Printing material drying uniformity detection method based on infrared thermogram

Abstract

The invention relates to the field of image processing technology, and in particular to a method for detecting the uniformity of drying of printed materials based on infrared thermal images, comprising: obtaining a grayscale image of a printed material during a drying process, performing multiple regional growth on the grayscale image, determining the degree of disorder of the grayscale value of the growing region by the grayscale value of a pixel point in the current growing region each time the regional growth is performed, dividing the growing region into two parts by using a median grayscale value, determining the distribution uniformity of the grayscale value of the growing region according to the distance between the pixel points included in the two parts, dynamically updating the growth threshold of the growing region by using the distribution uniformity of the grayscale value and the disorder degree of the grayscale value, so as to realize adaptive regional growth, improve the accuracy of the regional growth result, and accurately evaluate the heating uniformity in the drying process based on the accurate regional growth result.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method for detecting the drying uniformity of printed materials based on infrared thermal images. Background Art

[0002] In the field of quality control in the printing industry, the uniformity detection of printed material drying is a key link to ensure product quality. In recent years, the detection method based on infrared thermograms has attracted much attention and has been widely used. This is mainly due to the continuous development of infrared thermal imaging technology. With the advancement of science and technology, this technology has become mature and can obtain the temperature distribution of printed materials in real time and efficiently during the drying process. The infrared thermograms converted from these temperature data provide a reliable data basis for drying uniformity detection.

[0003] The region growing algorithm is a common image segmentation algorithm, which is suitable for detecting whether the temperature distribution on the surface of an object is uniform. For example, the Chinese patent application document with the existing application publication number CN116883433A discloses a real-time monitoring system for the surface temperature distribution of a photovoltaic module, including: a segmentation module, which is used to perform image segmentation processing on the photovoltaic module image to obtain an initial segmented image; a determination module, which is used to determine the fastest growth direction and the slowest growth direction of a preset number of basic seed pixels; a calculation module, which is used to calculate the first grayscale change rate and the second grayscale change rate corresponding to the basic seed pixels; an estimation module, which is used to calculate the first cutoff distance and the second cutoff distance of the basic seed pixels in the fastest growth direction and the slowest growth direction; a confirmation module, which is used to confirm the number of newly added seed pixels in the slowest growth direction of the seed pixels; a monitoring module, which is used to perform region growing processing on the photovoltaic module image to obtain the final segmented image to monitor the temperature of the photovoltaic module, thereby improving the accuracy of temperature monitoring and reducing the working cost of temperature monitoring.

[0004] The above-mentioned region growing algorithm performs region growing based on a fixed growth threshold, divides the image into multiple growth regions, and then realizes the monitoring of temperature distribution. However, due to the characteristics of the infrared image of the printed material itself, its grayscale change is gradual, and the above-mentioned fixed threshold region growing algorithm cannot accurately identify this grayscale gradient, and will mistakenly divide the grayscale gradient region into the same region, so that the pixels with large grayscale differences are classified into the same category, which reduces the accuracy of region division, and then leads to the misjudgment of the drying uniformity of the printed material, affecting the quality control effect of the printed material. Summary of the invention

[0005] In order to solve the problem that the conventional region growing algorithm does not accurately divide the region of the printed material image, resulting in misjudgment of the drying uniformity of the printed material, the present invention proposes a method for detecting the drying uniformity of the printed material based on an infrared thermogram, comprising:

[0006] Obtain a grayscale image corresponding to the infrared thermal image of the printed material during the drying process, perform multiple regional growth on the grayscale image, evaluate the drying uniformity of the printed material according to the results of regional growth, and dynamically update the growth threshold for any current growth area during each regional growth, including:

[0007] Calculate the disorder of the gray value of the growing area based on the gray value sequence of the pixels in the growing area , and are the maximum and minimum values ​​of the grayscale value sequence, respectively. is the mean of the gray value sequence, is the current growth threshold of the growth area, is the maximum value function;

[0008] The pixels in the growth area are divided into the first part and the second part based on the median gray value of the pixels in the growth area, and the distribution uniformity of the gray value of the growth area is calculated. ; and The first part of pixels, the second part The mean Euclidean distance between each pixel and other pixels in its part, and The first part of pixels, the second part The mean Euclidean distance between each pixel and other pixels in the growth area, and are the total number of pixels in the first part and the second part respectively;

[0009] The distribution uniformity and disorder level are used to update the current growth threshold of the growth region.

