A method, system, medium and product for detecting print quality

Abstract

The application provides a printing quality detection method, system, medium and product, and relates to the technical field of printing. The method establishes the corresponding relationship between the target printing result and the actual printing result by acquiring standard image data and printed product image data; the image data is divided by using a region identification model, key regions (such as text, complex patterns, color transition content) and non-key regions (such as solid color, background, white space content) are extracted, and the pixel coordinate ranges of the two are ensured to be one-to-one corresponding; further, the similarity of the key regions and the non-key regions is calculated by using a printing similarity matching model, and the comprehensive similarity is obtained by weighted calculation. Finally, by comparing the comprehensive similarity with a preset threshold, it is accurately judged whether the printing quality is qualified. The high-precision and differentiated detection of the printing quality is realized, and the accuracy, objectivity and efficiency of the printing quality detection are improved.

Description

Technical Field

[0001] This application relates to the field of printing technology, and in particular to a printing quality inspection method, system, medium, and product. Background Technology

[0002] With the rapid development of printing technology, the requirements for printing quality are constantly increasing. During the printing process, due to the influence of various factors such as equipment status, ink ratio, and paper quality, the actual effect of the printed product often deviates from the expected effect. This deviation not only affects the visual appearance and functionality of the printed product, but may also lead to reduced customer satisfaction and increased production costs. Therefore, achieving accurate printing quality inspection is of great significance for ensuring printing quality and improving production efficiency.

[0003] In related technologies, commonly used print quality inspection methods typically inspect the overall print quality of the printed content. This method works well for simple images or regular content, but when faced with highly complex print content, the accuracy and reliability of these technologies in print quality inspection are insufficient. Summary of the Invention

[0004] This application provides a printing quality inspection method, system, medium, and product to alleviate the problems of insufficient accuracy and reliability in printing quality inspection when dealing with complex printed content.

[0005] In a first aspect, this application provides a printing quality inspection method, comprising: acquiring standard image data and finished printed product image data corresponding to a current printed matter, wherein the standard image data is the target printing result of the current printed matter, and the finished printed product image data is the actual printing result of the current printed matter; inputting the standard image data into a preset region recognition model to obtain a first region segmentation result, wherein the first region segmentation result includes a first critical region and a first non-critical region, wherein the first critical region includes text, complex patterns, and color transition content, and the first non-critical region includes solid colors, background, and white space content; and inputting the finished printed product image data into the preset region recognition model to obtain a first region segmentation result. A region recognition model is used to obtain a second region segmentation result, which includes a second key region and a second non-key region. The pixel coordinate range of the second region segmentation result corresponds one-to-one with the pixel coordinate range of the first region segmentation result. The first region segmentation result and the second region segmentation result are input into a preset printing similarity matching model to obtain the key region similarity and the non-key region similarity. The key region similarity and the non-key region similarity are weighted and calculated to obtain a comprehensive similarity. When the comprehensive similarity is greater than or equal to a preset comprehensive similarity threshold, the printing quality of the current printed product is determined to be qualified. By employing the aforementioned technical solution, standard image data and finished printed product image data of the current printed matter can be obtained, clearly defining the correspondence between the target printing effect and the actual printing result. Next, by inputting the standard image data and finished printed product image data into a preset region recognition model, key and non-key regions are divided, allowing the content of the printed matter to be differentiated based on importance. Then, the two sets of region division results are input into a printing similarity matching model to calculate the similarity between key and non-key regions, thereby enabling precise matching based on the characteristics of different regions. Finally, a weighted calculation yields a comprehensive similarity score, which can comprehensively and objectively reflect whether the overall quality of the printed matter is up to standard. Different content has varying degrees of importance to printing quality; therefore, by differentiating the image data through region division, the influence and importance of key content are prevented from being diluted, thus avoiding a decrease in the reliability of the printing quality detection results. When dealing with complex printed content, this effectively ensures the accuracy and reliability of printing quality detection.

[0006] In conjunction with some embodiments of the first aspect, in some embodiments, the key region and the non-key region each include multiple map tiles. A weighted calculation is performed on the similarity of the key region and the similarity of the non-key region to obtain a comprehensive similarity. Specifically, this includes: determining the area influence factor corresponding to each map tile based on the tile area of ​​each map tile in the key region and the non-key region; adjusting the similarity of the key region or the non-key region corresponding to each map tile based on the area influence factor; and performing a weighted calculation on the adjusted similarity of the key region and the non-key region to obtain a comprehensive similarity. By employing the above technical solution, firstly, key and non-key regions are divided into multiple patches, and an area influence factor is determined based on the area of ​​each patch, effectively quantifying the impact of each patch on the overall similarity. Then, the similarity of the corresponding patches is adjusted using the area influence factor, ensuring that the similarity calculation for each patch reflects the importance of its actual area. Finally, by weighting the adjusted similarities, the resulting comprehensive similarity more accurately reflects printing quality, especially when printed materials contain large areas of similarity or difference, avoiding bias in the overall evaluation due to local details. Therefore, this method further improves the objectivity and regional specificity of the detection results.

