Automated Intelligent Printing Quality Detection Method and System

Abstract

The present invention relates to the technical field of printing quality detection, and discloses an automated intelligent printing quality detection method and system. The present invention obtains the printing design data of a target printing object, analyzes the display effect characteristics of the target printing object based on the design data, generates an image detection scheme according to the display characteristics, acquires the detection image data of the target printing object according to the scheme, analyzes the detection image by using a pre-trained machine learning model, generates a quality evaluation report, and evaluates the quality of the printing object. This method accurately detects the printing quality in an automated manner, avoids the errors of manual detection, improves the detection efficiency and consistency, and can accurately identify printing defects and quality problems by combining the printing design data and the machine learning model, solving the problem of low efficiency of manual detection of printing quality in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of printing quality detection, and particularly to an automated intelligent printing quality detection method and system. Background Art

[0002] In traditional printing quality detection, manual judgment is easily affected by subjective factors and has low efficiency. With the continuous improvement of the requirements for production efficiency and product quality in the modern printing industry, the traditional manual detection method can no longer meet the growing demand. Summary of the Invention

[0003] The purpose of the present invention is to provide an automated intelligent printing quality detection method and system, aiming to solve the problem of low efficiency of manual detection of printing quality in the prior art.

[0004] The present invention is implemented as follows. In the first aspect, the present invention provides an automated intelligent printing quality detection method, including:

[0005] Obtain the printing design data of the target printing object, and perform feature analysis on the printing finished product display effect of the target printing object according to the printing design data to obtain the printing display features of the target printing object;

[0006] Analyze the image detection scheme for the target printing object according to the printing display features to obtain the image data detection scheme for the target printing object;

[0007] Collect the image data of the target printing object according to the image data detection scheme to obtain the detection image data of the target printing object;

[0008] Analyze and process the detection image data according to a pre-trained machine learning model to obtain a quality assessment report for the target printing object.

[0009] In the second aspect, the present invention provides an automated intelligent printing quality detection system for implementing the automated intelligent printing quality detection method according to any item in the first aspect, including:

[0010] A display analysis module for obtaining the printing design data of the target printing object and performing feature analysis on the printing finished product display effect of the target printing object according to the printing design data to obtain the printing display features of the target printing object;

[0011] A scheme analysis module for analyzing the image detection scheme for the target printing object according to the printing display features to obtain the image data detection scheme for the target printing object;

[0012] A detection execution module, configured to collect image data of the target printed object according to the image data detection scheme, so as to obtain the detection image data of the target printed object;

[0013] A quality evaluation module, configured to analyze and process the detection image data according to a pre-trained machine learning model, so as to obtain a quality evaluation report of the target printed object.

[0014] The present invention provides an automated intelligent printing quality detection method, which has the following beneficial effects:

[0015] The present invention obtains the printing design data of the target printed object, analyzes the display effect characteristics of the target printed object based on the design data, generates an image detection scheme according to the display characteristics, collects the detection image data of the target printed object according to the scheme, analyzes the detection image by using a pre-trained machine learning model, generates a quality evaluation report, and evaluates the quality of the printed object. This method accurately detects the printing quality in an automated manner, avoids the errors of manual detection, improves the detection efficiency and consistency, and can accurately identify printing defects and quality problems by combining the printing design data and the machine learning model, solving the problem of low efficiency of manual detection of printing quality in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a schematic diagram of the steps of an automated intelligent printing quality detection method provided by an embodiment of the present invention;

[0017] Figure 2 is a schematic diagram of the structure of an automated intelligent printing quality detection system provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0019] The implementation of the present invention will be described in detail below with reference to specific embodiments.

[0020] Referring to Figure 1 、 Figure 2 shown, a preferred embodiment is provided by the present invention.

[0021] In a first aspect, the present invention provides an automated intelligent printing quality detection method, including:

[0022] S1: Obtain the printing design data of the target printing object, and perform feature analysis on the display effect of the printed finished product of the target printing object according to the printing design data, so as to obtain the printing display features of the target printing object;

[0023] S2: Analyze the image detection scheme for the target printing object according to the printing display features, so as to obtain the image data detection scheme for the target printing object;

[0024] S3: Collect the image data of the target printing object according to the image data detection scheme, so as to obtain the detected image data of the target printing object;

[0025] S4: Analyze and process the detected image data according to a pre-trained machine learning model, so as to obtain the quality evaluation report of the target printing object.

[0026] Specifically, in step S1 of the embodiment provided by the present invention, the printing design data of the target printing object is obtained. The target printing object is the finished product to be prepared through printing work and is also the object to be subjected to printing quality detection work.

[0027] More specifically, there are differences in the printing content and printing form of different target printing objects. Therefore, different detection schemes need to be used for different target printing objects to achieve the best detection effect.

[0028] It should be noted that in the technical solution provided by the present invention, the image data to be detected is obtained by collecting images of the target printing object, and these image data are recognized to perform quality detection on the target printing object. Under different optical environments, the image content displayed by the target printing object is different. That is to say, after applying different optical environments to the target printing object and collecting images of the target printing object in different optical environments, the collected images are different, and the different images displayed by the target printing object will be more conducive to certain types of defect detection. That is, some defects that are difficult to display by the target printing object under normal optical environments are more easily recognized under specific optical environments.