[0010] This technical solution uses infrared thermal imaging technology to obtain the temperature distribution of printed materials in the drying process in real time, which can accurately reflect the temperature changes on the surface of the printed materials and convert it into a grayscale image. When performing regional growth on the grayscale image, the growth threshold of each growing area is dynamically adjusted according to the degree of disorder of the grayscale value of the growing area and the uniformity of the grayscale value distribution, so that the regional growth process can adapt to the grayscale characteristics of different areas. In areas where the difference in grayscale values ​​is small and the distribution of grayscale values ​​is relatively uniform, the growth threshold will be appropriately increased to expand the area. Conversely, in areas where the difference in grayscale values ​​is large and the distribution of grayscale values ​​is uneven, the growth threshold will be appropriately reduced to limit the expansion of the area, so as to more accurately identify and divide the grayscale gradient area, avoid the limitations of the fixed threshold algorithm, improve the accuracy of regional division, and accurately evaluate whether the drying of printed materials is uniform.

[0011] Preferably, the method for evaluating the drying uniformity of the printed material according to the result of the regional growth is:

[0012] The result of the region growing is to divide the grayscale image into The final growth area, is the total number of final growth areas; the average grayscale values ​​of the pixels contained in each of these growth areas are ,in, is the average gray value of the pixels contained in the first growth area, is the average gray value of the pixels contained in the second growth area, For the The average gray value of the pixels contained in the growth area;

[0013] according to The mean and variance of the pixels in each of these growth areas are used to calculate the coefficient of variation of the average grayscale value:

[0014]

[0015] In the formula, For this The coefficient of variation of the average grayscale value of the pixels contained in each growth area, yes The mean of yes The variance of

[0016] If the coefficient of variation is greater than a preset coefficient of variation threshold, the printed material is dried evenly; if the coefficient of variation is not greater than the preset coefficient of variation threshold, the printed material is dried unevenly.

[0017] This technical solution can effectively measure the relative discreteness of the average grayscale values ​​of different growth areas by calculating the coefficient of variation. The coefficient of variation is very sensitive to the discreteness of the data. Even when the average grayscale values ​​are similar, it can detect subtle unevenness, which helps to accurately and intuitively reflect the uniformity problems in the drying process and provide a reliable basis for process optimization.

[0018] Preferably, the method of updating the current growth threshold of the growth region by using the distribution uniformity and disorder degree is:

[0019] The product of the distribution uniformity and the disorder degree is used as the gray distribution index of the growth area, and the current growth threshold of the growth area is multiplied by the gray distribution index, and the obtained value is used as the updated growth threshold of the growth area.

[0020] This technical solution updates the growth threshold of the growing region by combining the distribution uniformity of the grayscale values ​​of the growing region and the degree of clutter of the grayscale values ​​of the growing region, so that the growth threshold can dynamically adapt to the grayscale characteristics of different regions, improve the adaptability of regional growth, and more finely control the regional growth process, thereby improving the precision and accuracy of the regional growth results.

[0021] Preferably, before performing multiple region growing operations on the grayscale image, the method further includes selecting multiple seed points in the grayscale image and setting an initial growth threshold.

[0022] Preferably, the grayscale value sequence of the pixel points in the growth area is obtained by statistically calculating the grayscale values ​​of all the pixel points in the growth area, and these grayscale values ​​are sorted in order from large to small to form a grayscale value sequence.

[0023] This technical solution arranges the grayscale values ​​of all pixels in the growth area in order, and the resulting grayscale value sequence can intuitively display the distribution of grayscale values ​​in the growth area in terms of numerical size, and can clearly see the range and concentration trend of the grayscale values. If the grayscale values ​​contained in the grayscale value sequence are relatively concentrated and do not change much, it means that the growth area is more likely to be heated evenly during the drying process; conversely, if the grayscale values ​​contained in the grayscale value sequence are relatively wide and vary greatly, it means that the growth area is more likely to be heated unevenly during the drying process.

[0024] Preferably, the first part is the pixel points in the growth area whose grayscale value is less than or equal to the median grayscale value, and the second part is the pixel points in the growth area whose grayscale value is greater than the median grayscale value.

