding power evaluation. Through this channel, new holding power test results are quickly and accurately converted into holding power detection and evaluation coefficients. Once this evaluation channel is generated, subsequent holding power evaluations of self-adhesive labels can directly utilize this channel without undergoing a complex neural network training process again. Simply input the new holding power test results into this channel to quickly obtain the corresponding evaluation coefficients, which are then added to the adhesion performance test results, providing an important basis for comprehensively evaluating the quality of self-adhesive labels.

[0059] Based on the obtained tackiness test results, a tackiness evaluation coefficient is determined using a tackiness detection evaluation channel. The tackiness test results are data obtained through tackiness confidence tests on the adhesive layer, reflecting the adhesive's ability to maintain tack under certain conditions. The tackiness detection evaluation channel is generated when the tackiness evaluation loss coefficient is less than a threshold. It is a trained and optimized evaluation mechanism capable of accurately providing evaluation coefficients based on the tackiness test results. The tackiness test results are analyzed and processed through the tackiness detection evaluation channel to obtain the tackiness evaluation coefficient. This coefficient is a quantitative assessment of the adhesive layer's tackiness, reflecting the degree to which the adhesive layer's tackiness falls within a specific standard or expected range. Finally, the tackiness evaluation coefficient is added to the adhesion performance test results. The adhesion performance test results are a comprehensive consideration of the overall adhesion performance of the adhesive layer, including initial tack, tackiness, and adhesion migration. Including the tackiness evaluation coefficient makes the adhesion performance test results more comprehensive and accurate, providing richer information for subsequent quality assessment of self-adhesive labels.

[0060] In one possible implementation, step S310 further includes:

[0061] Step S317: Based on the multidimensional adhesion performance detection factors, generate adhesion migration detection instructions, and activate adhesion migration detection scenario factors according to the adhesion migration detection instructions, wherein the adhesion migration detection scenario factors include adhesion target, adhesion duration, test environment, and applied pressure; construct K adhesion migration detection scenarios according to the adhesion migration detection scenario factors, where K is a positive integer greater than 1; perform adhesion migration tests based on the adhesive layer according to the K adhesion migration detection scenarios to obtain K migration adhesive amounts; perform adhesion migration evaluation based on the K adhesion migration detection scenarios and the K migration adhesive amounts to obtain K adhesion migration evaluation coefficients; perform lumped value calculation based on the K adhesion migration evaluation coefficients to generate adhesion migration detection evaluation coefficients, and add the adhesion migration detection evaluation coefficients to the adhesion performance detection results.

[0062] Specifically, adhesion migration detection instructions are generated based on multi-dimensional adhesion performance detection factors, which cover aspects such as initial tack, holding tack, and adhesion migration. When the adhesion migration of an adhesive layer needs to be detected, specific detection instructions are generated based on these factors. The function of this detection instruction is to initiate the adhesion migration detection process. Once generated, it triggers the activation of adhesion migration detection scenario factors. These factors include the adhesion target, adhesion duration, test environment, and applied pressure. These factors are crucial for accurately evaluating the adhesion migration performance of the adhesive layer. The adhesion target refers to the object to which the adhesive layer adheres, which can be an object of different materials and surface properties. Different adhesion targets will have different effects on adhesion migration. The adhesion duration determines the observation of adhesion migration within a certain time range; a longer adhesion duration will lead to more adhesion migration phenomena. The test environment includes factors such as temperature, humidity, and light, which affect the performance of the adhesive and thus the degree of adhesion migration. The applied pressure simulates the stress conditions experienced by the adhesive layer in actual use, and different applied pressures will have different effects on adhesion and migration. By activating these scenario factors, specific detection conditions and parameters can be provided for subsequent adhesion and migration tests, ensuring the accuracy and reliability of the test results.

[0063] Adhesion migration detection scenario factors include important parameters such as adhesion target, adhesion duration, test environment, and applied pressure. Different combinations of these parameters can create various detection conditions to comprehensively evaluate the adhesion migration performance of the adhesive layer. Setting K to a positive integer greater than 1 means building multiple different detection scenarios. This is done to test the adhesion migration of the adhesive layer from multiple perspectives and under multiple conditions, avoiding the limitations that a single detection scenario may bring.