[0029] Furthermore, the printed content on the target printing object is different from the printing form, and the effects shown by target printing objects with different printed contents and printing forms are also different in different optical environments. This means that different target printing objects are more likely to expose printing quality defects in different optical environments. To obtain this information, it is necessary to acquire the printing design data of the target printing object, and then perform feature analysis on the display effect of the printed product of the target printing object based on the printing design data to obtain the printing display features of the target printing object, that is, what kind of display effect the target printing object will show in different optical environments, so as to provide information for subsequent steps to determine what kind of optical environment filter should be used to detect the target printing object.

[0030] It should be noted that the optical environment filter refers to the specific optical environment applied to the target printing object during image acquisition of the target printing object. The specific optical environment includes lighting effects such as light intensity, light frequency, and light color.

[0031] Specifically, in step S2 of the embodiment provided by the present invention, analyze the image detection scheme of the target printing object according to the printing display features to obtain the image data detection scheme of the target printing object.

[0032] More specifically, the printing display features of the target printing object reflect the display information of the target printing object under different optical environment filters. By analyzing the display information, the detection adaptability between the display information and various defect types can be obtained, that is, the detection effect of the display information of the target printing object under the optical environment filter on various defect types of the target printing object.

[0033] More specifically, the printed contents at different positions on the target printing object are different, so the most suitable optical environment filters for different positions are different. Based on this consideration, different optical environment filters need to be applied to the target printing object multiple times to perform comprehensive and accurate defect detection on each position of the target printing object.

[0034] More specifically, by analyzing the printing display features of the target printing object, obtain the optical environment filters that can best display various defects of the target printing object, and convert these optical environment filters into the image data detection scheme of the target printing object, that is, perform image data acquisition on the target printing object with these specified optical environment filters to obtain the detection image data to be detected.

[0035] More specifically, since the most suitable defect types to be detected by each optical environment filter are different, in order to be comprehensive, it is necessary to combine multiple optical environment filters to obtain multiple optical environment filters with the most comprehensive detection to be used as the image data detection scheme.

[0036] Specifically, in step S3 of the embodiment provided by the present invention, different optical environment filters are sequentially applied to the target printed object according to the image data detection scheme, and then image data of the target printed object under different optical environment filters is collected to obtain the detection image data of the target printed object under different optical environment filters.

[0037] It can be understood that the detection image data of the target printed object under different optical environment filters has different detection adaptabilities to various defect types. For example, the difference within a certain range of the brightness of a certain type of color printing is difficult to identify in the normal optical environment, and by adjusting the brightness and color of the optical environment, the brightness of this type of color printing can be more easily reflected by the image data. Therefore, multiple detections need to be carried out under different optical environment filters to ensure the comprehensiveness and accuracy of quality detection.

[0038] Specifically, in step S4 of the embodiment provided by the present invention, after obtaining the detection image data, the pre-trained machine learning model is used to analyze and process the detection image data to obtain the quality assessment report of the target printed object.

[0039] It can be understood that using the machine learning model and image analysis technology can automatically detect problems in the printing process, avoid subjective biases in manual detection, improve detection efficiency and accuracy, reduce the time of manual operation and inspection through automated quality analysis and report generation, and at the same time reduce the error rate and save costs.

[0040] The present invention provides an automated intelligent printing quality detection method, which has the following beneficial effects:

[0041] The present invention obtains the printing design data of the target printed object, analyzes the display effect characteristics of the target printed object based on the design data, generates an image detection scheme according to the display characteristics, collects the detection image data of the target printed object according to the scheme, uses the pre-trained machine learning model to analyze the detection image, generates a quality assessment report, and evaluates the quality of the printed object. This method accurately detects the printing quality in an automated manner, avoids the errors of manual detection, improves the detection efficiency and consistency, and can accurately identify printing defects and quality problems by combining the printing design data and the machine learning model, solving the problem of low efficiency of manual detection of printing quality in the prior art.

[0042] Preferably, the steps of obtaining the printing design data of the target printed object and analyzing the characteristics of the printing finished product display effect of the target printed object according to the printing design data to obtain the printing display characteristics of the target printed object include:

[0043] S11: Obtain the printing design data of the target printing object; wherein, the target printing object includes a substrate part and a printing part, and the printing design data is used to describe the specific form that the printing part is pre-planned to display on the substrate part;

[0044] S12: Perform data analysis on the printing design data in terms of the morphology dimension and the color dimension to obtain the morphology data features and color data features of the target printing object;

[0045] S13: According to a plurality of preset optical environment filters, perform simulation analysis on the morphology data features and color data features of the target printing object to obtain the morphology display effect atlas and color display effect atlas of the target printing object corresponding to various optical environment filters;

[0046] S14: Combine the morphology display effect atlas and color display effect atlas of the target printing object corresponding to various optical environment filters to obtain the printing display features of the target printing object.

[0047] Specifically, obtain the printing design data of the target printing object. These data describe the specific form of the printing part on the target printing object (including the substrate and the printing part), such as design patterns, fonts, colors, and textures. The design data usually comes from CAD files, vector graphic files (such as AI, SVG files), or PDF and other formats. The design data is extracted through appropriate file parsing tools, and the information such as the predetermined position, shape, and color of the printing part on the substrate is stored structurally to ensure that the data can reflect the detailed design of the printed item.