[0025] Preferably, the method of selecting multiple seed points in the grayscale image is:

[0026] The grayscale range is obtained according to the maximum grayscale value and the minimum grayscale value of the grayscale image, and the grayscale range is divided into Gray levels;

[0027] Get all the pixels of each gray level, and select them in turn according to the preset selection rules. pixels, as the gray level seed points, a total of Seed points, and All are preset values;

[0028] The preset selection rule is: first select a pixel point closest to the edge of the image, and then continue The sum of the distances between the pixel point selected each time and all the pixel points that have been selected is the largest.

[0029] This technical solution divides the grayscale range into multiple levels and selects seed points from each level, ensuring that the seed points can cover different grayscale intervals in the image, improving the representativeness of the seed points. The selection rules ensure the uniform distribution of seed points in the image, avoiding the situation where the seed points are too concentrated or sparse, thereby improving the comprehensiveness and accuracy of regional growth.

[0030] Preferably, when performing the first region growing, the current growth thresholds of all growing regions are the initial growth thresholds.

[0031] The present invention has the following effects:

[0032] This solution can effectively improve the accuracy and reliability of printing material drying uniformity detection through an adaptive region growing algorithm based on infrared thermal images. By dynamically adjusting the growth threshold and comprehensively evaluating the grayscale distribution characteristics, it solves the problem of inaccurate grayscale gradient region division in the traditional region growing algorithm, which not only improves the accuracy of region division, but also improves the accuracy of drying uniformity judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0034] Figure 1 It is a schematic flow chart of the method of the present invention;

[0035] Figure 2 It is a schematic flow chart of the method for adaptive region growing of the present invention. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0037] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0038] Reference Figure 1 The printing material drying uniformity detection method based on infrared thermogram provided by the present invention comprises steps S1 to S3:

[0039] S1: Obtain a grayscale image of the printed material during the drying process.

[0040] In the printing industry, it is crucial to ensure that the temperature distribution of printed materials during the drying process is uniform. Uneven temperature distribution may lead to quality problems of printed materials, such as uneven drying, deformation or inconsistent color. In order to improve product quality and production efficiency, it is necessary to analyze the heat conditions of printed materials during the drying process.

[0041] In this step, a high-resolution infrared thermal imager is used to collect images of the drying process of the printed material to obtain an infrared thermal map of the printed material during the drying process. The infrared thermal map can reflect the temperature distribution of the printed material. In order to reduce the complexity of subsequent image processing, this step grays the infrared thermal map and converts it into a grayscale map. Grayscaling can retain the detailed features in the image while reducing the storage space required for the image. Filtering technology (such as median filtering, Gaussian filtering, etc.) is used to remove noise in the grayscale map to improve image quality. Contrast enhancement technology (such as histogram equalization) is used to improve the contrast of the image, making the temperature difference more obvious and facilitating subsequent analysis.

[0042] There is a corresponding relationship between the temperature value of the pixel in the infrared thermal map and the grayscale value of the pixel in the grayscale map. In the infrared thermal map, the higher the temperature of the pixel, the larger the grayscale value in the grayscale map. This is because the grayscale value is usually proportional to the temperature. Specifically, in the infrared thermal map, the area with higher temperature will correspond to a higher grayscale value, which means that in the grayscale map, these areas will appear brighter (greater grayscale value, close to white), and in the infrared thermal map, the area with lower temperature will correspond to a lower grayscale value, and in the grayscale map, these areas will appear darker (smaller grayscale value, close to black).

[0043] S2: Perform multiple adaptive region growing operations on the grayscale image to divide the grayscale image into multiple final growing regions.

[0044] When performing drying uniformity detection based on the grayscale image obtained in step S1, a region growing algorithm can be used to divide different regions of the image. The region growing algorithm starts from some initial seed points and gradually merges adjacent pixels with similar features (such as grayscale values) into the same region. However, in practical applications, since there are regions with temperature gradients in infrared thermal images, there are correspondingly many regions with grayscale gradients in the grayscale image. Traditional region growing algorithms may mistakenly divide these gradient regions into one large region, resulting in inaccurate region division.

[0045] Therefore, this step introduces an adaptive region growing method in the process of region growing. Specifically, each time a region grows, for any current growing region, the growth threshold of the growing region is dynamically updated to achieve adaptive region growing.

[0046] Among them, for dynamically updating the growth threshold of the growth area, this step is to continuously check whether the grayscale values ​​of the pixels in each growth area are uniform during the regional growth process. If the grayscale values ​​of the pixels in a certain growth area change greatly (that is, the grayscale values ​​are relatively chaotic), the growth threshold of the growth area is reduced so that it does not continue to grow. If the grayscale values ​​of the pixels in a certain growth area change little (that is, the grayscale values ​​are relatively uniform), the growth threshold of the growth area can be appropriately increased to continue growing. In this way, different growth areas can be divided more accurately to avoid merging gradient areas together incorrectly.

