Printing quality control method and inspection system of printing machine equipment

Abstract

The invention relates to the technical field of image data processing, in particular to a printing quality control method of printing machine equipment and an inspection system. The method comprises the steps that a surface image of printing paper is obtained, a gray-level co-occurrence matrix is used for detecting the surface image, the defect degree of the surface image is obtained, and printing quality control over printing machine equipment is achieved; in the process of detecting the surface image by using the gray-level co-occurrence matrix, a printing binary image is obtained and a broken line graph is constructed; calculating the side length of the sliding window; calculating the image feature importance degree of the sliding window when the sliding window slides on the surface image every time; and calculating the defect degree of the surface image. According to the invention, defects in the image can be captured more comprehensively, especially tiny printing quality problems and paper jitter; through the gray level co-occurrence matrix, the spatial information of the image is fully considered, the problem that the pixel spatial relation is neglected in a traditional method is solved, and the precision of printing quality control is improved.

Description

Technical Field

[0001] This invention relates to the field of image data processing technology, and in particular to a printing quality control method and inspection system for printing press equipment. Background Technology

[0002] As the core execution carrier of the plate printing process, printing press equipment occupies a central position in the large-scale production of the printing industry. Printing operations relying on printing plates offer the technical advantages of stable print quality and high printing speed. Printing press equipment can achieve high-precision image and text reproduction through printing plates. For printed materials with high requirements for image detail and color accuracy, it can accurately reproduce the lines, patterns, and color levels in the design draft, ensuring that each printed product achieves a near-consistent visual effect. This meets the stringent requirements for product quality consistency in large-scale production, and precise control of print quality has become a core demand for the large-scale operation of printing press equipment. In the industrial printing field, printing press equipment equipped with printing plates is an indispensable core piece of equipment for production operations. For example, in the packaging printing industry, the production of a large number of product packaging boxes, bags, and other categories relies on the plate printing operations of printing press equipment for efficient and stable processing. Its rapid production capacity can adapt to large-scale market demand, ensuring timely product supply to the market.

[0003] Patent application CN111401280A discloses an image recognition method based on adjusting the learning rate using a gray-level co-occurrence matrix. The method includes: acquiring athlete images through a camera; summarizing the acquired images; performing Gaussian filtering on the original image to obtain the incident image; subtracting the incident image from the original image to obtain the reflected image; updating the learning rate of the Gaussian mixture model based on the four-direction eight-neighbor gray-level co-occurrence matrix; and athlete extraction.

[0004] However, the aforementioned patent applications do not address the issue that during the process of acquiring surface images, printing equipment may experience defects due to various reasons during printing, such as plate wear, uneven ink distribution, and uneven printing pressure. These defects are reflected in the surface image, and the printing paper may vibrate irregularly as the printing roller rotates, resulting in a certain degree of blurring in the acquired image, affecting the judgment of surface image quality and thus hindering accurate printing quality control. Furthermore, the gray-level co-occurrence matrix mainly focuses on the gray-level value relationship between pixel pairs, neglecting the spatial relationship of pixels in the image. However, the spatial layout of pixels in the image is equally important for the description and recognition of texture features. Summary of the Invention

[0005] To address the issue that surface image quality deteriorates during image acquisition due to printing press malfunctions or paper vibration, and that grayscale co-occurrence matrix (GLCM) inspection ignores the spatial relationships of pixels in the image, thus reducing the accuracy of printing quality control, this invention provides a printing quality control method and inspection system for printing presses.