[0007] In conjunction with some embodiments of the first aspect, in some embodiments, the similarity of the key region or the similarity of the non-key region corresponding to each of the map tiles is adjusted according to the area influence factor, specifically including: determining the shape influence factor corresponding to each of the map tiles according to the shape regularity of each of the map tiles; and adjusting the similarity of the key region or the similarity of the non-key region corresponding to each of the map tiles according to the shape influence factor and the area influence factor. By adopting the above technical solution, a shape influence factor is added to the existing adjustment of tile similarity based on the area influence factor. This allows the similarity calculation to consider not only tile area but also the shape complexity of the tiles. Complex-shaped tiles typically involve more printing details or key content; the introduction of the shape influence factor allows for a more refined evaluation of the similarity of these tiles. The combined adjustment of both area and shape influence factors makes the similarity calculation results more sensitive to complex content, thereby improving the quality detection capability for printing details and ensuring the applicability and accuracy of the technical solution in the detection of complex printed content.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, after weighting the adjusted key region similarity and the non-key region similarity to obtain a comprehensive similarity, the method further includes: when the comprehensive similarity is less than a preset comprehensive similarity threshold, determining that the printing quality of the current printed matter is unqualified; comparing each of the blocks of the current printed matter one by one to determine whether the key region similarity or the non-key region similarity corresponding to each block is lower than the average block similarity, the average block similarity being the ratio of the sum of the key region similarity and the non-key region similarity to the number of blocks; if so, identifying the corresponding block as a defective block and generating an alarm signal, the alarm signal including the pixel coordinate range and defect type of the defective block. By employing the above technical solution, when the overall similarity is below a preset threshold, specific defective blocks can be quickly located by comparing the similarity of each block individually. Furthermore, by determining whether the block similarity is lower than the average block similarity, potential defective areas are accurately identified, and alarm signals are generated, including the pixel coordinate range and defect type of the defective block. This refined defect location and marking method not only helps to quickly identify printing problems but also clearly defines the specific defect location and type, facilitating subsequent adjustments and improvements, and significantly improving detection efficiency and the accuracy of problem feedback.

[0009] In conjunction with some embodiments of the first aspect, in some embodiments, after obtaining the standard image data corresponding to the current printed matter and the printed product image data, the method further includes: performing a shooting quality assessment on the printed product image data based on the standard image data to obtain a quality assessment value, wherein the shooting quality assessment includes detecting the resolution, contrast, and brightness distribution of the printed product image data; if the quality assessment value is lower than a preset quality assessment threshold, the printed product image data is determined to be unqualified, and a reshoot is immediately requested. By adopting the above technical solution, after acquiring standard image data and printed product image data, the first step is to assess the image quality of the printed product image data. This effectively ensures that the image data used for inspection has sufficient resolution, contrast, and brightness distribution. When the quality assessment value is lower than a preset threshold, the image data is immediately deemed unqualified and a re-engraving is requested, thus preventing errors in the inspection results caused by data quality issues from the outset. This mechanism ensures that subsequent inspection processes are based on high-quality data, thereby improving the overall accuracy and reliability of the inspection.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, after obtaining the standard image data corresponding to the current printed matter and the printed product image data, the method further includes: performing edge detection on the standard image data to obtain image edge features; and performing feature correction on the printed product image data based on the image edge features to generate corrected printed product image data, wherein the feature correction includes rotation correction and scale normalization of the image. By employing the aforementioned technical solution, edge detection after acquiring standard image data can extract clear image edge features. Based on these edge features, feature correction is performed on the printed product image data, including image rotation correction and scale normalization. This effectively eliminates deviations introduced during the shooting process due to angle or proportion issues. The corrected printed product image data more closely approximates the true characteristics of the standard image data, thus providing a more accurate and consistent basis for subsequent quality inspection. This process significantly improves the accuracy and consistency of printing quality inspection.

[0011] In conjunction with some embodiments of the first aspect, in some embodiments, after performing feature correction on the printed product image data based on the image edge features to generate corrected printed product image data, the method further includes: obtaining the standard resolution of the standard image data and the finished product resolution of the corrected printed product image data; determining whether the standard resolution is consistent with the finished product resolution; if they are inconsistent, adjusting the standard resolution and / or the finished product resolution to make the standard resolution equal to the finished product resolution. By employing the above technical solution, after feature correction, the standard resolution and the finished product resolution are further obtained, and their consistency is determined. This ensures the uniformity of resolution between the standard image and the corrected finished product image. If the resolutions are inconsistent, they can be adjusted to be equal, effectively avoiding matching errors caused by resolution differences. This method can further improve the matching accuracy between the standard image and the finished product image during the comparison process, ensuring the objectivity and consistency of the detection results, and providing more precise technical support for printing quality inspection.

[0012] In a second aspect, embodiments of this application provide a printing quality inspection system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the printing quality inspection system to perform the method described in the first aspect and any possible implementation thereof.

[0013] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a printing quality inspection system, cause the printing quality inspection system to perform the method described in the first aspect and any possible implementation thereof.

[0014] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a printing quality inspection system, cause the printing quality inspection system to perform the method described in the first aspect and any possible implementation thereof.

[0015] Understandably, the printing quality inspection system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0016] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By adopting the acquisition of standard image data and printed product image data, the application of region segmentation models, and the weighted calculation of comprehensive similarity, it is possible to differentiate between key and non-key regions of printed materials, and comprehensively reflect the quality of printing through comprehensive similarity. This effectively solves the problem of insufficient overall evaluation of printed product images in related technologies, thereby achieving accuracy and objectivity in the detection of printing quality for complex printed content.

[0017] 2. By using a method that divides key and non-key areas into patches and adjusting similarity based on the area of ​​the patches, the influence of each patch on the overall similarity can be quantified more accurately. This effectively solves the problem of inaccurate detection results due to differences in regional characteristics in related technologies, thereby achieving a more refined and targeted assessment of printing quality.

[0018] 3. By employing a quality assessment mechanism for the image data of printed products, including the detection of resolution, contrast, and brightness distribution, and requesting re-capture when the quality assessment value is below the threshold, high-quality input of detection data can be ensured from the source. This effectively solves the problem of low-quality image data affecting the printing quality detection results in related technologies, thereby achieving high accuracy and reliability in printing quality detection. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of a printing quality inspection device in an embodiment of this application; Figure 2 This is a flowchart illustrating a printing quality inspection method in an embodiment of this application; Figure 3 This is another schematic flowchart of the printing quality inspection method in the embodiments of this application; Figure 4 This is a schematic diagram of the physical device structure of a printing quality inspection system in the embodiments of this application. Detailed Implementation

[0020] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0021] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0022] To facilitate understanding, the following will be combined with Figure 1 This paper introduces an application scenario for printing quality inspection.