[0047] In order to perform adaptive region growing, this step first sets some basic conditions for region growing, including selecting multiple seed points in the grayscale image and setting an initial growth threshold.

[0048] The method for selecting seed points is:

[0049] In order to divide the grayscale image into several areas with different grayscale levels, first obtain the maximum grayscale value and the minimum grayscale value of all pixels in the grayscale image, obtain the grayscale range of the grayscale image, and divide the grayscale value range into This step sets , the value is the experience value;

[0050] For each grayscale level, select all pixels corresponding to the grayscale level in the grayscale image, and select from all pixels according to the preset selection rules The pixel points are used as the seed points of the grayscale level. The seed points in each grayscale level can represent the typical grayscale features in the level, thereby improving the accuracy and reliability of the region growing algorithm. This step sets , this value is an experience value.

[0051] The selection rule set here is: select the pixel point closest to the edge of the grayscale image among all the above pixels as the first seed point, select the pixel point farthest from the first seed point as the second seed point, select the pixel point with the largest sum of distances from the first two seed points as the third seed point, and repeat this process until the pixel point is selected. The selection rule ensures that the distance between the selected seed points is maximized, thereby increasing the diversity of seed points, which helps capture subtle grayscale changes in the image and reduces misjudgment. And first selecting the pixel points closest to the edge of the grayscale image as seed points helps to better handle image boundary conditions and ensure that the image boundary is also fully considered.

[0052] Implementers can adjust the number of grayscale levels and seed points according to actual conditions. According to the above operations, this step selects 8 seed points from each of the 8 grayscale levels, and a total of 64 seed points are obtained, ensuring that the seed points are evenly distributed throughout the grayscale range. This helps to more comprehensively cover different grayscale areas in the image and avoid missing important information. Through evenly distributed seed points and adaptive selection rules, it is possible to effectively reduce the situation where grayscale gradient areas are mistakenly merged into the same area, thereby improving the accuracy of area division.

[0053] Among them, the initial growth threshold is set , the implementer can set the initial growth threshold according to the actual situation. The function of the growth threshold is to include pixels in the growth area. For example, there is a seed point with a grayscale value of 100, and the growth threshold is set to 10, which means that among the adjacent pixels around the seed point, the pixels with grayscale values ​​between 90 and 110 are added to the growth area corresponding to the seed point. For example, if the grayscale value of an adjacent pixel is 95, because it differs from the grayscale value of the seed point by only 5 (less than the threshold 10), this pixel will be added to the growth area. If the grayscale value of an adjacent pixel is 85, it will not be added to the growth area because the difference exceeds the threshold.

[0054] For any growth area, the grayscale values ​​of all pixels it contains are similar. This method is often used in scenes such as image segmentation and object recognition.

[0055] After setting the seed point and the initial growth threshold, perform the adaptive region growing operation, refer to Figure 2 , including steps S21 to S23:

[0056] S21: During each region growth, for any current growth region, determine whether it can be expanded.

[0057] The specific judgment method is:

[0058] Define grayscale value interval , is the gray value of the seed point corresponding to the growth area, is the current growth threshold of the growth area.

[0059] Obtaining pixel points that are directly adjacent to the boundary of the growth region and have not been included in the growth region, and checking whether the grayscale values ​​of these pixel points fall within the grayscale value interval one by one, if there is at least one pixel point whose grayscale value falls within the grayscale value interval, the growth region can be expanded, then expanding the growth region, that is, including the pixel points whose grayscale values ​​fall within the grayscale value interval among these pixel points into the growth region, and then executing steps S22 to S23 after expansion;

[0060] If there is no pixel point whose grayscale value falls within the grayscale value interval, the growth region cannot be expanded, the region growth of the growth region ends, and the subsequent steps are no longer executed.

[0061] S22: Evaluate the disorder of the grayscale values ​​of the growth area and the distribution uniformity of the grayscale values ​​of the growth area.

[0062] In this step, the clutter of the grayscale values ​​of the growing area reflects whether the numerical distribution of the grayscale values ​​of all pixels in the growing area is uniform. The greater the clutter of the grayscale values ​​of the growing area, the more uneven the numerical distribution of the grayscale values ​​of all pixels in the growing area, and vice versa.