[0006] In a first aspect, the present invention provides a printing quality control method for printing press equipment, which adopts the following technical solution: A printing quality control method for printing press equipment includes: acquiring a surface image of printing paper; detecting the surface image using a gray-level co-occurrence matrix to obtain the defect degree of the surface image, thereby achieving printing quality control of the printing press equipment; during the detection of the surface image using the gray-level co-occurrence matrix: preprocessing the surface image to obtain a printing binary image; constructing a line graph based on the printing binary image, wherein the line graph includes a horizontal axis line graph and a vertical axis line graph; calculating the sliding value based on the average peak value, average valley value, mean width of the range corresponding to all peak values, mean width of the range corresponding to all valley values, and the maximum value among all peak values ​​of each line graph. The side length of the window; a sliding window is constructed based on the side length; based on the mean of the gradient magnitude, the variance of the gray value, and the information entropy of all pixels within the sliding window when the sliding window slides on the surface image each time, the importance of the image features of the sliding window when it slides on the surface image each time; based on the importance of the image features, the energy features in each direction when the sliding window slides on the surface image each time, and the mean of the energy features, the defect degree of the surface image is calculated; the energy features are the energy features of the gray-level co-occurrence matrix generated when the corresponding surface image within the sliding window matches pixels in the four directions of upward, downward, left, and right when the sliding window slides on the surface image each time.

[0007] This invention utilizes sliding window technology, combining gradient magnitude mean, grayscale variance, and energy characteristics in each direction, to more comprehensively capture defects in images, especially minor printing quality issues and paper jitter. Through the grayscale co-occurrence matrix, it fully considers the spatial information of the image, overcoming the problem of traditional methods neglecting pixel spatial relationships, thereby improving the accuracy of printing quality control. The fusion of multiple image features makes defect assessment more comprehensive and improves the sensitivity of quality detection. Automated processing reduces manual intervention, improves quality control efficiency, and is suitable for large-scale automated applications.

[0008] Furthermore, the preprocessing involves converting the surface image into a grayscale image, and then using a grayscale value threshold to convert the grayscale image into a printed binary image.

[0009] Furthermore, the method for obtaining the horizontal axis line chart and the vertical axis line chart is as follows: taking the horizontal dimension of the printed binary image as the horizontal axis and the vertical dimension as the vertical axis, the pixels in the printed binary image are projected onto the horizontal axis and the vertical axis respectively. The sum of the gray values ​​of the pixels projected onto each coordinate on the horizontal axis and the vertical axis is counted. The sums of the gray values ​​of the pixels at each coordinate on the horizontal axis and the vertical axis are connected in the coordinate order to obtain the horizontal axis line chart and the vertical axis line chart.

[0010] Furthermore, the side length satisfies the following relationship: In the formula, Let be the side length of the sliding window. and The first The average peak and average trough of the line chart. and The first The mean of the range width corresponding to all peaks and the mean of the range width corresponding to all troughs in the line chart. For the first The maximum value among all peaks in the line graph. To prevent hyperparameters with a denominator of 0.

[0011] This invention, by combining the peak and valley values ​​of a line graph with their corresponding range widths, can dynamically calculate the appropriate side length of a sliding window. This helps to flexibly adjust the window size according to different image characteristics, thereby improving the accuracy of image processing. The sliding window constructed based on the side length of the more adaptable sliding window can more accurately capture detailed changes in the image, reduce the possibility of missed detections and false detections, and improve the accuracy of print quality inspection.

[0012] Furthermore, the importance of the image features satisfies the following relationship: In the formula, For the sliding window on the surface image, the first Image feature importance during each slide For the sliding window on the surface image, the first The local information entropy of the pixel grayscale distribution within the sliding window during each sliding motion. For the sliding window on the surface image, the first The average gradient magnitude of all pixels within the sliding window during each sliding motion. For the sliding window on the surface image, the first The variance of the grayscale values ​​of all pixels within the sliding window during each slide. For normalization function, It is a natural exponential function.

[0013] This invention calculates the importance of image features by fusing local information entropy, gradient magnitude mean, and gray value variance. After normalizing and exponentially mapping the local information entropy, it can accurately distinguish between printed patterns and background areas, focusing on key detection areas with rich textures. Furthermore, the normalization result of the gradient magnitude mean is dynamically corrected by the gray value variance, effectively avoiding the problem of background edges being misjudged as high-importance areas and improving the accuracy of defect detection.