[0048] More specifically, perform analysis on the obtained design data in terms of the morphology dimension and the color dimension to extract the morphology data features and color data features of the target printing object. The morphology dimension includes patterns, sizes, proportions, etc., and the color dimension includes hue, saturation, brightness, etc. Use image analysis and pattern recognition techniques to analyze the design data. For the morphology dimension, features such as the contour, shape, and size of the printing design can be extracted through geometric analysis (such as graphic segmentation, edge detection, etc.); for the color dimension, the specific features of the color can be obtained through color space conversion (such as RGB to CMYK or Lab color space conversion).

[0049] More specifically, according to a variety of preset optical environment filters, simulate and analyze the morphology data features and color data features of the target printing object, observe the changes in the printing effect by simulating different optical conditions, so as to evaluate the performance of the target printing object in different environments.

[0050] More specifically, by simulating different optical environment filters through lighting models (such as the Phong model or the Blinn-Phong model), rendering techniques in computer graphics can be used to simulate the surface reflection effects (such as highlights, diffuse reflections, refractions, etc.) of printed materials under different lighting conditions, and apply them to the topography and color characteristics of the target printed object to generate corresponding display effect atlases.

[0051] More specifically, based on the results of simulation analysis, topographic display effect atlases and color display effect atlases of the target printed object under various optical environment filters are generated. These atlases show the topography and color performance of the printed object under different optical environments, which can help designers evaluate whether the design meets the expected effect. Using image processing and data visualization techniques (such as heat maps, color difference maps, etc.), topographic display effect atlases and color display effect atlases are generated. These atlases can show the detailed changes under different optical conditions (such as surface glossiness, color changes, contrast changes, etc.).

[0052] More specifically, the topographic display effect atlases and color display effect atlases generated under different optical environment filters are combined with data to comprehensively evaluate the overall display characteristics of the target printed object. This process helps to identify potential problems in the design and evaluate the performance of the printed material in various environments. The topographic and color characteristics can be combined through multi-dimensional data fusion techniques (such as principal component analysis PCA, weighted average, etc.). In this way, comprehensive printed display characteristic data can be generated and output as charts or reports to help designers quickly understand the display effect of the target printed object in different environments.

[0053] More specifically, through the combined atlases and data, the printed display characteristics of the target printed object are finally obtained. These characteristics can reflect the topographic changes and color performance of the printed material in different optical environments, providing a scientific basis for subsequent quality inspection.

[0054] Preferably, the steps of performing simulation analysis on the display effects of the topographic data characteristics and color data characteristics of the target printed object according to a preset number of optical environment filters to obtain topographic display effect atlases and color display effect atlases of the target printed object corresponding to various optical environment filters include:

[0055] S131: Perform sub - location processing based on the grid coordinate system on the topographic data characteristics and color data characteristics of the target printed object respectively to obtain display versions of the topographic data characteristics and color data characteristics corresponding to the grid coordinate system. The display versions of the topographic data characteristics and color data characteristics corresponding to the grid coordinate system are respectively marked as topographic grid feature distributions and color grid feature distributions;

[0056] S132: Respectively perform environmental correction effect simulation on the morphological grid feature distribution and the color grid feature distribution according to a preset optical environment filter, and respectively assign grid feature parameters to the morphological grid feature distribution and the color grid feature distribution according to the results of the environmental correction effect simulation, so as to obtain the morphological grid parameter distribution and the color grid parameter distribution of the target printing object corresponding to the optical environment filter;

[0057] S133: Repeat the above steps for various preset optical environment filters to obtain the morphological grid parameter distribution and the color grid parameter distribution of the target printing object corresponding to various optical environment filters;

[0058] S134: Conduct a comprehensive evaluation of the grid independent display effect and the grid adjacent display effect on the morphological grid parameter distribution and the color grid parameter distribution of the target printing object corresponding to various optical environment filters, so as to obtain the morphological display effect level and the color display effect level of the target printing object corresponding to various optical environment filters;

[0059] S135: The morphological display effect levels of the target printing object corresponding to various optical environment filters jointly form a morphological display effect atlas, and the color display effect levels of the target printing object corresponding to various optical environment filters jointly form a color display effect atlas.

[0060] Specifically, the morphological data features and color data features of the target printing object are subdivided and positioned in the grid coordinate system, which means mapping the morphology and color data of the object into a predetermined grid. Usually, a two-dimensional or three-dimensional grid coordinate system is used to represent the features of each local area. According to the size and complexity of the target printing object, a suitable grid coordinate system is created, and the features such as the shape, texture, and pattern of the target printing object and the color information are mapped into the grid to generate the morphological grid feature distribution and the color grid feature distribution. Each grid node will contain the morphological and color data of the target object in that area.

[0061] More specifically, through the subdivided data after gridification, the morphological grid feature distribution and the color grid feature distribution are respectively marked to ensure that the morphological and color information is stored and managed according to the grid distribution. The morphological grid feature distribution will contain data such as the geometric shape, texture, and surface smoothness of each grid cell; the color grid feature distribution will contain information such as the color value, brightness, and saturation of each grid cell.