[0023] After the printing press completes printing, the printed materials are transported to a printing quality inspection device for quality control. When the printed materials reach the image acquisition area of ​​the printing quality inspection device, the printing quality inspection system acquires the finished printed image data through the device's image acquisition equipment. The system then processes this image data to obtain the final printing quality inspection result.

[0024] Please see Figure 1 , Figure 1 This is a schematic diagram of a printing quality inspection device in an embodiment of this application. The printing quality inspection device 110 includes a finished product image capturing device 111 and a printed matter conveying device 112.

[0025] In practical applications of print quality inspection, the main function of the print quality inspection device 110 is to perform real-time detection and quality assessment of the actual printing effect of printed materials. It includes the following core components: Finished Product Image Capture Equipment 111: This equipment is a core component of the print quality inspection device, typically consisting of hardware such as an industrial-grade camera and a high-precision scanner. It can capture or scan printed materials at high resolution across the entire surface, converting the physical image of the printed material into digital data to obtain the actual image data of the printed product (i.e., the image data of the finished printed product). In application, the finished product image capturing device 111 ensures that the captured printed finished product image data has sufficient clarity and detail by means of precise focusing, adjusting exposure time, and zone acquisition.

[0026] Printed matter conveying equipment 112: This equipment is used to transport printed matter from the output end of the printing press to the imaging area of ​​the inspection device. It usually consists of an automated conveyor belt, positioning device, etc., and can control the conveying speed, stopping position and angle of the printed matter to ensure that the printed matter is in the accurate position and orientation during imaging, and avoid the image acquisition quality being affected by positional deviation or tilt.

[0027] In application, the printing conveyor 112 can be seamlessly integrated with the printing production line to achieve integrated printing and inspection operations.

[0028] To facilitate understanding, the following description, based on the above scenario, outlines a printing quality inspection method provided in this implementation, applicable to a printing quality inspection system (hereinafter referred to as the inspection system), and provides a workflow. Please refer to [link / reference]. Figure 2 This is a flowchart illustrating a printing quality inspection method in an embodiment of this application.

[0029] S201. Obtain the standard image data and the finished printed product image data corresponding to the current printed product. The standard image data is the target printing result of the current printed product, and the finished printed product image data is the actual printing result of the current printed product. Here, "current printed matter" refers to the printed product awaiting quality inspection currently being processed by the inspection system; "standard image data" refers to the ideal image file of the current printed matter, a digital representation of the expected visual effect; and "finished product image data" refers to the actual image data acquired by image acquisition equipment after the current printed matter has been produced by the printing equipment, reflecting the actual presentation effect of the printed matter. Image acquisition equipment includes industrial-grade cameras, high-precision scanners, etc., capable of converting the physical form of the printed matter into digital image information. In the printing production process, when printed materials are transported to the quality inspection station after printing, the inspection system sends image acquisition commands to the image acquisition device. For standard image data, the inspection system directly retrieves it from a pre-built database that stores the standard image corresponding to the current printed material. When acquiring the image data of the finished printed product, the inspection system controls the image acquisition device to take high-resolution pictures of the printed material. If the printed material is large, a method of acquiring images in sections and then stitching them together is used to ensure complete acquisition of the surface information of the printed material, ultimately obtaining the standard image data and the image data of the finished printed product corresponding to the current printed material.

[0030] S202. Input the standard image data into the preset region recognition model to obtain the first region division result. The first region division result includes a first key region and a first non-key region. The first key region includes text, complex patterns and color transition content. The first non-key region includes solid color, background and white space content. Among them, the region recognition model is an intelligent model built into the detection system and trained based on deep learning algorithms (such as convolutional neural networks CNN). It has been trained on a large amount of image data of different types and has the ability to accurately analyze image content and divide regions. The first region division result refers to the image region classification result obtained after the standard image data is processed by the region recognition model. The first key region is used to represent the part of the standard image that plays a decisive role in printing quality and visual communication, including text (such as the title and body content of a book), complex patterns (such as illustrations on a poster and trademark graphics), and color transition content (such as gradient backgrounds and light and shadow effect areas). The first non-key region refers to the relatively minor part of the standard image, including solid color filled areas (such as backgrounds with a single background color), background decorative elements (such as simple line textures), and white space (blank areas reserved to highlight the main subject). The region recognition model is built based on deep learning technology, aiming to achieve accurate region segmentation of input image data. First, a training dataset is constructed, collecting image samples containing various types of printed materials. These samples require precise annotation, clearly indicating the category of each region (e.g., text, complex patterns, color transitions, solid backgrounds, white space, etc.) and its corresponding pixel coordinate range. To improve the model's robustness and generalization ability, data augmentation operations are performed on the dataset, including rotation, scaling, translation, and noise addition, to simulate diverse changes in real-world scenarios.

[0031] The core structure of the model employs a Convolutional Neural Network (CNN). The input is image data, and the output is the region category and corresponding coordinates for each pixel. During training, the image first passes through several convolutional layers to extract local features such as texture, edges, and shape. Dimensionality is then progressively reduced through pooling layers to retain key feature information. Subsequently, after multiple convolutional layers and activation functions, the network gradually learns the relationship between global and local features. Finally, the feature map is mapped to the classification space through fully connected layers, normalized using the Softmax function, and the output is the probability that each pixel belongs to different region categories.

[0032] The model training employs the cross-entropy loss function of labeled data, combined with the backpropagation algorithm to optimize network weights, ensuring high accuracy of the model on the training data. After training, when input image data is received, the model automatically analyzes the image and outputs segmentation results including key regions (such as text, complex patterns, and color transition content) and non-key regions (such as solid colors, backgrounds, and white space content). It also provides the category label and pixel coordinate range for each region, ensuring the accuracy and consistency of the results.