[0063] Provide a method to obtain the chaos of the gray value of the growth area:

[0064] Arrange the grayscale values ​​of all pixels in the growth area in descending order to obtain the grayscale value sequence of the growth area and calculate the disorder of the grayscale value of the growth area :

[0065]

[0066] In this formula, and are the maximum and minimum values ​​of the grayscale value sequence, respectively. is the mean of the gray value sequence, is the current growth threshold of the growth area, is the maximum value function.

[0067] In the formula Part of it reflects the symmetry and concentration of the grayscale values ​​of the pixels in the growth area, while Some of them take into account the influence of the current growth threshold. This comprehensive evaluation method helps to judge the uniformity of the area more accurately, thereby improving the accuracy of the overall detection.

[0068] If the grayscale values ​​of pixels in a growth area vary greatly, that is, If the grayscale value of the growth area is very large, it means that the grayscale value of the growth area is very different. For example, if a part of an image is a stripe with a strong black-white contrast, the grayscale value range of this part will be very large. is very small, which means that only pixels with very close grayscale values ​​can be included in the same area. This will lead to the formation of more small areas, and the grayscale value in each small area varies less, but there will be more different small areas overall. When the grayscale value range of the growing area is large and the current growth threshold of the growing area is very small, that is, Larger, at the same time If the value is higher, the grayscale value of the growth area will be very chaotic, which means that the grayscale value distribution in the area is very uneven, and there are many pixels with very different grayscale values.

[0069] If the gray value difference of pixels in a growth area is small, that is, is very small, indicating that the grayscale value distribution in this growth area is very uniform. For example, in the grayscale image, there is a part of the background with a very consistent color, and the grayscale value range of this part will be very small. If the current growth threshold of this growth area is set very large, it means that even pixels with some differences in grayscale values ​​can be included in the growth area, which will result in a small number of large growth areas in the grayscale image, and the grayscale values ​​of the pixels in the growth area vary greatly. When the grayscale value range of the growth area is very small and the growth threshold is very large, that is, Smaller, at the same time If the value is lower, the calculated clutter level will be very low, which means that the grayscale value distribution in the growth area is very uniform and there is no significant grayscale value difference.

[0070] The formula is calculated by calculating the maximum value of the gray value sequence of the growing area and minimum value , and the mean , which can effectively quantify the distribution of grayscale values ​​in the growing region. This quantification method not only considers the extreme value of the grayscale value, but also the average value of the grayscale value, and obtains the accurate degree of chaos of the grayscale value. By introducing the current growth threshold, the region growing process can adaptively adjust the growth threshold according to the actual grayscale value distribution, so as to better adapt to the local changes in the image and improve the accuracy and robustness of the region division.

[0071] In the process of region growing, if each growing region has a larger grayscale value range and a smaller growth threshold, the grayscale value of the growing region is more chaotic, because in this case the grayscale value distribution in the growing region is very uneven and can only contain pixels with very close grayscale values. If it has a smaller grayscale value range and a larger growth threshold, the grayscale value of the growing region is less chaotic, because in this case the grayscale value distribution in the growing region is very uniform.

[0072] This step reflects whether the distribution of the grayscale values ​​of all pixels in the growth area is uniform in physical position through the uniformity of the distribution of the grayscale values ​​in the growth area. The greater the uniformity of the distribution of the grayscale values ​​in the growth area, the more uniform the distribution of the grayscale values ​​of all pixels in the growth area is in physical position, and vice versa. In addition, the distribution of grayscale values ​​in physical position is characterized by the different positions of the pixels corresponding to the grayscale values ​​in the growth area.

[0073] Provide a method for obtaining the distribution uniformity of the grayscale value of the growth area:

[0074] The growth region is divided into a first part and a second part by using the median grayscale value of the pixels contained in the growth region. The first part is the pixels in the growth region whose grayscale value is less than or equal to the median grayscale value, and the second part is the pixels in the growth region whose grayscale value is greater than the median grayscale value. This division helps to evaluate the distribution of pixels in different grayscale ranges. The median grayscale value is the average value of the maximum grayscale value and the minimum grayscale value rounded off.