[0062] More specifically, an environmental correction effect simulation is performed on the topographic grid feature distribution and the color grid feature distribution according to a preset optical environment filter. This step aims to simulate the changes in topography and color in different grid cells under different optical environments. An optical simulation technique (such as a ray-tracing-based rendering technique) is used to simulate the environmental correction effect. The environmental correction takes into account the effects of different lighting, reflection, refraction, etc. on the topography and color of each grid cell, and this will calculate the new values of the topography and color of each grid cell according to the type of filter.

[0063] More specifically, according to the results of the environmental correction effect simulation, parameters are assigned to the grid features of the topographic grid feature distribution and the color grid feature distribution respectively. This means that new topography and color parameters are assigned to each grid cell according to the results of different optical environment corrections. The topography and color feature parameters of each grid cell will be assigned new values based on the data after environmental simulation. These values include changes in form (such as surface gloss, concavity and convexity, etc.) and changes in color (such as color difference, saturation change, etc.).

[0064] More specifically, the above steps are repeated to perform a correction effect simulation on the target printed object under multiple preset optical environment filters, and a topographic grid parameter distribution and a color grid parameter distribution are obtained respectively. This will generate different topography and color effects for each environmental filter. Different lighting conditions, reflection modes, etc. are configured for each optical environment filter, and the above process is repeated using a rendering engine or an optical simulation algorithm. A new set of topography and color parameter distributions is generated under each filter.

[0065] More specifically, a comprehensive evaluation is performed on the topographic grid parameter distribution and the color grid parameter distribution under each optical environment filter, considering the independent display effect of the grid cells and the interaction effect of adjacent grids. Image processing techniques such as high-pass filtering, edge detection, and adjacency analysis are used to independently evaluate the topography and color changes of each grid cell. At the same time, the transition effect and mutual influence between adjacent grids should also be considered. The comprehensive evaluation will consider local details and overall effects.

[0066] More specifically, according to the results of the comprehensive evaluation, a topographic display effect level and a color display effect level are generated for each optical environment filter. A level represents the display performance of the target printed object under an optical environment filter. The topographic display effect levels and the color display effect levels under different optical environment filters are combined to generate a final topographic display effect atlas and a color display effect atlas. Through data visualization techniques (such as heat maps, color scale maps, three-dimensional maps, etc.), the display effect levels under different environmental filters are combined to generate a topographic display effect atlas and a color display effect atlas. These atlases can display the topography and color changes under each environment.

[0067] Preferably, the step of analyzing an image detection solution for the target printing object according to the printing display feature to obtain an image data detection solution for the target printing object includes:

[0068] S21: Perform an optical environment filter evaluation on the target printing object using the topography display effect spectrum and the color display effect spectrum in the printing display feature to obtain several optical environment filters prepared to participate in the detection work;

[0069] S22: Perform a test execution plan conversion on several optical environment filters prepared to participate in the detection work to obtain a filter execution plan for applying the optical environment filter to the target printing object; wherein, the filter execution plan includes filter execution parameters and filter execution time;

[0070] S23: Construct an image acquisition execution plan corresponding to the filter execution time in each filter execution plan to obtain an image acquisition execution plan corresponding to each filter execution plan;

[0071] S24: Each filter execution plan and the image acquisition execution plan corresponding to each filter execution plan together constitute the image data detection solution.

[0072] Specifically, the step of evaluating the optical environment filter for the target printing object using the topography display effect spectrum and the color display effect spectrum aims to identify the optical environment conditions related to the display effect of the target printing object in actual application. By comparing the topography and color performance of the target printing object under different optical environments, the most suitable optical environment filters for detection are selected. By simulating the display effects under different optical environment filters, the topography and color characteristics of the target printing object are evaluated, so as to select the best set of optical environment filters for evaluation. These filters will be used for subsequent image detection work.

[0073] More specifically, for the best optical environment filters evaluated in the previous step, formulate a test execution plan and convert it into a specific filter execution plan. These filter execution plans will include the execution parameters of each optical environment filter (how to apply the optical environment filter to the target printing object) and the execution time (the action time of each filter). According to the characteristics of different optical environment filters, convert them into specific filter execution plans. In this process, the execution parameters of the filters need to be clearly set, and at the same time, the execution time of each filter is determined to ensure that the test effect matches the display effect of the target printing object.

[0074] More specifically, according to the execution time of each optical environment filter, a corresponding image acquisition execution plan is constructed. The image acquisition execution plan will be temporally planned according to the filter execution time to ensure that the image acquisition process can collect the required image data within the correct time. According to the filter execution time, parameters such as the time interval of image acquisition, image resolution, and acquisition frequency are formulated. Each filter execution plan may correspond to one or more image acquisition time points, and the construction of the acquisition plan will take into account the performance changes of image data under different filters. For example, for some filters, long exposure may be required to capture light changes, while for other filters, only short-time acquisition may be needed.

[0075] More specifically, the execution plan of each optical environment filter and the corresponding image acquisition execution plan are combined to form a final image data detection plan. This plan will include the specific operation steps of applying filters to the target printed object in different optical environments, as well as the corresponding image acquisition plan for each filter. The combination of each filter execution plan and the image acquisition execution plan forms a complete detection process. This detection process will include the parameter settings, execution time, acquisition plan, etc. of each optical environment filter. During the whole process, an automated detection system (such as a computer vision and image processing platform) can be used to execute the operations in the plan to ensure that image data can be obtained efficiently and accurately.