[0033] Once the detection system acquires standard image data, it provides a region recognition model for region segmentation. Specifically, the detection system preprocesses the standard image data according to the model's required format and parameters, including adjusting image resolution and color mode conversion. The preprocessed image data is then input into the preset region recognition model. Upon receiving the data, the model uses its internal multi-layer neural network structure to analyze and extract pixel information, color distribution, and shape features layer by layer. Based on the key features and pattern recognition rules learned during training, the model divides the standard image data into a first key region and a first non-key region, assigning each region a corresponding identifier and pixel coordinate information.

[0034] S203. Input the printed product image data into the region recognition model to obtain the second region division result. The second region division result includes the second key region and the second non-key region, and the pixel coordinate range of the second region division result corresponds one-to-one with the pixel coordinate range of the first region division result. The second region segmentation result refers to the image region classification result generated after the printed product image data is processed by the region recognition model; the second key region is the part of the printed product image corresponding to the first key region of the standard image, including the actual printed text, complex patterns, and color transition content; the second non-key region is the part of the printed product image corresponding to the first non-key region of the standard image, including the actual printed solid colors, background, and white space content; the pixel coordinate range represents the position and size information of each region in the image on the two-dimensional plane, and is a key parameter to ensure that the first region segmentation result and the second region segmentation result can accurately correspond and be effectively compared. After completing the region segmentation of the standard image data, the detection system then processes the printed product image data. Specifically, the detection system first preprocesses the printed product image data to meet the input requirements of the region recognition model, and then inputs the processed image data into the region recognition model. The model restarts its analysis program, using the same algorithms and rules as when processing the standard image data, to perform feature extraction and region segmentation on the printed product image data, obtaining the second region segmentation result. During the segmentation process, the detection system strictly controls the pixel coordinate range of the second region segmentation result output by the model, ensuring a one-to-one correspondence with the pixel coordinate range of the first region segmentation result. This guarantees that the key and non-key regions in the printed product image perfectly match the corresponding regions in the standard image in terms of position and size.

[0035] S204. Input the first region division result and the second region division result into a preset printing similarity matching model to obtain the key region similarity and non-key region similarity. The printing similarity matching model is an intelligent model built based on image processing and pattern recognition technologies in the detection system. It is used to compare and analyze the similarity between the region segmentation results of two images. Key region similarity represents the numerical similarity between the second key region of the printed image and the first key region of the standard image; a higher value indicates that the printing effect of the key parts is closer to the standard. Non-key region similarity refers to the numerical similarity between the second non-key region of the printed image and the first non-key region of the standard image, reflecting the printing quality of non-key parts. For example, when printing a card with complex patterns and text, key region similarity is used to measure the difference between the printing clarity and color accuracy of the patterns and text on the card and the standard image, while non-key region similarity is used to evaluate the printing quality of non-key parts such as the card's background color. After the detection system obtains the first region segmentation result of the standard image data and the second region segmentation result of the printed product image data, it enters the printing quality similarity evaluation stage. Specifically, the detection system standardizes and organizes the first and second region segmentation results according to the input requirements of the printing similarity matching model, ensuring the accuracy of region information, pixel coordinates, and other data in both results. Subsequently, the detection system inputs the organized data into the preset printing similarity matching model. After receiving the data, the model first performs a pixel-by-pixel or feature-by-feature comparison of key regions in the two results, calculating multiple indicators such as color difference, shape matching degree, and texture similarity to comprehensively derive the key region similarity value; then, it analyzes and compares non-key regions in the same way to obtain the non-key region similarity value.

[0036] S205. The similarity of the key region and the similarity of the non-key region are weighted and calculated to obtain the comprehensive similarity. The overall similarity score is a comprehensive numerical value used to evaluate the overall print quality, obtained by the detection system through weighted calculation of the similarity scores of key and non-key regions. It takes into account the degree of influence of different parts of the printed material on the overall quality. For example, when evaluating the print quality of a color picture book, the key region similarity score reflects the print quality of the pictures and text in the book, the non-key region similarity score reflects the print quality of the background color of blank pages, borders, etc., while the overall similarity score is a comprehensive evaluation of the print quality of the entire book. After obtaining the similarity scores for key and non-key regions, the detection system performs a comprehensive evaluation of the print quality. Specifically, the system first obtains pre-set weighting parameters based on factors such as the type of printed material and the relative importance of key and non-key regions to quality. For example, for a promotional poster, since text and images are the core of the information conveyed, the weight for key regions might be set to 0.7, and the weight for non-key regions to 0.3. Then, the system calculates the similarity score by multiplying the key region similarity by its weight and the non-key region similarity by its weight, using a weighted calculation formula. The two products are then added together to obtain the overall similarity score.

[0037] S206. When the overall similarity is greater than or equal to the preset overall similarity threshold, the printing quality of the current printed matter is determined to be qualified.

[0038] The preset comprehensive similarity threshold is a standard value pre-set in the detection system to judge whether the printing quality is qualified. This threshold is determined based on factors such as the quality requirements of the printed matter and the level of production technology. For example, for high-precision art book printing, the preset comprehensive similarity threshold may be set to 0.9, while for ordinary promotional leaflet printing, the threshold may be 0.8. Qualified printing quality means that the actual printing effect of the current printed matter has a similarity to the standard image that meets or exceeds the preset standard, thus satisfying the quality requirements.

[0039] After calculating the overall similarity, the detection system assesses the printing quality. Specifically, the system compares the calculated overall similarity value with a preset overall similarity threshold. If the overall similarity is greater than or equal to the preset threshold, it indicates that the difference between the overall printing effect of the current printed product in critical and non-critical areas and the standard image is within an acceptable range, and the printing quality meets the expected requirements. At this point, the detection system determines that the printing quality of the current printed product is qualified, generates a qualified report, records relevant detection data, and sends a signal to the subsequent production process, allowing the printed product to enter the next stage such as packaging and warehousing. If the overall similarity is less than the preset threshold, the printing quality is determined to be unqualified, triggering the unqualified handling process, and the printed product is isolated, reworked, or otherwise processed.