[0075] The distribution uniformity of the grayscale value of the growth area is calculated based on the distance between the pixels in these two parts. The calculation formula is:

[0076]

[0077] In the formula, is the distribution uniformity of the gray value in the growth area, The first part of the growth area The average Euclidean distance between the pixel and other pixels in the first part, and are the number of pixels contained in the first and second parts of the growth area, respectively. The second part of the growth area For the first or second part, if the average Euclidean distance between all pixels in a part is small, it means that the pixels in this part are relatively concentrated and the grayscale value does not change much, that is, the local uniformity of the grayscale values ​​corresponding to the pixels in this part is high.

[0078] In this formula, For the first part The average Euclidean distance between a pixel and other pixels in the growth area, For the second part For the first part or the second part, if the average Euclidean distance between the pixels in a certain part and other pixels in the entire area is large, it means that this part is very different from other parts of the entire growth area, and the overall uniformity of the grayscale values ​​corresponding to the pixels in this part is high.

[0079] In this formula, through To measure the consistency between the local uniformity of the grayscale values ​​corresponding to the pixels of the first part and the overall uniformity of the grayscale values ​​corresponding to the pixels of the first part, To measure the consistency between the local uniformity of the grayscale values ​​corresponding to the pixels of the second part and the overall uniformity of the grayscale values ​​corresponding to the pixels of the second part, finally, the consistency of the two parts is added and the average is taken to obtain the distribution uniformity of the grayscale values ​​of the growth area. , comprehensively quantify the distribution uniformity of the gray value of the growth area from the overall and local aspects, The larger the value of is, the more uniform the distribution of the grayscale value of the growth area is in the physical position. This method can more accurately evaluate the uniformity of the distribution of the grayscale value of the growth area, especially when processing gradient areas.

[0080] S23: Using the distribution uniformity and disorder level, the current growth threshold of the growth region is updated.

[0081] The product of the distribution uniformity of the grayscale values ​​of the growing area and the degree of disorder of the grayscale values ​​of the growing area is used as the grayscale distribution index of the growing area. The current growth threshold of the growing area is multiplied by the grayscale distribution index, and the obtained value is used as the updated growth threshold of the growing area. This means that in the growing area with high uniformity of grayscale value distribution and low disorder of grayscale value, the growth threshold is relatively high, thereby reducing missegmentation, while in the area with low uniformity of grayscale value distribution and high disorder of grayscale value, the growth threshold is relatively low, thereby better capturing details. This dynamic adjustment mechanism can better balance global and local information and improve the accuracy and consistency of regional growth results.

[0082] Further, after the growth threshold is updated, the process returns to step S21.

[0083] During the first region growing, the current growth threshold of all growing regions is the initial growth threshold. Then, for any growing region, each time region growing is performed, the operations of step S21-step S23 are executed. Step S21 is the judgment condition for the end of region growing. The operation of dynamically updating the growth threshold is implemented through step S22-step S23. Through this set of processes, the process of adaptive region growing of the grayscale image is realized. Finally, when all the growing regions end region growing, the grayscale image is divided into multiple final growth regions.

[0084] This adaptive region growing method improves the accuracy of the region growing results. For example, in the infrared thermal map, there is a region with gradually rising temperature. The traditional region growing algorithm uses a fixed growth threshold, and this gradually rising temperature region will be mistakenly divided into a large region. The adaptive region growing method of this scheme analyzes the change of grayscale value in this region during the region growing process. If the change of grayscale value is large, the growth threshold is automatically reduced, so that this region is further subdivided into several small regions, thereby more accurately reflecting the actual temperature distribution.

[0085] S3: Determine the drying uniformity of the printed material based on the coefficient of variation of the average gray value of the growth area.

[0086] Determining the drying uniformity of printed materials is to evaluate whether the printed materials are dried uniformly. Suppose the total number of final growth areas is The average grayscale values ​​of the pixels in these growth areas are ,in, is the average gray value of the pixels contained in the first growth area, is the average gray value of the pixels contained in the second growth area, For the The average gray value of the pixels contained in the growth area is The mean and variance of the pixels in each of these growth areas are used to calculate the coefficient of variation of the average grayscale value;

[0087] If the coefficient of variation is greater than the preset coefficient of variation threshold, it means that the printed material is heated unevenly during the drying process, that is, the printed material is dried unevenly; if the coefficient of variation is not greater than the preset coefficient of variation threshold, the printed material is heated evenly during the drying process, that is, the printed material is dried evenly. This step sets the coefficient of variation threshold to 5%, which is an empirical value.