[0076] Preferably, the steps of applying the topography display effect atlas and the color display effect atlas in the printed display features to the optical environment filter evaluation of the target printed object to obtain several optical environment filters prepared to participate in the detection work include:

[0077] S211: According to the topography display effect atlas and the color display effect atlas in the printed display features, perform a detection adaptability analysis on each optical environment filter to obtain the detection adaptability feature distribution of various optical environment filters reflected by the topography display effect atlas and the color display effect atlas in the printed display features; wherein, the detection adaptability feature distribution is used to describe the detection effect of the optical environment filter on various quality defect types.

[0078] S212: Retrieve various quality defect types of the target printed object and their corresponding detection weights from a preset database, and use various quality defect types and their corresponding detection weights as the detection objectives.

[0079] S213: Analyze the achievement status of the detection objectives according to the detection adaptability feature distribution of various optical environment filters to obtain the detection objective achievement features of various optical environment filters corresponding to the detection objectives.

[0080] S214: Filter combinations are made for various optical environment filters according to the detection purpose achievement characteristics of various optical environment filters, and multi-dimensional analysis of the detection purpose achievement status, filter execution resource consumption, and filter detection error self-correction of the filter combinations is carried out to obtain several filter combinations and combination evaluation values;

[0081] S215: Each optical environment filter in the filter combination with the best combination evaluation value is used as several optical environment filters prepared to participate in the detection work.

[0082] Specifically, in combination with the topography display effect diagram and the color display effect diagram, the adaptability of each optical environment filter is analyzed respectively. These display diagrams will reflect the performance of the target printed object under different optical environments, including topography and color characteristics. According to these characteristics, the detection adaptability of each optical environment filter for various quality defects (such as color difference, uneven printing, morphological defects, etc.) is evaluated. Computer vision and image processing algorithms are used to analyze the topography and color diagrams, and according to the characteristics of each filter, the detection effects under different quality defect types are simulated. The adaptability feature distribution of the filter is established through machine learning or statistical methods, and the performance characteristics of each filter for different defect types are generated.

[0083] More specifically, according to the detection adaptability feature distribution of the optical environment filter obtained in the first step, combined with the target detection task, that is, the preset quality defect types and their corresponding detection weights, the performance of each filter in actual detection is analyzed. Here, the weights of the quality defect types reflect the importance of various defects in the actual production process, so as to provide a basis for subsequent filter evaluation. By matching the adaptability features of each optical environment filter with the preset quality defect types and their weights, multi-dimensional target achievement analysis is carried out. This process calculates the matching degree between the features of the filter and the target, and generates a "target achievement status" report of the filter to measure the utility of each filter in the detection task.

[0084] More specifically, according to the "goal achievement status" in the second step and the detection purpose achievement characteristics of various filters, filter combination analysis is carried out. In this step, various optical environment filters are combined, and multi-dimensional factors such as the detection effect, resource consumption, detection error, and error self-correction of different combinations are evaluated. Finally, the filter combination with the best combined evaluation value is selected. The multi-objective optimization algorithm (such as genetic algorithm, simulated annealing algorithm, etc.) is used to evaluate different filter combinations. During the evaluation process, the following multiple factors are considered: Goal achievement status: The ability of the filter combination to achieve the preset goal (i.e., effectively detect various defects), Filter execution resource consumption: including the execution time, computing resources, hardware requirements, etc. of the filter, Filter detection error and self-correction: Evaluate the error of each filter combination and select those filter combinations that can automatically correct the detection error. These factors are used to generate a comprehensive evaluation value through weighted calculation, and the optimal filter combination is selected based on this value.

[0085] More specifically, based on the filter combination evaluation results obtained through multi-dimensional analysis, the optical environment filter combination with the best comprehensive evaluation value is selected. These filters will be used as the optical environment filters prepared to participate in the detection work. After comprehensively considering each evaluation dimension, the best combination is selected and applied to the actual detection. The selection of the filter combination not only ensures the optimization of the detection effect but also improves the resource utilization efficiency during the detection process and minimizes errors and unnecessary resource waste.

[0086] Preferably, the step of collecting image data of the target printed object according to the image data detection scheme to obtain the detection image data of the target printed object includes:

[0087] S31: Apply the optical environment filters to the target printed object in sequence according to each filter execution scheme in the image data detection scheme, so that the target printed object is successively in each specified optical environment filter;

[0088] S32: Collect images of the target printed object in each specified optical environment filter according to the image acquisition execution scheme in the image data detection scheme to obtain the detection image data of the target printed object; wherein, the detection image data includes the detection images under various optical environment filters.

[0089] Specifically, according to the filter execution scheme in the image data detection scheme, each optical environment filter is applied to the target printed object in sequence. Each optical environment filter corresponds to a specific detection environment, which may include different optical characteristics such as light, color temperature, contrast, etc. These filters can simulate different environmental conditions, aiming to test the quality of the target printed object under various optical environments. An optical filter device or a combination of adjustable light sources, cameras, and optical components is used to apply different optical environment filters to the target printed object in sequence. These filters simulate different detection conditions by changing the light attributes (such as brightness, hue, contrast, etc.) to ensure a comprehensive and multi-dimensional detection of the target printed object.