[0064] Before conducting the adhesion migration test, the initial weights of each sample and its corresponding adhesion target are accurately measured and denoted as W1 (initial sample weight) and W2 (initial adhesion target weight), respectively. Then, the samples are subjected to an adhesion test under a specific adhesion migration testing scenario (e.g., applying pressure to an adhesion target of a specific material and maintaining adhesion for a specific duration under specific temperature and humidity conditions). After the test, the weights of the sample and the adhesion target are measured again and denoted as W3 (sample weight after test) and W4 (adhesion target weight after test), respectively. The amount of migrated adhesive is then calculated using the following formula: Migrated adhesive amount = (W4 - W2) - (W1 - W3). This method determines the amount of adhesive that migrates from the adhesive layer to the adhesion target by measuring the change in weight, making it relatively intuitive and accurate.

[0065] When conducting adhesion migration assessments to obtain K adhesion migration evaluation coefficients, K adhesion migration detection scenarios and their corresponding K migration adhesive amounts are comprehensively considered. For each detection scenario, analysis is performed from multiple aspects. First, the migration adhesive amount is compared with a preset standard value. If the migration adhesive amount is low, it indicates that the adhesion migration degree of the adhesive in that scenario is small, and its adhesion stability is good, thus a higher evaluation coefficient is assigned. Conversely, if the migration adhesive amount is high, it indicates that the adhesion migration is significant, and the adhesion stability of the adhesive is problematic, thus the evaluation coefficient will decrease. Simultaneously, the variation of the migration adhesive amount is monitored. If the variation of the migration adhesive amount is relatively stable at different times or stages within the same detection scenario, it indicates that the adhesion migration performance of the adhesive has good consistency and reliability, which will lead to an increase in the evaluation coefficient. However, if the migration adhesive amount fluctuates greatly, it means that the performance of the adhesive in that scenario is not stable enough and is easily affected by external factors, thus the evaluation coefficient will decrease accordingly. Furthermore, other factors in the testing scenario cannot be ignored, such as the material and surface characteristics of the adhesive target, adhesion duration, and the testing environment (including temperature, humidity, and pressure). Different adhesive targets have different effects on adhesive migration, the duration of adhesion also affects the amount of migrated adhesive, and changes in the testing environment can accelerate or slow down the adhesive migration process. By comprehensively analyzing these factors, an accurate adhesive migration evaluation coefficient is determined for each adhesive migration testing scenario, ultimately obtaining K evaluation coefficients. These coefficients will provide important basis for subsequent product quality evaluation and process improvement.

[0066] A pooled value calculation is performed based on K adhesion migration evaluation coefficients to generate an adhesion migration detection evaluation coefficient. The pooled value calculation uses the average method to extract an index that can represent the overall adhesion migration performance from multiple evaluation coefficients. This adhesion migration detection evaluation coefficient is then added to the adhesion performance test results, which consider the comprehensive performance of the adhesive layer, including initial tack, holding tack, and adhesion migration. Adding the adhesion migration detection evaluation coefficient makes the adhesion performance test results more comprehensive and accurate, providing richer information for subsequent quality assessment of self-adhesive labels.

[0067] In one possible implementation, step S400 further includes:

[0068] Step S410: Perform anti-sticking property testing on the base paper layer to obtain the anti-sticking property coefficient of the base paper.

[0069] Step S420: Perform strength characteristic testing on the base paper layer to obtain the base paper strength characteristic coefficient.

[0070] Step S430: Normalize the base paper anti-stick characteristic coefficient and the base paper strength characteristic coefficient to obtain the test results of the base paper's two-sided characteristics.

[0071] Step S440: The backing paper bilateral characteristic detection results are weighted and calculated based on the backing paper bilateral characteristic detection weights to generate the label backing paper quality coefficient, wherein the backing paper bilateral characteristic detection weights include a predetermined weight for anti-sticking characteristics and a predetermined weight for strength characteristics.