[0014] Furthermore, the defect degree satisfies the following relationship: In the formula, The defect degree of the surface image. For the sliding window on the surface image, the first Image feature importance during each slide , , , The sliding window is respectively The energy characteristics of upward, downward, leftward, and rightward movements during each slide. The mean of all energy characteristics. It is the hyperbolic tangent function.

[0015] This invention combines the energy features of an image in four directions to more comprehensively capture defects in surface images; the calculation of energy differences helps distinguish normal areas from defective areas, especially in areas with inconsistent textures or irregular changes in the image, which helps improve the accuracy of defect detection; the introduction of image feature importance as a weight allows the detection algorithm to dynamically adjust according to the feature importance of each area, focusing on areas more likely to have defects, thus improving the efficiency of print quality inspection.

[0016] Furthermore, the method for achieving printing quality control of the printing press equipment includes: If the defect level of the surface image exceeds the preset abnormal threshold, the printing quality of the printing equipment is deemed substandard, and an early warning is issued to complete the printing quality control inspection.

[0017] This invention can automatically detect whether the printing quality of printing equipment meets the standards by setting an abnormal threshold, without the need for manual intervention, thus improving the efficiency of quality control. When the defect rate exceeds the set threshold, the system will immediately issue an early warning to help operators take timely measures to prevent the problem from escalating.

[0018] Furthermore, the range width corresponding to the peak value and the range width corresponding to the valley value are obtained by determining the range based on the coordinates near the coordinates of the peak or valley value in the pixel distribution map where the difference between the sum of the gray values ​​corresponding to the coordinates of the peak or valley value and the peak or valley value is less than 0.1.

[0019] Furthermore, the normalization function adopts the maximum-minimum normalization function.

[0020] Secondly, the present invention provides a printing quality control system for printing press equipment, which adopts the following technical solution: A printing quality control system for a printing press includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the printing quality control method for the printing press is implemented.

[0021] By adopting the above technical solution, a computer program is generated from the printing quality control method of the above-mentioned printing machine equipment and stored in the memory so that it can be loaded and executed by the processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.

[0022] The present invention has the following technical effects: To address the limitation of gray-level co-occurrence matrices in representing the spatial information between pixels, a sliding window is constructed and slid across the surface image to obtain the gray-level co-occurrence matrix of the corresponding local region within the sliding window. During the generation of the gray-level co-occurrence matrix, multiple gray-level co-occurrence matrices are generated at different positions of the sliding window in the surface image. By utilizing the energy characteristics of these multiple gray-level co-occurrence matrices, the impact of printing paper jitter on plate quality inspection is reduced, improving the identification effect of surface image defects and enhancing the accuracy of plate inspection. Attached Figure Description

[0023] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts.

[0024] Figure 1 This is a flowchart of a printing quality control method for a printing press according to an embodiment of the present invention.

[0025] Figure 2 This is a flowchart of step S2 in a printing quality control method for a printing machine according to an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram of a horizontal axis broken line graph in a printing quality control method for a printing machine according to an embodiment of the present invention.

[0027] Figure 4 This is a schematic diagram of the vertical axis broken line graph in a printing quality control method for a printing machine according to an embodiment of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] The core of this invention is to construct a sliding window that slides across a surface image, restricting the calculation of the gray-level co-occurrence matrix to a local area of ​​the surface image. By generating multiple gray-level co-occurrence matrices at different positions on the surface image during the sliding window's movement, the texture features at different locations on the surface image are captured, which improves the ability to recognize local textures of the surface image to a certain extent. At the same time, based on the energy characteristics of all gray-level co-occurrence matrices, it can be analyzed whether the blurring of the surface image is caused by the shaking of the printing paper, thus better identifying whether there are defects in the surface image, and further determining whether the printing quality of the printing press meets the standards.