[0090] More specifically, according to the image acquisition execution scheme in the image data detection scheme, the performance of the target printed object under each optical environment filter is respectively imaged to obtain the detection image data under each filter condition. Each detection image data reflects the printing quality characteristics of the target printed object under this specific optical environment. A high-precision camera or image acquisition device is used to perform image acquisition under the action of different optical environment filters. This process usually involves using automated equipment to scan and photograph the target object to ensure that the image data under each optical environment filter is accurately recorded. It may involve multiple imaging technologies such as visible light imaging, infrared imaging, ultraviolet imaging, etc. to capture different quality characteristics.

[0091] More specifically, through the above acquisition process, the detection image data of the target printed object under different optical environment filters is obtained. These detection image data constitute a multi-level image data set. Each image reflects the performance of the target printed object under a certain specific optical environment. Multiple image data are generated and stored. Each image corresponds to the quality performance of the target printed object under a specific filter. High-resolution image formats (such as TIFF, PNG, JPEG) can be used for saving, and relevant optical environment filter information and image acquisition metadata are attached for subsequent analysis and processing.

[0092] Preferably, the pre-training step of the machine learning model includes:

[0093] S41: Collect the model learning data of the machine learning model; wherein, the model learning data includes detection image data and corresponding image evaluation annotations. The image evaluation annotations are used for the annotation processing of the quality evaluation results of the detection image data, and the image evaluation annotations correspond to the quality evaluation report;

[0094] S42: Partition the model learning data to obtain a training set, a validation set, and a test set;

[0095] S43: Construct a machine learning model composed of an input layer, multiple recurrent layers, and an output layer;

[0096] S44: Substitute the training set into the machine learning model, and let the machine learning model perform feature analysis on the data mapping relationship between the detected image data and the image evaluation annotation according to the training set, so as to obtain the data mapping relationship between the detected image data and the image evaluation annotation;

[0097] S45: Substitute the validation set into the machine learning model, and let the machine learning model perform verification processing on the data mapping relationship according to the validation set, so as to obtain the confidence level of the data mapping relationship;

[0098] S46: Adjust the model parameters of the machine learning model according to the confidence level of the data mapping relationship, so as to obtain a new data mapping relationship;

[0099] S47: Substitute the test set into the machine learning model, and let the machine learning model perform test processing on the data mapping relationship according to the test set, to obtain the data fitting condition of the data mapping relationship;

[0100] S48: When the data fitting condition of the data mapping relationship meets the training standard, deploy the machine learning model with the data mapping relationship to complete the training of the machine learning model;

[0101] S49: When the data fitting condition of the data mapping relationship does not meet the training standard, re - perform the training of the data mapping relationship until a data mapping relationship with a data fitting condition that meets the training standard is obtained.

[0102] Specifically, it is necessary to collect model learning data suitable for training, which includes detected image data and corresponding image evaluation annotations. The image evaluation annotation is used to evaluate the quality of each detected image data and record the corresponding evaluation results. The image evaluation annotation may include information such as the type of defects, severity, printing quality grade, etc. in the image. These data will be used as label data for supervised learning. Each image is evaluated through manual annotation or automated tools (such as image processing techniques, computer vision algorithms, etc.), and the quality annotation information is attached to the image data to form a pair of training samples.

[0103] More specifically, the model learning data (including detection image data and image evaluation annotations) is partitioned into sets, usually divided into a training set, a validation set, and a test set: Training set: Used to train the machine learning model and adjust the model's parameters. Validation set: Used to monitor the model's performance during training, adjust hyperparameters, and select the optimal model. Test set: Used to evaluate the final performance of the model and verify whether the model has good generalization ability. A common partitioning ratio is 70% for the training set, 15% for the validation set, and 15% for the test set. When partitioning, ensure the representativeness and uniformity of the data to ensure that the data in each subset is representative.

[0104] More specifically, according to the complexity of the problem and the characteristics of the data, construct a suitable machine learning model. According to the description, the model structure is a neural network model composed of an input layer, multiple recurrent layers (such as LSTM, GRU, etc.), and an output layer. Input layer: Accepts the processed image data as input. Multiple recurrent layers: Used to learn the temporal information and spatial relationships in the image data, usually used for tasks such as image sequences and video analysis. Output layer: Outputs the quality evaluation annotations of the image according to the task requirements, such as classification labels or regression results. Select a suitable framework (such as TensorFlow, PyTorch, etc.) to build the neural network model and configure hyperparameters such as the appropriate number of layers, activation functions, and loss functions.

[0105] More specifically, substitute the training set data into the constructed machine learning model and let the model learn according to the relationship between the image data and the image evaluation annotations in the training set. The model performs feature analysis and parameter updates on the data mapping relationship through the backpropagation algorithm, and optimizes the model parameters through the gradient descent method or other optimization algorithms (such as Adam, RMSprop, etc.) to minimize the error between the prediction result and the actual annotation.

[0106] More specifically, substitute the validation set into the model. The model validates the learned mapping relationship according to the validation set, evaluates the confidence of the model in the data mapping relationship through the validation set, calculates the error or loss function value, and determines whether it is necessary to adjust the model hyperparameters. After each round of training, use the validation set for prediction and calculate metrics such as accuracy, F1 score, and loss to evaluate the performance of the model. If the model's performance on the validation set is not ideal, adjust the hyperparameters (such as learning rate, batch size, etc.).