[0072] Specifically, the anti-stick properties of the backing paper layer are tested, as these properties directly affect the ease of use and reliability of self-adhesive labels in practical applications. To accurately obtain the anti-stick property coefficient of the backing paper, a series of testing methods and techniques are employed. First, the actual label peeling process is simulated. The self-adhesive label is adhered to the backing paper layer, and then a professional peeling device is used to peel the label off the backing paper at a certain speed and angle. During this process, the force required to peel the label is measured. If the peeling force is small, it indicates a good anti-stick effect between the backing paper layer and the adhesive, and the anti-stick property coefficient of the backing paper will be correspondingly high; conversely, if the peeling force is large, it means that the anti-stick performance of the backing paper is poor, and the anti-stick property coefficient of the backing paper will be low. In addition, the surface condition of the backing paper after the label is peeled off is observed. If the backing paper surface is clean and there is no adhesive residue, this also indicates good anti-stick properties; while if there is a lot of adhesive residue on the backing paper surface, it indicates that the anti-stick performance of the backing paper has problems. Simultaneously, multiple repeated tests were conducted to obtain more accurate and reliable data. Statistical analysis of the test results was performed to calculate the mean and standard deviation, providing a more comprehensive evaluation of the anti-stick properties of the backing paper layer. Finally, based on the results of these tests and analyses, an anti-stick property coefficient for the backing paper was determined. This coefficient quantitatively reflects the quality of the anti-stick performance of the backing paper layer, providing an important reference for subsequent backing paper quality assessment and product application.

[0073] The strength characteristics of the backing paper have a significant impact on the stability and reliability of self-adhesive labels during production, storage, transportation, and use. To obtain the strength characteristic coefficient of the backing paper, its strength is tested from multiple aspects. Tensile strength is a crucial indicator. Using a specialized material tensile testing machine, the backing paper sample is fixed on the machine and stretched along a specific direction at a certain tensile speed until it breaks. During this process, the testing machine records the maximum tensile force the backing paper withstands. Based on this maximum tensile force and parameters such as the cross-sectional area of ​​the backing paper, the tensile strength value is obtained. A higher tensile strength value indicates a stronger resistance to tensile deformation, and consequently, a higher strength characteristic coefficient. Besides tensile strength, the tear strength of the backing paper is also important, reflecting its ability to resist tearing under localized external forces. The Elmendorf tear test is used to test the backing paper. In this test, a small slit is cut into a sample of the backing paper according to a specified method, and then the paper is torn at a certain speed. The force required to tear the paper is measured; the greater the tear force, the higher the tear strength of the backing paper, and its strength characteristic coefficient will be correspondingly higher. In addition, the bending strength of the backing paper is tested. The sample is placed on a bending test apparatus, a certain bending moment is applied, and the deformation and damage of the paper during bending are observed. Backing paper with high bending strength is less likely to break or deform under bending force, which also reflects its better strength characteristics. Multiple repeated tests are required during strength characteristic testing to reduce testing errors and improve the accuracy and reliability of the test results. The test results of tensile strength, tear strength, and bending strength are comprehensively considered, and the data analysis and processing methods are used to finally determine the strength characteristic coefficient of the backing paper. This coefficient can comprehensively and accurately reflect the strength characteristics of the backing paper layer, providing an important basis for evaluating the overall quality of self-adhesive labels.

[0074] The anti-stick property coefficient and the strength property coefficient of the backing paper are normalized to ensure that the two coefficients are in the same numerical range, between [0, 1]. After normalization, the test results of the two-sided properties of the backing paper comprehensively reflect the performance of the backing paper in both anti-stick and strength properties.

[0075] The weighting of the backing paper's bilateral characteristic tests includes predetermined weights for anti-stick properties and strength properties. These weights are pre-set based on actual needs and importance, adjusting the contribution ratio of anti-stick and strength properties in the overall evaluation. For example, if the anti-stick properties of the backing paper are more important in a specific application scenario, then a higher weight is assigned to anti-stick properties; conversely, if strength properties are more critical, then the weight of strength properties is increased. By weighting the backing paper's bilateral characteristic test results with the corresponding weights, a label backing paper quality coefficient is obtained. This coefficient comprehensively considers both the anti-stick and strength properties of the backing paper, more accurately reflecting the quality level of the backing paper and providing an important basis for the overall quality assessment of self-adhesive labels.

[0076] In one possible implementation, step S410 further includes:

[0077] Step S411: Obtain multiple base paper samples, wherein the multiple base paper samples have the same production parameters as the base paper layer.