[0040] By adopting the above technical solution, this application provides a printing quality inspection method. It establishes a direct correspondence between the target printing effect and the actual printing result by acquiring standard image data and printed product image data, providing a precise data foundation for subsequent quality inspection. The method uses a region recognition model to divide the image data into regions, performing differentiated processing on key and non-key regions while ensuring a one-to-one correspondence between them in pixel coordinate ranges, thus improving the precision of the inspection and the accuracy of region matching. Furthermore, a printing similarity matching model is used to calculate the similarity between key and non-key regions, accurately reflecting the differences in printing effects across different regions. Finally, a weighted calculation is performed to obtain a comprehensive similarity, which is then compared with a preset threshold to quickly determine whether the overall quality of the printed product is acceptable. This method effectively solves the problems of insufficient accuracy in printing quality inspection of complex printed content and imperfect handling of regional characteristic differences in related technologies, thereby improving the accuracy of printing quality inspection.

[0041] In the above embodiments, this application obtains standard image data of the printed matter and image data of the printed product, divides the area, calculates the similarity of corresponding areas, and judges the printing quality. In practical applications, the acquisition of image data of the printed product may result in unclear images, leading to misjudgments of printing quality. In some embodiments, the image quality can be evaluated after the acquisition of the image data of the printed product to avoid misjudgments of printing quality due to errors in the acquisition of the image data of the printed product.

[0042] The following provides supplementary information regarding the scenario in this embodiment. Based on the above scenario, the method provided in this embodiment will be described in more detail below. Please refer to... Figure 3 This is another flowchart illustrating the printing quality inspection method in this application.

[0043] S301. Obtain the standard image data and the finished printed image data corresponding to the current printed matter. The standard image data is the target printing result of the current printed matter, and the finished printed image data is the actual printing result of the current printed matter. Step S301 is similar to step S201 in the above embodiments, and will not be repeated here. Please refer to the description in the corresponding step.

[0044] S302. Based on the standard image data, perform a shooting quality assessment on the printed product image data to obtain a quality assessment value. The shooting quality assessment includes the detection of the resolution, contrast, and brightness distribution of the printed product image data. Shooting quality assessment refers to the comprehensive detection of resolution, contrast, and brightness distribution of printed image data. Resolution represents the density of pixels in the image, contrast is the ratio of the brightness of the brightest to the darkest parts of the image, and brightness distribution represents the distribution of brightness across different areas of the image. The quality assessment value is a numerical value obtained by quantifying the shooting quality assessment results using a specific algorithm; a higher value indicates better shooting quality of the printed image data.

[0045] After acquiring the standard image data and the finished printed image data corresponding to the current printed material, the inspection system performs an image quality assessment on the finished printed image data. Specifically, the system first detects the resolution of the finished printed image data by calculating the number of pixels per unit length or unit area and comparing it with a preset resolution standard to determine if the image is clear. Next, it detects the image contrast, analyzing the brightness difference between bright and dark areas to assess the vividness of colors. Then, it detects the brightness distribution by statistically analyzing the brightness values ​​of different areas of the image to determine if there is any uneven brightness. Finally, the system substitutes the detection results of resolution, contrast, and brightness distribution into a pre-set evaluation algorithm to calculate a quality assessment value, and compares this value with a preset quality assessment threshold. If the value is lower than the threshold, the finished printed image data is deemed unqualified.

[0046] S303. If the quality assessment value is lower than the preset quality assessment threshold, the printed image data is deemed unqualified, and a reshoot is immediately requested. The preset quality assessment threshold is a standard value pre-set in the detection system to measure whether the image data of the printed product is up to standard. This value is determined comprehensively based on factors such as actual printing quality inspection needs and equipment performance. A failure to meet the standard means that the currently acquired image data cannot meet the requirements for subsequent printing quality inspection. Once the detection system completes the quality assessment of the printed product image data and calculates the quality assessment value, it compares this value with a preset quality assessment threshold. Specifically, if the quality assessment value is lower than the preset threshold, the detection system determines that the currently acquired printed product image data is unqualified and immediately sends a re-shooting request to the imaging equipment responsible for capturing the printed product image, such as an industrial camera. Simultaneously, it records the unqualified information, including the shooting time, potential problems, etc., and adjusts the parameter settings of the imaging equipment. After receiving the re-shooting request, the imaging equipment will re-shoot the printed product, generating new printed product image data. The detection system then re-assesses the quality of the new data until the image data is qualified.

[0047] S304. Perform edge detection on the standard image data to obtain image edge features.

[0048] Edge detection refers to the operation of a detection system using mathematical algorithms to identify and extract the boundaries of objects in an image. By analyzing the changes in pixel grayscale values ​​in an image, it finds the locations where the grayscale values ​​change drastically; these locations are the edges of the image. Image edge features refer to the relevant information about the boundaries of objects in an image obtained through edge detection, including the length, direction, and curvature of the edges. These features can reflect the shape and structure of the object. For example, in an image with a printed trademark, edge detection can extract the boundary outline of the trademark pattern, and the shape, corners, and other information of the trademark boundary constitute the image edge features. After ensuring the printed product image data is qualified, the inspection system begins processing the standard image data to obtain key information for subsequent image analysis. Specifically, the inspection system uses mature edge detection algorithms such as the Canny algorithm and the Sobel operator to process each pixel in the standard image data. The algorithm calculates the grayscale changes of each pixel and its neighboring pixels. When the grayscale change exceeds a certain threshold, the point is considered to be located on the edge of the image. By traversing and calculating the entire standard image data, the inspection system can accurately extract the image edges.