[0030] This invention discloses a printing quality control method for printing press equipment, referring to... Figure 1 This includes steps S001-S003: S001: Obtain a surface image of the printed paper.

[0031] S002: The surface image is detected using the gray-level co-occurrence matrix to obtain the defect degree of the surface image.

[0032] It should be noted that the color distribution varies in different areas of the surface image of printed paper. Therefore, the size of the sliding window needs to be determined based on the color distribution in the surface image so that the gray-level co-occurrence matrix generated by the sliding window can better reflect the characteristics of each position in the surface image. As the sliding window slides across the surface image, the color features and image gradient of the corresponding local image within the sliding window can reflect the importance of that local image. The gray-level co-occurrence matrix generated by the corresponding local image within the sliding window can express the blurring degree of that local image. The defect level of the entire surface image can be obtained by combining the importance and blurring degree of the local image.

[0033] In the process of detecting surface images using the gray-level co-occurrence matrix, refer to Figure 2 This includes steps S201-S204: S201: Obtain the printed binary image and construct a line chart.

[0034] The surface image is preprocessed to obtain a printed binary image. A line graph is then constructed based on the printed binary image, including a horizontal axis line graph and a vertical axis line graph.

[0035] Specifically, the preprocessing involves converting the surface image into a grayscale image, and then using a grayscale value threshold to convert the grayscale image into a printed binary image (containing only pixels with pixel values ​​of 0 and 255).

[0036] Implementers can set the grayscale threshold according to the specific implementation situation, for example, 110.

[0037] Specifically, the method for obtaining the horizontal axis line chart and the vertical axis line chart is as follows: Using the horizontal dimension of the printed binary image as the horizontal axis and the vertical dimension as the vertical axis, project the pixels of the printed binary image onto the horizontal and vertical axes respectively. Sum the grayscale values ​​of the pixels projected onto each coordinate on both axes. Then, connect the sums of the grayscale values ​​of the pixels at each coordinate on both axes in coordinate order to obtain the horizontal and vertical line graphs. (See [reference]). Figure 3 and Figure 4 .

[0038] S202: Calculate the side length of the sliding window.

[0039] It should be noted that the gray-level co-occurrence matrix generated by sliding the window across the surface image represents the characteristics of each region of the image. To ensure the generated gray-level co-occurrence matrix better reflects the characteristics of each region of the surface image, the size of the sliding window should ideally exclude the pattern region when sliding over the background area of ​​the printed paper, while allowing the surface image to occupy a larger proportion of the sliding window when sliding over the pattern region. Therefore, the size of the sliding window is determined based on the color distribution of the surface image in the printed paper. Since the pattern and background regions of the printed paper are distributed in blocks, these two types of regions can be roughly distinguished in the printed binary image. Then, by using the width of a certain range near the peak and the width of a certain range corresponding to the valley in the line graph corresponding to the printed binary image, the approximate size of each region can be obtained. When a peak in the horizontal or vertical line graph is higher, it indicates that the pixels near a row or column of the image corresponding to that peak have a high similarity, suggesting that the pixels in that row or column may be the background region in the surface image. Therefore, the approximate width or length of the background can be obtained. A lower valley value in a graph or vertical axis line chart indicates that the pixels near a row or column of printed paper corresponding to that coordinate also have a high degree of similarity. However, since the position with high similarity corresponds to the valley value in the line chart, it means that the pixels corresponding to that row or column may be a patterned area of ​​the surface image. Thus, the approximate width or length of the patterned area can be obtained. Based on the approximate width or length of the background and the approximate width or length of the patterned area, the size of the sliding window can be determined. Therefore, the size of the sliding window can be determined by connecting the sum of the gray values ​​of the pixels projected onto each coordinate on the horizontal and vertical axes.

[0040] The side length of the sliding window is calculated based on the average peak value, average valley value, average width of the range corresponding to all peak values, average width of the range corresponding to all valley values ​​for each line chart, and the maximum value among all peak values ​​for each line chart.