[0078] Step S412: Using a peel force testing machine, perform peel force tests on the multiple backing paper samples respectively to obtain multiple test peel forces.

[0079] Step S413: Calculate the concentrated value based on the multiple test peel forces to obtain the peel force test results.

[0080] Step S414: Activate the pre-built anti-stick property detection model.

[0081] Step S415: Input the peel force test result into the anti-stick property detection model and output the anti-stick property coefficient of the base paper.

[0082] Specifically, multiple backing paper samples are obtained, which have the same production parameters as the actual backing paper layer used. This is done to ensure that the test results can accurately reflect the true performance of the backing paper layer, as the same production parameters minimize the interference of other factors.

[0083] Peel force tests were performed on multiple backing paper samples using a peel force tester. The peel force tester accurately measures the force required to peel a label from the backing paper. During the testing process, standard testing methods and operating procedures were strictly followed to ensure consistent testing conditions for each backing paper sample. Through this testing, multiple peel force values ​​were obtained.

[0084] The concentrated value is calculated based on the peel force from multiple tests. The concentrated value calculation adopts the average value method. By calculating the concentrated value, the influence of the deviation of individual test data on the result is reduced, thereby obtaining the peel force test result that is more representative of the overall situation.

[0085] Activate the pre-built anti-stick property detection model, which has been trained and optimized with a large amount of data and can accurately evaluate the anti-stick properties of the backing paper based on the input peel force data.

[0086] The peel force test results obtained earlier are input into the anti-stick property detection model. After internal calculation and analysis, the model outputs the anti-stick property coefficient of the backing paper. This coefficient quantitatively reflects the anti-stick performance of the backing paper, thereby providing a better understanding of the backing paper's performance and offering strong support for the overall quality control of self-adhesive labels.

[0087] Example 2, based on the same inventive concept as the image analysis-based self-adhesive label production quality monitoring method in the foregoing examples, such as... Figure 2 As shown, this application provides an image analysis-based self-adhesive label production quality monitoring system. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0088] Self-adhesive label acquisition module 10 is used to acquire self-adhesive labels, wherein the self-adhesive label includes a face material layer, an adhesive layer and a backing paper layer.

[0089] The label surface material quality coefficient determination module 20 is used to perform image analysis based on the surface material layer monitoring image of the surface material layer to determine the label surface material quality coefficient.

[0090] The label adhesive quality coefficient acquisition module 30 is used to perform comprehensive adhesive performance testing on the adhesive layer based on multidimensional adhesive performance testing factors to obtain the label adhesive quality coefficient. The multidimensional adhesive performance testing factors include initial tack, holding tack and adhesive migration.

[0091] The label backing paper quality coefficient acquisition module 40 is used to perform bilateral characteristic detection and integration on the backing paper layer according to the bilateral characteristic detection weight of the backing paper to obtain the label backing paper quality coefficient.

[0092] The label quality expectation condition setting module 50 sets label quality expectation conditions based on the self-adhesive label, wherein the label quality expectation conditions include the expected face material quality coefficient, the expected label adhesive quality coefficient, and the expected label backing paper quality coefficient.

[0093] The self-adhesive label quality inspection result output module 60, based on the label quality expectation conditions, performs expectation compliance judgment on the label face material quality coefficient, the label adhesive quality coefficient and the label backing paper quality coefficient, and outputs the self-adhesive label quality inspection result.

[0094] Furthermore, the label adhesive quality coefficient acquisition module 30 also includes:

[0095] An adhesion performance test result acquisition unit is used to perform adhesion performance testing on the adhesive layer according to the multidimensional adhesion performance test factor, and obtain the adhesion performance test result.

[0096] The first weight condition determination unit is used to initialize weight settings based on the multidimensional adhesion performance detection factor and determine the first weight condition for adhesion performance detection.

[0097] The second weighting condition determination unit is used to calculate the proportion based on the adhesion performance test results to obtain the second weighting condition for adhesion performance test.

[0098] The third weight condition determination unit calculates the concentrated value of each dimension of the weight based on the first weight condition and the second weight condition of the adhesion performance detection to obtain the third weight condition of the adhesion performance detection.