[0088] For example, for 10 growth areas, the average grayscale values ​​of the pixels in each of the 10 growth areas are calculated, which are , and then calculate to The mean and variance of and , calculate the coefficient of variation:

[0089]

[0090] In the formula, is the coefficient of variation of the average grayscale value of the pixels contained in each of the 10 growth areas.

[0091] if , the printed material is heated evenly during the drying process. , then there is significant uneven heating of the printed material during the drying process, and the staff needs to be notified to make relevant adjustments to ensure the quality of the printed material.

[0092] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0093] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A printing material drying uniformity detection method based on infrared thermogram, characterized in that: include: Obtain a grayscale image corresponding to the infrared thermal image of the printed material during the drying process, perform multiple regional growth on the grayscale image, evaluate the drying uniformity of the printed material according to the results of regional growth, and dynamically update the growth threshold for any current growth area during each regional growth, including: Calculate the disorder of the gray value of the growing area based on the gray value sequence of the pixels in the growing area , and are the maximum and minimum values ​​of the grayscale value sequence, respectively. is the mean of the gray value sequence, is the current growth threshold of the growth area, is the maximum value function; The pixels in the growth area are divided into the first part and the second part based on the median gray value of the pixels in the growth area, and the distribution uniformity of the gray value of the growth area is calculated. , and The first part of pixels, the second part The mean Euclidean distance between each pixel and other pixels in its part, and The first part of pixels, the second part The mean Euclidean distance between each pixel and other pixels in the growth area, and are the total number of pixels in the first part and the second part respectively; The distribution uniformity and disorder level are used to update the current growth threshold of the growth region.

2. The printing material drying uniformity detection method based on infrared thermogram according to claim 1 is characterized in that: The method for evaluating the drying uniformity of printed materials based on the results of regional growth is: The result of the region growing is to divide the grayscale image into The final growth area, is the total number of final growth areas; the average grayscale values ​​of the pixels contained in each of these growth areas are ,in, is the average gray value of the pixels contained in the first growth area, is the average gray value of the pixels contained in the second growth area, For the The average gray value of the pixels contained in the growth area; according to The mean and variance of the pixels in each of these growth areas are used to calculate the coefficient of variation of the average grayscale value: ； In the formula, For this The coefficient of variation of the average grayscale value of the pixels contained in each growth area, yes The mean of yes The variance of If the coefficient of variation is greater than a preset coefficient of variation threshold, the printed material is dried evenly; if the coefficient of variation is not greater than the preset coefficient of variation threshold, the printed material is dried unevenly.

3. The printing material drying uniformity detection method based on infrared thermogram according to claim 1 is characterized in that: The method for updating the current growth threshold of the growth region by using the distribution uniformity and disorder degree is: The product of the distribution uniformity and the disorder degree is used as the grayscale distribution index of the growth area, and the current growth threshold of the growth area is multiplied by the grayscale distribution index to obtain a value as the updated growth threshold of the growth area.

4. The printing material drying uniformity detection method based on infrared thermogram according to claim 1 is characterized in that: Before performing multiple region growing operations on the grayscale image, the method further includes selecting multiple seed points in the grayscale image and setting an initial growth threshold.

5. The printing material drying uniformity detection method based on infrared thermogram according to claim 1 is characterized in that: The gray value sequence of the pixel points in the growth area is obtained by statistically calculating the gray values ​​of all the pixel points in the growth area, and these gray values ​​are sorted in order from large to small to form a gray value sequence.

6. The printing material drying uniformity detection method based on infrared thermogram according to claim 1 is characterized in that: The first part is the pixel points in the growth area whose grayscale value is less than or equal to the median grayscale value, and the second part is the pixel points in the growth area whose grayscale value is greater than the median grayscale value.

7. The printing material drying uniformity detection method based on infrared thermogram according to claim 4 is characterized in that: The method of selecting multiple seed points in the grayscale image is: The grayscale range is obtained according to the maximum grayscale value and the minimum grayscale value of the grayscale image, and the grayscale range is divided into Gray levels; Get all the pixels of each gray level, and select them in turn according to the preset selection rules. pixels, as the gray level seed points, a total of Seed points, and All are preset values; The preset selection rule is: first select a pixel point closest to the edge of the image, and then continue The sum of the distances between the pixel point selected each time and all the pixel points that have been selected is the largest.

8. The printing material drying uniformity detection method based on infrared thermogram according to claim 4 is characterized in that: When performing the first region growing, the current growing thresholds of all growing regions are the initial growing thresholds.

Images