[0049] S305. Based on the edge features of the image, perform feature correction on the printed product image data to generate corrected printed product image data. The feature correction includes rotation correction and scale normalization of the image.

[0050] Feature correction refers to the operation by which the detection system performs geometric transformations and resizing on the printed product image data based on the image edge features of the standard image data. The aim is to ensure that the printed product image data matches the standard image data in angle and size, facilitating accurate similarity comparison later. Rotation correction, a part of feature correction, involves the detection system calculating the tilt angle of the printed product image data and rotating it to the correct orientation, such as adjusting a tilted printed image to a horizontal or vertical position. Scale normalization is also a crucial step in feature correction. It involves the detection system enlarging or reducing the printed product image data according to the size specifications of the standard image data to match their dimensions, such as adjusting an oversized or undersized printed product image to the same resolution and size as the standard image. After acquiring the image edge features of the standard image data, the detection system begins processing the printed product image data to eliminate image deviations caused by factors such as shooting angle and distance. Specifically, the detection system first analyzes whether the printed product image data is tilted based on the image edge features. By calculating the angular deviation of the edges, it determines the tilt direction and angle of the image. Then, it uses rotation algorithms, such as affine transformation, to rotate the image to the correct angle, completing the rotation correction. Next, the detection system compares the size information of the standard image data, including resolution and aspect ratio, with the size differences of the printed product image data. Based on the differences, it selects an appropriate scaling algorithm, such as bilinear interpolation or bicubic interpolation, to scale the printed product image data, making the two consistent in size. Finally, it generates the corrected printed product image data, laying the foundation for subsequent accurate printing quality inspection.

[0051] S306. Obtain the standard resolution of the standard image data and the finished resolution of the corrected printed image data. Standard resolution refers to the pixel density specification set for the standard image data of the current printed matter. It measures the number of pixels per unit length or unit area of ​​the standard image, determining its sharpness and level of detail. Finished product resolution refers to the pixel density of the printed product image data after feature correction, reflecting the sharpness of the actual printed image. For example, the finished product resolution of an actual photographed and corrected printed image may differ from the standard resolution due to factors such as the performance of the photographing equipment and the correction process. The acquisition operation refers to the process by which the detection system extracts the standard resolution and finished product resolution data from storage locations or processing results. S307. Determine whether the standard resolution is consistent with the finished product resolution. Once the detection system obtains the standard resolution and the finished product resolution, it determines whether the two are equal. If they are equal, the subsequent region division step is performed; if they are not equal, the resolution is adjusted before proceeding with the next step.

[0052] S308. If they are inconsistent, the standard resolution and / or the finished product resolution shall be adjusted to make the standard resolution equal to the finished product resolution. When the consistency between the standard resolution and the finished product resolution is determined and the result is inconsistent, the detection system adjusts the resolution. Specifically, the system first analyzes the numerical differences between the standard resolution and the finished product resolution. If the standard resolution is higher than the finished product resolution, the system selects an appropriate magnification algorithm to process the printed image data. While increasing the number of pixels, it uses an interpolation algorithm to calculate the grayscale or color values ​​of the newly added pixels to ensure a smooth image transition. If the standard resolution is lower than the finished product resolution, the printed image data is reduced in size to remove redundant pixels, while using an algorithm to retain key image information. In some cases, the standard image data can also be adjusted accordingly, or both can be processed simultaneously, ultimately bringing the standard resolution and the finished product resolution to an equal state, preparing for subsequent print quality detection based on image region segmentation and similarity matching.

[0053] S309. Input the standard image data into the preset region recognition model to obtain the first region division result. The first region division result includes a first key region and a first non-key region. The first key region includes text, complex patterns and color transition content. The first non-key region includes solid colors, background and white space content. S310. Input the printed product image data into the region recognition model to obtain the second region division result. The second region division result includes a second key region and a second non-key region, and the pixel coordinate range of the second region division result corresponds one-to-one with the pixel coordinate range of the first region division result. S311. Input the first region division result and the second region division result into a preset printing similarity matching model to obtain the key region similarity and non-key region similarity. Steps S309 to S311 are similar to steps S202 to S204 in the above embodiments, and will not be repeated here. Please refer to the description of the corresponding steps.

[0054] In steps S309 and S310, a first key region and a first non-key region, a second key region and a second non-key region are obtained, respectively. The key regions (first key region and second key region) and the non-key regions (first non-key region and second non-key region) each include multiple image blocks. An image block refers to an image segment with different content in the corresponding region, such as text fragments, independent images, and white space.

[0055] S312. Based on the area of ​​each tile in the key area and non-key area, determine the area influence factor corresponding to each tile. The tile area represents the number of image pixels occupied by each tile, calculated by the total number of pixels contained within the tile. For example, a text tile consisting of 100×100 pixels has a tile area of ​​10,000 pixels. The area influence factor refers to the weighting coefficient assigned to each tile by the detection system according to specific rules based on its area size, used to measure the degree of influence of tile area on print quality assessment.

[0056] After completing the region segmentation and similarity determination of the standard image data and the printed product image data, the detection system determines the area influence factor of each patch. Specifically, the system first traverses all patches in both critical and non-critical regions, calculates the patch area and the total patch area by counting the number of pixels contained in each patch. The ratio of the patch area to the total patch area is then calculated and determined as the area influence factor for each patch, denoted as A. i .

[0057] S313. Determine the shape influence factor corresponding to each block based on the shape regularity of each block.

[0058] Shape regularity is a quantitative description of the regularity of a block's shape. It measures how closely a block's shape resembles a standard geometric shape (such as a rectangle or circle). The closer the shape is to a standard geometric shape, the higher the regularity. For example, a regular square text box block has a higher shape regularity than an irregular hand-drawn pattern block. Shape influence factor refers to the weighting coefficient assigned to each block by the detection system based on its shape regularity using a specific algorithm. It reflects the influence of block shape on print quality assessment. Blocks with high regularity generally have more stable reference value in print quality assessment, and their shape influence factors are also relatively higher.