[0099] The weighted calculation unit is used to perform weighted calculation on the adhesion performance test results according to the third weighting condition of the adhesion performance test, and output the label adhesive quality coefficient.

[0100] Furthermore, the adhesion performance test result acquisition unit also includes:

[0101] The initial tack tester activation unit generates an initial tack test command based on the multidimensional adhesion performance detection factor, and activates the initial tack tester based on the initial tack test command.

[0102] The initial tack test data acquisition unit, based on the adhesive layer, performs multiple initial tack tests using the initial tack tester to obtain multiple initial tack test data.

[0103] The initial tack test result acquisition unit is used to perform a set value calculation based on the multiple initial tack test data to obtain the initial tack test result.

[0104] The initial tack detection and evaluation model building unit is used to build an initial tack detection and evaluation model by performing deep learning based on the initial tack detection result record set and the initial tack detection evaluation record set.

[0105] The initial tack test evaluation coefficient output unit outputs an initial tack test evaluation coefficient based on the initial tack test result and according to the initial tack test evaluation model, and adds the initial tack test evaluation coefficient to the adhesion performance test result.

[0106] Furthermore, the adhesion performance test result acquisition unit also includes:

[0107] The holding power tester activation unit generates a holding power test command based on the multidimensional adhesion performance detection factor, and activates the holding power tester based on the holding power test command.

[0108] The tack test result acquisition unit obtains the tack test result by performing a tack confidence test on the adhesive layer using the tack tester.

[0109] The stickiness detection and evaluation loss coefficient acquisition unit is used to train the residual neural network with the stickiness test result record set as input information and the stickiness detection and evaluation record set as output information, and to acquire the stickiness detection and evaluation loss coefficient after each predetermined number of training iterations.

[0110] A tackiness detection and evaluation channel generation unit is used to generate a tackiness detection and evaluation channel when the tackiness detection and evaluation loss coefficient is less than the tackiness detection and evaluation loss threshold.

[0111] The tackiness test evaluation coefficient acquisition unit obtains the tackiness test evaluation coefficient based on the tackiness test results and according to the tackiness test evaluation channel, and adds the tackiness test evaluation coefficient to the adhesion performance test results.

[0112] Furthermore, the adhesion performance test result acquisition unit also includes:

[0113] An adhesion migration detection scenario factor activation unit generates an adhesion migration detection instruction based on the multidimensional adhesion performance detection factor, and activates the adhesion migration detection scenario factor according to the adhesion migration detection instruction. The adhesion migration detection scenario factor includes the adhesion target, adhesion duration, test environment, and applied pressure.

[0114] An adhesion migration detection scenario building unit is used to build K adhesion migration detection scenarios based on the adhesion migration detection scenario factor, where K is a positive integer greater than 1.

[0115] The migration adhesive amount acquisition unit, based on the adhesive layer, performs adhesion migration tests according to the K adhesion migration detection scenarios to obtain K migration adhesive amounts.

[0116] An adhesion migration evaluation coefficient acquisition unit performs adhesion migration evaluation based on the K adhesion migration detection scenarios and the K migration adhesive amounts to obtain K adhesion migration evaluation coefficients.

[0117] An adhesion migration detection evaluation coefficient generation unit is used to calculate the aggregate value based on the K adhesion migration evaluation coefficients, generate adhesion migration detection evaluation coefficients, and add the adhesion migration detection evaluation coefficients to the adhesion performance detection results.

[0118] Furthermore, the label backing paper quality coefficient acquisition module 40 also includes:

[0119] A backing paper anti-stick characteristic coefficient acquisition unit is used to detect the anti-stick characteristics of the backing paper layer and obtain the backing paper anti-stick characteristic coefficient.

[0120] A backing paper strength characteristic coefficient acquisition unit is used to perform strength characteristic detection on the backing paper layer to obtain the backing paper strength characteristic coefficient.

[0121] The backing paper bilateral characteristic test result acquisition unit is used to perform normalization processing based on the backing paper anti-stick characteristic coefficient and the backing paper strength characteristic coefficient to obtain the backing paper bilateral characteristic test result.

[0122] The label backing paper quality coefficient generation unit performs weighted calculations on the bilateral characteristic detection results of the backing paper based on the bilateral characteristic detection weights of the backing paper to generate the label backing paper quality coefficient. The bilateral characteristic detection weights of the backing paper include predetermined weights for anti-stick characteristics and predetermined weights for strength characteristics.