[0107] More specifically, according to the evaluation results of the confidence on the validation set, adjust the model's parameters (such as network weights, learning rate, etc.) to optimize the data mapping relationship. This process may include multiple trainings and validations until the model reaches the expected performance level. Use hyperparameter optimization algorithms (such as grid search, random search, or Bayesian optimization) to find the optimal parameter combination.

[0108] More specifically, the test set data is substituted into the model for testing to evaluate the fitting of the data mapping relationship and check the performance of the model on unseen data. The role of the test set is to verify the generalization ability of the model and ensure its good performance in real-world applications. In the testing phase, the test set is used for model prediction, and common evaluation metrics (such as precision, recall, F1-score, AUC, etc.) are calculated to evaluate whether the model meets the expected standards.

[0109] More specifically, when the data fitting condition of the model on the test set reaches the training standard, it is considered that the model has been trained and has sufficient accuracy for deployment and application. The task in the deployment phase is to apply the trained model to the actual scenario for real-time prediction and decision-making.

[0110] More specifically, if it is found in the testing phase that the data fitting condition does not meet the training standard, the model needs to be retrained until a satisfactory fitting effect is achieved. Adjustments are made by analyzing the training results (such as overfitting or underfitting), increasing the data volume, changing the model structure, or using regularization techniques, etc., until the performance of the model meets the standard.

[0111] Refer to Figure 2 As shown, in the second aspect, the present invention provides an automated intelligent printing quality detection system for implementing the automated intelligent printing quality detection method described in any one of the first aspects, including:

[0112] A display and analysis module for obtaining the printing design data of the target printing object and performing feature analysis on the printing finished product display effect of the target printing object according to the printing design data to obtain the printing display features of the target printing object;

[0113] A scheme analysis module for analyzing the image detection scheme of the target printing object according to the printing display features to obtain the image data detection scheme of the target printing object;

[0114] A detection execution module for collecting the image data of the target printing object according to the image data detection scheme to obtain the detection image data of the target printing object;

[0115] A quality evaluation module for analyzing and processing the detection image data according to a pre-trained machine learning model to obtain a quality evaluation report of the target printing object.

[0116] In this embodiment, for the specific implementation of each module in the above system embodiment, please refer to that described in the above method embodiment, and details are not repeated here.

[0117] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An automated intelligent printing quality detection method, characterized in that Including: Obtain the printing design data of the target printing object, and perform feature analysis on the display effect of the printed finished product of the target printing object according to the printing design data, so as to obtain the printing display characteristics of the target printing object; Analyze the image detection scheme of the target printing object according to the printing display characteristics, so as to obtain the image data detection scheme of the target printing object; Collect the image data of the target printing object according to the image data detection scheme, so as to obtain the detected image data of the target printing object; Analyze and process the detected image data according to the pre-trained machine learning model, so as to obtain the quality evaluation report of the target printing object; The steps of obtaining the printing design data of the target printing object and performing feature analysis on the display effect of the printed finished product of the target printing object according to the printing design data to obtain the printing display characteristics of the target printing object include: Obtain the printing design data of the target printing object; wherein, the target printing object includes a substrate part and a printing part, and the printing design data is used to describe the specific form that the printing part is pre-planned to display on the substrate part; Perform data parsing on the printing design data in the morphology dimension and color dimension to obtain the morphology data characteristics and color data characteristics of the target printing object; Perform simulation analysis on the display effect of the morphology data characteristics and color data characteristics of the target printing object according to a plurality of preset optical environment filters, so as to obtain the morphology display effect spectrum and color display effect spectrum of the target printing object corresponding to various optical environment filters; wherein, the optical environment filter is a specific optical environment applied to the target printing object during image acquisition of the target printing object, and the specific optical environment includes light intensity, light frequency, and light color; Combine the morphology display effect spectrum and color display effect spectrum of the target printing object corresponding to various optical environment filters to obtain the printing display characteristics of the target printing object; The steps of analyzing the image detection scheme of the target printing object according to the printing display characteristics to obtain the image data detection scheme of the target printing object include: Evaluate the optical environment filters applied to the target printing object for the morphology display effect spectrum and color display effect spectrum in the printing display characteristics to obtain several optical environment filters to be prepared to participate in the detection work; Convert the test execution scheme for several optical environment filters to be prepared to participate in the detection work to obtain a filter execution scheme for applying the optical environment filter to the target printing object; wherein, the filter execution scheme includes filter execution parameters and filter execution time; Construct an image acquisition execution scheme corresponding to the filter execution time in each filter execution scheme to obtain an image acquisition execution scheme corresponding to each filter execution scheme; Each filter execution scheme and the image acquisition execution scheme corresponding to each filter execution scheme together constitute the image data detection scheme.