[0059] After determining the area influence factor for each patch, the detection system determines the shape influence factor. Specifically, the system uses edge detection and contour fitting algorithms to analyze the shape of each patch. First, the edge contour of the patch is extracted, and then an attempt is made to fit it to a standard geometric shape. By calculating the difference between the actual contour and the fitted shape, the shape regularity is quantified. For example, the least squares method is used to calculate the error between the patch contour and the ideal rectangle; the smaller the error, the higher the shape regularity. Next, the detection system determines the shape influence factor for each patch, denoted as S, according to a pre-defined shape regularity-influence factor mapping table. i .

[0060] S314. Based on the shape influence factor and the area influence factor, adjust the similarity of the key region or the non-key region corresponding to each of the map tiles. After obtaining the shape and area influence factors for each tile, the detection system adjusts the similarity between key and non-key regions. This process comprehensively considers both tile area and shape factors to correct the original similarity calculation results, making the assessment more closely reflect actual printing conditions. Specifically, for each tile, the detection system weights and combines its shape and area influence factors, then multiplies this weighted combination with the original tile similarity value to obtain the adjusted similarity value. The detection system sequentially performs the above adjustment calculations on all tiles within both key and non-key regions, updating the similarity value for each tile to obtain more accurate key and non-key region similarity data.

[0061] S315. The adjusted similarity of the key region and the similarity of the non-key region are weighted and calculated to obtain the comprehensive similarity. After adjusting the similarity of key and non-key regions, the detection system calculates the overall similarity. Specifically, based on the adjusted similarity of the key and non-key regions, as well as the weight coefficient α for the key region and the weight coefficient β for the non-key region, the overall similarity is calculated using the following formula:

[0062] Where α+β=1 and α>β, X1 is the similarity of key regions, X2 is the similarity of non-key regions, and N k and N n The number of tiles in key areas and non-key areas are respectively, and the area influence factor A is... i satisfy Shape Influence Factor S i Satisfying 0≤S i ≤1.

[0063] S316. Determine whether the overall similarity is less than the preset overall similarity threshold.

[0064] S317. When the overall similarity is less than the preset overall similarity threshold, the printing quality of the current printed matter is determined to be unqualified. Printing quality failure means that the detection system, by comparing the overall similarity with the overall similarity threshold, concludes that the overall similarity between the printed image data and the standard image data does not meet the acceptable standard, which may include problems such as blurry text, incomplete patterns, and color deviation.

[0065] S318. Compare each of the blocks in the current printed matter one by one, and determine whether the similarity of the key area or the similarity of the non-key area corresponding to each block is lower than the average block similarity. The average block similarity is the ratio of the sum of the similarity of the key area and the similarity of the non-key area to the number of blocks.

[0066] After the inspection system determines that the printing quality of the current printed material is substandard, a detailed analysis of the block similarity is performed to accurately locate the specific area causing the quality problem. By comparing the similarity of each block with the overall average level, blocks with obvious defects are identified. Specifically, the inspection system first traverses each block in both critical and non-critical areas, obtaining the critical area similarity or non-critical area similarity value for each block. Then, based on the previously calculated critical area similarity, non-critical area similarity, and the total number of blocks, the inspection system calculates the average block similarity. Next, the inspection system compares the similarity value of each block one by one with the average block similarity to determine whether it is below the average level, thereby identifying which blocks may have printing defects. S319. If so, the corresponding block will be identified as a defect block and an alarm signal will be generated. The alarm signal includes the pixel coordinate range and defect type of the defect block. Among them, defective blocks refer to image blocks whose similarity is below the average level and indicates printing quality problems, as determined by the detection system after comparing the similarity of each block with the average block similarity. Examples include blocks with missing strokes in text or blocks with color deviations in patterns. Alarm signals are prompts generated by the detection system after identifying defective blocks. These signals are used to notify operators of the specific quality problem and include key information such as the pixel coordinate range and defect type of the defective block, enabling operators to quickly locate and resolve the issue.

[0067] The inspection system first assigns a unique identifier to each object identified as a defective block and records its pixel coordinate range in detail in the system's defective block record list. The specific location is determined by calculating the pixel coordinates of the top-left and bottom-right corners of the block. Simultaneously, the system analyzes the characteristics of the defective block based on pre-defined defect judgment rules to determine its defect type, such as comparing the color distribution of a standard image with the finished product image to determine if color distortion exists. Finally, the system integrates the pixel coordinate range and defect type of the defective block, generates an alarm signal, and sends it to operators or relevant quality control personnel through various means such as display screen prompts, SMS notifications, and pop-ups in the production management system, enabling timely processing of the corresponding printed materials.

[0068] S320. When the overall similarity is greater than or equal to the preset overall similarity threshold, the printing quality of the current printed matter is determined to be qualified.

[0069] By adopting the above technical solutions, this application provides a printing quality inspection method that, through image quality assessment of printed product image data, ensures from the source that the resolution, contrast, and brightness distribution of the image data meet the inspection requirements, avoiding distortion of inspection results due to low-quality data. By edge detection and feature correction of standard image data, it solves the deviation problem caused by differences in shooting angle or size, ensuring that the printed product image data and standard images are consistent in scale and geometric characteristics, laying an accurate foundation for subsequent comparison. Furthermore, by adjusting the area influence factor and shape influence factor of the image blocks to adjust regional similarity, it accurately quantifies the impact of each image block on the overall quality assessment, improving detection accuracy. By locating defective image blocks and generating alarm signals, it can quickly mark the location and type of printing defects, facilitating timely feedback and correction. This method effectively solves the detection deviation problems caused by insufficient data quality, image correction, and detection accuracy in related technologies, achieving high reliability, precision, and efficiency in printing quality inspection of complex printed content.