[0123] Furthermore, the unit for obtaining the anti-stick property coefficient of the base paper also includes:

[0124] A backing paper sample acquisition unit is used to acquire multiple backing paper samples, wherein the multiple backing paper samples have the same production parameters as the backing paper layer.

[0125] A peel force acquisition unit is used to perform peel force tests on the multiple backing paper samples according to a peel force testing machine to obtain multiple test peel forces.

[0126] A peel force test result acquisition unit is used to calculate the concentrated value based on the multiple test peel forces to obtain the peel force test result.

[0127] A detection model activation unit is used to activate a pre-built anti-stick property detection model.

[0128] The paper backing anti-stick characteristic coefficient output unit is used to input the peel force test result into the anti-stick characteristic detection model and output the paper backing anti-stick characteristic coefficient.

[0129] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0130] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0131] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for monitoring the production quality of self-adhesive labels based on image analysis, characterized in that, The method includes: Obtain a self-adhesive label, wherein the self-adhesive label includes a face layer, an adhesive layer, and a backing paper layer; Image analysis is performed on the surface material monitoring images to determine the label surface material quality coefficient; The adhesive layer is subjected to comprehensive adhesion performance testing based on multidimensional adhesion performance testing factors to obtain the label adhesive quality coefficient. The multidimensional adhesion performance testing factors include initial tack, holding tack and adhesion migration. Based on the weighting of the bilateral characteristic detection of the backing paper, the bilateral characteristic detection of the backing paper layer is integrated to obtain the label backing paper quality coefficient; Based on the self-adhesive label, label quality expectation conditions are set, wherein the label quality expectation conditions include the expected face material quality coefficient, the expected label adhesive quality coefficient, and the expected label backing paper quality coefficient; Based on the label quality expectation conditions, the label face material quality coefficient, the label adhesive quality coefficient, and the label backing paper quality coefficient are judged to meet the expectations, and the self-adhesive label quality test results are output. The adhesive layer is subjected to comprehensive adhesion performance testing based on multidimensional adhesion performance testing factors to obtain the label adhesive quality coefficient, including: The adhesive layer is tested for adhesion performance based on the multidimensional adhesion performance testing factors to obtain the adhesion performance test results; The initial weights are set according to the multidimensional adhesion performance detection factors to determine the first weight condition for adhesion performance detection. Based on the adhesion performance test results, the proportion is calculated to obtain the second weighting condition for adhesion performance test. Based on the first weight condition and the second weight condition for adhesion performance detection, the concentrated value of each dimension weight is calculated to obtain the third weight condition for adhesion performance detection. The adhesion performance test results are weighted according to the third weighting condition of the adhesion performance test, and the label adhesive quality coefficient is output. The adhesive layer is tested for adhesion performance based on the multidimensional adhesion performance testing factors to obtain adhesion performance test results, including: Based on the multidimensional adhesion performance detection factors, an initial tack test command is generated, and based on the initial tack test command, the initial tack tester is activated. Based on the adhesive layer, multiple initial tack tests are performed using the initial tack tester to obtain multiple initial tack test data. The initial tack test result is obtained by calculating the pooled value based on the multiple initial tack test data. Deep learning is used to build an initial tack test and evaluation model based on the initial tack test result record set and the initial tack test evaluation record set. Based on the initial tack test results, according to the initial tack test evaluation model, the initial tack test evaluation coefficient is output, and the initial tack test evaluation coefficient is added to the adhesion performance test results. The adhesive layer is tested for adhesion performance based on the multidimensional adhesion performance testing factors to obtain adhesion performance test results, including: Based on the multidimensional adhesion performance detection factors, a holding power detection command is generated, and based on the holding power detection command, the holding power tester is activated. Based on the adhesive layer, a holding confidence test is performed using the holding power tester to obtain the holding power test result; Using the stickiness test result record set as input information and the stickiness detection evaluation record set as output information, the residual neural network is trained, and the stickiness detection evaluation loss coefficient is obtained after each predetermined number of training iterations. When the holding strength detection and evaluation loss coefficient is less than the holding strength detection and evaluation loss threshold, a holding strength detection and evaluation channel is generated; Based on the holding properties test results, a holding properties test evaluation coefficient is obtained according to the holding properties test evaluation channel, and the holding properties test evaluation coefficient is added to the adhesion performance test results; The adhesive layer is tested for adhesion performance based on the multidimensional adhesion performance testing factors to obtain adhesion performance test results, including: Based on the multidimensional adhesion performance detection factors, an adhesion migration detection instruction is generated, and according to the adhesion migration detection instruction, an adhesion migration detection scenario factor is activated, wherein the adhesion migration detection scenario factor includes adhesion target, adhesion duration, test environment, and applied pressure. Based on the adhesion migration detection scenario factor, K adhesion migration detection scenarios are constructed, where K is a positive integer greater than 1; Based on the adhesive layer, adhesion migration tests are performed according to the K adhesion migration detection scenarios to obtain K migration adhesive amounts; Adhesion migration is evaluated based on the K adhesion migration detection scenarios and the K migration adhesive amounts to obtain K adhesion migration evaluation coefficients; Based on the K adhesion and migration evaluation coefficients, a central value is calculated to generate an adhesion and migration detection evaluation coefficient, and the adhesion and migration detection evaluation coefficient is added to the adhesion performance detection result.