[0041] Specifically, the side length satisfies the following relationship: ; In the formula, Let be the side length of the sliding window. and The first The average peak and average trough of the line chart. and The first The mean of the range width corresponding to all peaks and the mean of the range width corresponding to all troughs in the line chart. For the first The maximum value among all peaks in the line graph. To prevent hyperparameters with a denominator of 0.

[0042] Implementers can set the step size and hyperparameters of the sliding window according to the specific implementation situation. For example, the step size is... The hyperparameters are .

[0043] in, for The corresponding weights are the average of all peaks in the horizontal axis line chart. It is a relatively large value, in order to pass Effectively regulate It needs to be passed adjust The effects, The principle is the same, so I will not repeat it here; and They represent the first The average width of the range corresponding to all peaks and the average width of the range corresponding to all valleys in the horizontal line graph. The width of the range corresponding to peaks and valleys is determined by the range of coordinates near the peak or valley in the pixel distribution map where the sum of the gray values ​​corresponding to the coordinates is less than 0.1 compared to the peak or valley value. For example, in the horizontal line graph, the sum of gray values ​​at coordinate 100 is the peak value, with a value of 10000. If the sum of gray values ​​corresponding to coordinate 101 is 9500, ... If the position at coordinate 101 falls within the range corresponding to the peak value, and the sum of the grayscale values ​​corresponding to coordinates 90 to 110 differs from the peak value by less than 0.1, while the sum of the grayscale values ​​corresponding to coordinates outside 90 to 110 differs from the peak value by more than 0.1, then the range corresponding to the peak value is [value missing]. The calculation of the range width corresponding to the valley value is similar to that corresponding to the peak value, and will not be repeated here.

[0044] S203: Calculate the importance of image features as the sliding window slides across the surface image for each time.

[0045] It should be noted that printed paper is usually composed of patterns and backgrounds of various colors. Since people tend to focus more on important areas such as patterns when observing paper, features generated by patterned areas in an image should be given higher importance and weight, while features generated by the background should be given relatively lower importance.

[0046] A sliding window is constructed based on the stated side length. The importance of image features is calculated based on the mean of the gradient magnitude, the variance of the grayscale value, and the information entropy of all pixels within the sliding window as the sliding window slides across the surface image each time.

[0047] Specifically, the importance of the image features satisfies the following relationship: ; In the formula, For the sliding window on the surface image, the first Image feature importance during each slide For the sliding window on the surface image, the first The local information entropy of the pixel grayscale distribution within the sliding window during each sliding motion. For the sliding window on the surface image, the first The average gradient magnitude of all pixels within the sliding window during each sliding motion. For the sliding window on the surface image, the first The variance of the grayscale values ​​of all pixels within the sliding window during each slide. For the normalization function, the minimum-max normalization function is used in this embodiment. It is a natural exponential function.

[0048] Among them, local information entropy is introduced. It can reflect the richness of image texture. When the sliding window is in the area of ​​the printed pattern, the texture is complex and the information entropy of the grayscale distribution is high. Higher, through The importance of features is amplified, but when the sliding window is in the background or has simple edges, the texture is monotonous and the information entropy is low. Lower The magnification effect is reduced, thus locking the printing plate pattern area and eliminating interference from the paper background. Since determining the importance of image features as the sliding window slides across the printing paper solely based on image gradients can lead to a higher importance of background features if the sliding window contains a complete background but its boundary lies precisely at the boundary between the background blank and the surface image, the variance of the grayscale values ​​of the corresponding pixels within the pattern area of ​​the sliding window is used instead. Using gamma transform to Make corrections when When it is larger, The smaller the window size, the greater the difference in grayscale values ​​of the corresponding pixels within the sliding window, the more complex the texture of the area within the window, and the more likely the area within the window is to be a pattern region. right The greater the expansion, the more... The smaller, The larger the value, the smaller the difference in grayscale values ​​of corresponding pixels within the sliding window, and the simpler the texture of the corresponding area within the window. This makes the area more likely to be the background. right The smaller the extent of expansion.