2. The automated intelligent printing quality detection method according to claim 1, wherein The steps of simulating and analyzing the display effects of the morphological data features and color data features of the target printed object according to a plurality of preset optical environment filters to obtain the morphological display effect atlas and color display effect atlas of the target printed object corresponding to various optical environment filters include: Performing sub - localization processing based on a grid coordinate system on the morphological data features and color data features of the target printed object respectively to obtain the display versions of the morphological data features and color data features corresponding to the grid coordinate system. The display versions of the morphological data features and color data features corresponding to the grid coordinate system are respectively marked as morphological grid feature distributions and color grid feature distributions; Performing environmental correction effect simulation on the morphological grid feature distribution and the color grid feature distribution respectively according to the preset optical environment filters, and endowing parameters of grid features to the morphological grid feature distribution and the color grid feature distribution respectively according to the results of the environmental correction effect simulation to obtain the morphological grid parameter distribution and color grid parameter distribution of the target printed object corresponding to the optical environment filters; Repeating the above steps for various preset optical environment filters to obtain the morphological grid parameter distribution and color grid parameter distribution of the target printed object corresponding to various optical environment filters; Comprehensively evaluating the grid - independent display effect and grid - adjacent display effect of the morphological grid parameter distribution and color grid parameter distribution of the target printed object corresponding to various optical environment filters to obtain the morphological display effect level and color display effect level of the target printed object corresponding to various optical environment filters; The morphological display effect levels of the target printed object corresponding to various optical environment filters together form a morphological display effect atlas, and the color display effect levels of the target printed object corresponding to various optical environment filters together form a color display effect atlas.

3. The automated intelligent printing quality inspection method according to claim 1, wherein The steps of evaluating the optical environment filters applied to the target printed object for the morphological display effect atlas and color display effect atlas in the printing display features to obtain several optical environment filters to be prepared for the detection work include: Performing detection adaptability analysis of corresponding quality defect types on each optical environment filter according to the morphological display effect atlas and color display effect atlas in the printing display features to obtain the detection adaptability feature distributions of various optical environment filters reflected by the morphological display effect atlas and color display effect atlas in the printing display features; wherein, the detection adaptability feature distribution is used to describe the detection effect of the optical environment filter on various quality defect types; Retrieving various quality defect types of the target printed object and corresponding detection weights from a preset database, and using various quality defect types and corresponding detection weights as detection purposes; Analyzing the achievement status of the detection purposes according to the detection adaptability feature distributions of various optical environment filters to obtain the detection purpose achievement features of various optical environment filters corresponding to the detection purposes; Filter combinations are made for various optical environment filters according to the detection purpose achievement characteristics of various optical environment filters, and a multi-dimensional analysis of the detection purpose achievement status, filter execution resource consumption, and filter detection error self-correction of the filter combinations is carried out to obtain several filter combinations and combination evaluation values; Each optical environment filter in the filter combination with the best combination evaluation value is used as several optical environment filters prepared to participate in the detection work.

4. The automated intelligent printing quality detection method according to claim 1, wherein The steps of collecting image data of the target printed object according to the image data detection scheme to obtain the detection image data of the target printed object include: Applying optical environment filters to the target printed object in sequence according to each filter execution scheme in the image data detection scheme, so that the target printed object is successively in each specified optical environment filter; Performing image acquisition on the target printed object in each specified optical environment filter according to the image acquisition execution scheme in the image data detection scheme to obtain the detection image data of the target printed object; wherein, the detection image data includes detection images under each optical environment filter.

5. The automated intelligent printing quality detection method according to claim 1, characterized in that, The pre-training steps of the machine learning model include: Collecting model learning data of the machine learning model; wherein, the model learning data includes detection image data and corresponding image evaluation annotations, the image evaluation annotations are used for the annotation processing of the quality evaluation results of the detection image data, and the image evaluation annotations correspond to the quality evaluation report; Performing set partitioning on the model learning data to obtain a training set, a validation set, and a test set; Constructing a machine learning model composed of an input layer, multiple recurrent layers, and an output layer; Substituting the training set into the machine learning model, and enabling the machine learning model to perform feature analysis on the data mapping relationship between the detection image data and the image evaluation annotations according to the training set to obtain the data mapping relationship between the detection image data and the image evaluation annotations; Substituting the validation set into the machine learning model, and enabling the machine learning model to perform verification processing on the data mapping relationship according to the validation set to obtain the confidence of the data mapping relationship; Adjusting the model parameters of the machine learning model according to the confidence of the data mapping relationship to obtain a new data mapping relationship; Substituting the test set into the machine learning model, and enabling the machine learning model to perform test processing on the data mapping relationship according to the test set to obtain the data fitting status of the data mapping relationship; When the data fitting status of the data mapping relationship meets the training standard, performing model deployment on the machine learning model with the data mapping relationship to complete the training of the machine learning model; When the data fitting status of the data mapping relationship does not meet the training standard, re-training the data mapping relationship until a data mapping relationship with a data fitting status meeting the training standard is obtained.

6. An automated intelligent printing quality detection system, characterized in that, An automated intelligent printing quality detection method for implementing any one of claims 1-5 includes: A display analysis module, which is used to obtain the printing design data of a target printing object, and perform feature analysis on the display effect of the printed finished product of the target printing object according to the printing design data, so as to obtain the printing display features of the target printing object; A scheme analysis module, which is used to analyze the image detection scheme of the target printing object according to the printing display features, so as to obtain the image data detection scheme of the target printing object; A detection execution module, which is used to collect image data of the target printing object according to the image data detection scheme, so as to obtain the detected image data of the target printing object; A quality evaluation module, which is used to analyze and process the detected image data according to a pre-trained machine learning model, so as to obtain a quality evaluation report of the target printing object.

Images