2. The method as described in claim 1, characterized in that, Based on the weighted bilateral characteristic detection of the backing paper layer, the bilateral characteristic detection of the backing paper layer is integrated to obtain the label backing paper quality coefficient, including: The anti-stick properties of the base paper layer are tested to obtain the anti-stick property coefficient of the base paper. The strength characteristics of the base paper layer are tested to obtain the base paper strength characteristic coefficient. The base paper anti-stick characteristic coefficient and the base paper strength characteristic coefficient are normalized to obtain the test results of the base paper's two-sided characteristics. The quality coefficient of the label backing paper is generated by weighting the detection results of the two-sided characteristics of the backing paper based on the detection weights of the two-sided characteristics of the backing paper. The two-sided characteristics detection weights of the backing paper include predetermined weights for anti-stick characteristics and predetermined weights for strength characteristics.

3. The method as described in claim 2, characterized in that, Based on the anti-stick properties of the base paper layer, the anti-stick property coefficient of the base paper is obtained by testing, including: Multiple base paper samples are obtained, wherein the multiple base paper samples have the same production parameters as the base paper layer; Using a peel force testing machine, peel force tests were performed on the multiple backing paper samples to obtain multiple test peel forces. The peel force test results are obtained by calculating the concentrated value based on the multiple test peel forces. Activate the pre-built anti-stick property detection model; The peel force test results are input into the anti-stick property detection model, and the anti-stick property coefficient of the base paper is output.

4. A self-adhesive label production quality monitoring system based on image analysis, characterized in that, The system is used to perform the image analysis-based self-adhesive label production quality monitoring method according to any one of claims 1 to 3, the system comprising: A self-adhesive label acquisition module is used to acquire self-adhesive labels, wherein the self-adhesive label includes a face material layer, an adhesive layer, and a backing paper layer; A label surface material quality coefficient determination module is used to perform image analysis based on the surface material layer monitoring image to determine the label surface material quality coefficient. The label adhesive quality coefficient acquisition module is used to perform comprehensive adhesive performance testing on the adhesive layer based on multidimensional adhesive performance testing factors to obtain the label adhesive quality coefficient. The multidimensional adhesive performance testing factors include initial tack, holding tack and adhesive migration. A label backing paper quality coefficient acquisition module is used to perform bilateral characteristic detection and integration on the backing paper layer according to the bilateral characteristic detection weight of the backing paper to obtain the label backing paper quality coefficient. A label quality expectation condition setting module, which sets label quality expectation conditions based on the self-adhesive label, wherein the label quality expectation conditions include an expected face material quality coefficient, an expected label adhesive quality coefficient, and an expected label backing paper quality coefficient; The self-adhesive label quality inspection result output module, based on the label quality expectation conditions, performs expectation compliance judgment on the label face material quality coefficient, the label adhesive quality coefficient and the label backing paper quality coefficient, and outputs the self-adhesive label quality inspection result.

Images