[0049] S204: Calculate the defect degree of the surface image.

[0050] It should be noted that the gray-level co-occurrence matrix of each region of the printed paper obtained through the sliding window can represent the energy characteristics of each region of the printed paper. Based on the energy characteristics, it can be determined whether the blurring phenomenon of the printed paper is caused by irregular shaking of the paper during printing. At the same time, the image features of different regions of the printed paper have different importance. Therefore, the defect degree of the surface image can be obtained based on the image feature importance corresponding to different regions of the printed paper and the energy characteristics of the gray-level co-occurrence matrix corresponding to different regions of the surface image.

[0051] Based on the importance of the image features, the energy features in each direction when the sliding window slides for each time, and the mean of the energy features, the defect degree of the surface image is calculated. The energy features are the energy features of the gray-level co-occurrence matrix generated when the sliding window slides on the surface image for each time, based on the matching of pixels in the four directions of up, down, left, and right.

[0052] Specifically, the defect degree satisfies the following relationship: ; In the formula, The defect degree of the surface image. For the sliding window on the surface image, the first Image feature importance during each slide , , , The sliding window is respectively The energy characteristics of upward, downward, leftward, and rightward movements during each slide. The mean of all energy characteristics. It is the hyperbolic tangent function.

[0053] in, The difference between the energy features of the gray-level co-occurrence matrix generated when matching pixels vertically (upward and downward) and the energy features of the gray-level co-occurrence matrix generated when matching pixels horizontally (leftward and rightward) is the result of vertical matching. The larger the difference, the greater the difference between the horizontal and vertical blurring in that area of ​​the surface image. In this case, the blurring is more likely to be caused by printing quality defects, and the image defect level should be higher. Conversely, the smaller the difference, the smaller the difference between the horizontal and vertical blurring in that area of ​​the surface image. In this case, the blurring is more likely to be caused by irregular shaking of the printing paper. Therefore, the more likely the paper printing is to be normal, the lower the image defect level should be. To comprehensively consider the average and difference values ​​of energy features in relation to the degree of image defects, the higher the energy of a gray-level co-occurrence matrix, the larger the value, and the more likely the corresponding image is to be blurry. In this case, the surface image is more likely to have printing quality defects, and the image defect degree should be higher. Conversely, the lower the energy of a gray-level co-occurrence matrix, the smaller the value, and the more likely the corresponding image is to be clear. In this case, the image defect degree should be lower. For the reaction sliding window During the second slide, the value of the defect degree of the image region corresponding to the sliding window is determined. If the sliding window... During each slide, the greater the importance of the features generated in the image region corresponding to the sliding window, the greater the impact of the defect degree of that part of the image region on the surface image defect degree.

[0054] S3: To achieve printing quality control of printing press equipment.

[0055] Specifically, the method for achieving printing quality control of printing press equipment includes: If the defect level of the surface image exceeds the preset abnormal threshold, the printing press may have insufficient printing plate or ink, indicating that the printing quality of the printing press is substandard. An early warning will be issued, prompting staff to inspect the printing press to avoid losses and waste of consumables, thus completing the printing quality control test.

[0056] Implementers can set an anomaly threshold based on the specific implementation situation, for example, 0.7.

[0057] This invention also discloses a printing quality control system for a printing press, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a printing quality control method for a printing press according to the present invention.