[0070] The printing quality inspection system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference needed]. Figure 4 This is a schematic diagram of the physical device structure of a printing quality inspection system in an embodiment of this application.

[0071] It should be noted that, Figure 4 The structure of the printing quality inspection system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0072] like Figure 4 As shown, the electronic device includes a Central Processing Unit (CPU) 401, which can perform various appropriate actions and processes based on a program stored in Read-Only Memory (ROM) 402 or a program loaded from storage portion 408 into Random Access Memory (RAM) 403, such as performing the methods described in the above embodiments. The Random Access Memory (RAM) 403 also stores various programs and data required for system operation. The CPU 401, ROM 402, and RAM 403 are interconnected via a bus 404. An Input / Output (I / O) interface 405 is also connected to the bus 404.

[0073] The following components are connected to the input / output (I / O) interface 405: an input section 406 including audio input devices, push-button switches, etc.; an output section 407 including displays, audio output devices, indicator lights, etc.; a storage section 408 including hard disks, etc.; and a communication section 409 including network interface cards such as LAN (Local Area Network) cards, modems, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output (I / O) interface 405 as needed. A removable medium 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 410 as needed so that computer programs read from it can be installed into the storage section 408 as needed.

[0074] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 409, and / or installed from removable medium 411. When the computer program is executed by the Central Processing Unit (CPU) 401, it performs the various functions defined in this application. It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0075] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0076] Specifically, the printing quality inspection system of this embodiment includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the printing quality inspection method provided in the above embodiment.

[0077] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the print quality inspection system described in the above embodiments; or it may exist independently and not assembled into the print quality inspection system. The storage medium carries one or more computer programs that, when executed by a processor of the print quality inspection system, cause the print quality inspection system to implement the print quality inspection method provided in the above embodiments.

[0078] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0079] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0080] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for printing quality inspection, characterized in that, include: Obtain standard image data and finished print image data corresponding to the current printed matter. The standard image data is the target printing result of the current printed matter, and the finished print image data is the actual printing result of the current printed matter. The standard image data is input into a preset region recognition model to obtain a first region division result. The first region division result includes a first key region and a first non-key region. The first key region includes text, complex patterns and color transition content, and the first non-key region includes solid color, background and white space content. The printed product image data is input into the region recognition model to obtain a second region segmentation result. The second region segmentation result includes a second key region and a second non-key region, and the pixel coordinate range of the second region segmentation result corresponds one-to-one with the pixel coordinate range of the first region segmentation result. The first region segmentation result and the second region segmentation result are input into a preset printing similarity matching model to obtain the key region similarity and non-key region similarity. The first critical region and the second critical region, the first non-critical region and the second non-critical region each include multiple corresponding tiles. Based on the tile area of ​​each tile in the key area and non-key area, determine the area influence factor corresponding to each tile; Based on the shape regularity of each block, the shape influence factor corresponding to each block is determined; the shape regularity is obtained by extracting the edge contour of the block, fitting the edge contour to a standard geometric shape, and calculating the difference between the actual contour and the fitted shape. Based on the shape influence factor and the area influence factor, adjust the similarity of the key regions or the similarity of the non-key regions corresponding to each of the map tiles; The adjusted similarity scores of the key regions and the non-key regions are weighted and calculated to obtain the comprehensive similarity score. The calculation formula is as follows: Among them, the weight coefficient α of the key area and the weight coefficient β of the non-key area satisfy α+β=1 and α>β. For key region similarity, For non-critical region similarity, and The number of tiles in key areas and non-key areas, respectively, and the area influence factor. satisfy Shape Influence Factor satisfy ; When the overall similarity is greater than or equal to the preset overall similarity threshold, the printing quality of the current printed matter is determined to be qualified. When the overall similarity is less than a preset overall similarity threshold, the printing quality of the current printed matter is determined to be unqualified. Each of the blocks in the current printed matter is compared one by one to determine whether the similarity of the key area or the similarity of the non-key area corresponding to each block is lower than the average block similarity. The average block similarity is the ratio of the sum of the similarity of the key area and the similarity of the non-key area to the number of blocks. If so, the corresponding block is identified as a defective block and an alarm signal is generated. The alarm signal includes the pixel coordinate range and defect type of the defective block.

2. The method according to claim 1, characterized in that, After acquiring the standard image data and the printed product image data corresponding to the current printed matter, the method further includes: Based on the standard image data, the image data of the printed product is subjected to a shooting quality assessment to obtain a quality assessment value. The shooting quality assessment includes the detection of the resolution, contrast and brightness distribution of the image data of the printed product. If the quality assessment value is lower than the preset quality assessment threshold, the printed finished product image data is determined to be unqualified, and a reshoot is immediately requested.

3. The method according to claim 1, characterized in that, After acquiring the standard image data and the printed product image data corresponding to the current printed matter, the method further includes: Edge detection is performed on the standard image data to obtain image edge features; Based on the image edge features, feature correction is performed on the printed product image data to generate corrected printed product image data. The feature correction includes rotation correction and scale normalization of the image.

4. The method according to claim 3, characterized in that, After performing feature correction on the printed product image data based on the image edge features to generate corrected printed product image data, the method further includes: Obtain the standard resolution of the standard image data and the finished resolution of the corrected printed product image data; Determine whether the standard resolution is consistent with the finished product resolution; If they are inconsistent, the standard resolution and / or the finished product resolution shall be adjusted to make the standard resolution equal to the finished product resolution.

5. A printing quality inspection system, characterized in that, The printing quality inspection system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the printing quality inspection system to perform the method as described in any one of claims 1-4.

6. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the printing quality inspection system, the printing quality inspection system performs the method as described in any one of claims 1-4.

7. A computer program product, characterized in that, When the computer program product is run on the printing quality inspection system, the printing quality inspection system performs the method as described in any one of claims 1-4.

Images