[0058] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0059] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for controlling printing quality in a printing press, characterized in that, include: The surface image of the printing paper is acquired, and the surface image is detected using a gray-level co-occurrence matrix to obtain the defect degree of the surface image, so as to achieve printing quality control of the printing press equipment. In the process of detecting surface images using the gray-level co-occurrence matrix: The surface image is preprocessed to obtain a printed binary image. A line graph is constructed based on the printed binary image, including a horizontal axis line graph and a vertical axis line graph. The side length of the sliding window is calculated based on the average peak value, average valley value, average width of the range corresponding to all peak values, average width of the range corresponding to all valley values, and the maximum value among all peak values ​​of each line graph. The sliding window is constructed based on the side length. The image feature importance of the sliding window at each slide on the surface image is calculated based on the average gradient magnitude, variance of gray values, and information entropy of all pixels within the sliding window at each slide on the surface image. The defect degree of the surface image is calculated based on the image feature importance, the energy features in each direction of the sliding window at each slide, and the average value of the energy features. The energy features are the energy features of the gray-level co-occurrence matrix generated when the corresponding surface image within the sliding window matches pixels in the four directions of upward, downward, left, and right at each slide on the surface image.

2. The printing quality control method for a printing press according to claim 1, characterized in that, The preprocessing involves converting the surface image into a grayscale image, and then using a grayscale value threshold to convert the grayscale image into a printed binary image.

3. The printing quality control method for a printing press according to claim 1, characterized in that, The method for obtaining the horizontal axis line chart and the vertical axis line chart is as follows: Using the horizontal dimension of the printed binary image as the horizontal axis and the vertical dimension as the vertical axis, the pixels in the printed binary image are projected onto the horizontal and vertical axes respectively. The sum of the gray values ​​of the pixels projected onto each coordinate on the horizontal and vertical axes is counted. The sums of the gray values ​​of the pixels at each coordinate on the horizontal and vertical axes are then connected in coordinate order to obtain the horizontal axis line chart and the vertical axis line chart.

4. The printing quality control method for a printing press according to claim 1, characterized in that, The side lengths satisfy the following relationship: ； In the formula, Let be the side length of the sliding window. and The first The average peak and average trough of the line chart. and The first The mean of the range width corresponding to all peaks and the mean of the range width corresponding to all troughs in the line chart. For the first The maximum value among all peaks in the line graph. To prevent hyperparameters with a denominator of 0.

5. The printing quality control method for a printing press according to claim 1, characterized in that, The importance of the image features satisfies the following relationship: ； In the formula, For the sliding window on the surface image, the first Image feature importance during each slide For the sliding window on the surface image, the first The local information entropy of the pixel grayscale distribution within the sliding window during each sliding motion. For the sliding window on the surface image, the first The average gradient magnitude of all pixels within the sliding window during each sliding motion. For the sliding window on the surface image, the first The variance of the grayscale values ​​of all pixels within the sliding window during each slide. For normalization function, It is a natural exponential function.

6. The printing quality control method for a printing press according to claim 1, characterized in that, The defect degree satisfies the following relationship: ； In the formula, The defect degree of the surface image, For the sliding window on the surface image, the first Image feature importance during each slide , , , The sliding window is respectively The energy characteristics of upward, downward, leftward, and rightward movements during each slide. The mean of all energy characteristics. It is the hyperbolic tangent function.

7. The printing quality control method for a printing press according to claim 1, characterized in that, The method for achieving printing quality control of printing press equipment includes: If the defect level of the surface image exceeds the preset abnormal threshold, the printing quality of the printing equipment is deemed substandard, and an early warning is issued to complete the printing quality control inspection.

8. A printing quality control method for a printing press according to claim 1 or 4, characterized in that, The method for obtaining the range width corresponding to the peak value and the range width corresponding to the valley value is as follows: The range is determined by the coordinates where the sum of the gray values ​​corresponding to the coordinates of the peak or valley value in the pixel distribution map differs from the peak or valley value by less than 0.

1.

9. A printing quality control method for a printing press according to claim 5, characterized in that, The normalization function is the minimum-maximum normalization function.

10. A printing quality control system for a printing press, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a printing quality control method for a printing press according to any one of claims 1-9.

Images