A method for detecting damage of printing scraper blades on-line

Abstract

The present application relates to the technical field of image data processing, and in particular to an online detection method for damage to a printing scraper. The method comprises: obtaining a surface image of a scraper to be detected, and for a target pixel in the surface image, determining a parameter value of the target pixel according to the pixel values ​​of other pixels in the neighborhood of the target pixel in the red channel; adjusting a preset blur radius according to the parameter value of the target pixel to obtain a target blur radius corresponding to the target pixel; graying the surface image to obtain a surface gray image, and using the target blur radius corresponding to the target pixel to perform unsharp mask (USM) enhancement processing on the surface gray image to obtain an enhanced image, and performing threshold segmentation on the enhanced image to obtain the rust area in the enhanced image. Through the above technical scheme, the difficulty of detecting damage to the scraper caused by rust can be reduced.

Description

Technical Field

[0001] The present application relates to the technical field of image data processing, and in particular to an online detection method for scraper damage used in printing. Background Art

[0002] Gravure printing is a printing method that engraves or etches grooves on the printing plate to carry the ink, and then transfers the ink to the substrate through pressure. During the gravure printing process, the excess ink on the surface of the printing plate can be scraped off by a scraper to ensure that the right amount of ink is transferred to the substrate; by adjusting the pressure and angle of the scraper, the thickness of the ink layer can be controlled, thereby adjusting the color density and clarity of the printed product; in addition, the scraper can also clean the surface of the printing plate to prevent ink accumulation on the surface of the printing plate.

[0003] Printing scrapers are usually made of metal. The surface of the scraper may rust due to being placed in a humid environment for a long time, causing damage to the scraper, thereby affecting the scraper's control over the ink and the color of the printed matter. Therefore, it is necessary to detect the rust on the surface of the scraper.

[0004] In order to detect possible rust on the surface of metal, in the relevant technology, for example, the Chinese patent application document with publication number CN118691554A provides a method for real-time detection of metal corrosion depth based on image recognition algorithm, including: obtaining a mesoscopic morphology image of the detection point; obtaining a corrosion depth sample of the detection point; establishing a mesoscopic database according to the mesoscopic morphology image and the corrosion depth of the detection point; the mesoscopic database is used to store the correspondence between the position of the detection point and the corrosion depth; obtaining a wide-angle morphology image of the detection point of the metal component; the wide-angle morphology image is a mesoscopic morphology image containing several detection points; constructing a hierarchical tree matching model, the hierarchical tree matching model is used to index the mesoscopic morphology image of the detection point in the wide-angle morphology image; and determining the corrosion depth according to the mesoscopic morphology image in the indexed mesoscopic database.

[0005] However, when collecting image samples of rusted printing scrapers, it is difficult to collect image samples of rusted scrapers in all cases, making it difficult to detect rust on the surface of the scraper and making it difficult to detect damage to the scraper caused by rust. Summary of the invention

[0006] In order to overcome the problem of difficulty in detecting damage to scrapers caused by rust in related technologies, the present application provides an online detection method for scraper damage for printing, including: obtaining a surface image of the scraper to be detected, and for a target pixel in the surface image, determining a parameter value of the target pixel according to the pixel values ​​of other pixels in the neighborhood of the target pixel in a red channel; adjusting a preset blur radius according to the parameter value of the target pixel to obtain a target blur radius corresponding to the target pixel; graying the surface image to obtain a surface grayscale image, and performing unsharp mask (USM) enhancement processing on the surface grayscale image using the target blur radius corresponding to the target pixel to obtain an enhanced image, and performing threshold segmentation on the enhanced image to obtain a rust area in the enhanced image; wherein the parameter value of the target pixel , norm is the normalization function; G is the difference between the target pixel and other pixels in the neighborhood; U is the number of other pixels in the neighborhood of the target pixel. is the pixel value of the u-th pixel in the red channel within the neighborhood of the target pixel. is the pixel value of the u-th pixel in the green channel within the neighborhood of the target pixel. It is the pixel value of the u-th pixel in the blue channel within the neighborhood of the target pixel.

[0007] In this way, a surface image of the scraper to be detected is obtained, and the preset blur radius is adjusted according to the parameter value of the target pixel in the surface image to obtain the target blur radius corresponding to the target pixel; since the parameter value of the target pixel is determined according to the pixel values ​​of other pixels in the neighborhood of the target pixel in the red channel, the pixel value of the rust on the red channel is larger, and therefore the target blur radius can better match the target pixel; the surface image is grayed to obtain a surface grayscale image, and the surface grayscale image is enhanced by unsharp masking (USM) using the target blur radius corresponding to the target pixel to obtain an enhanced image, in which the features of the rust in the enhanced image are more obvious, and threshold segmentation is performed on the enhanced image, so that the rust area in the enhanced image can be obtained more accurately, thereby reducing the difficulty of detecting rust on the surface of the scraper, thereby reducing the difficulty of detecting damage to the scraper caused by rust.

[0008] Optionally, adjusting the preset blur radius according to the parameter value of the target pixel point to obtain the target blur radius corresponding to the target pixel point includes: , where Z is the target blur radius corresponding to the target pixel, P is the preset blur radius, exp is an exponential function with a natural constant as the base, and V is the parameter value of the target pixel. To round up.

[0009] In this way, since the target blur radius corresponding to the target pixel point is obtained by adjusting the preset blur radius according to the parameter value of the target pixel point, the target blur radius corresponding to target pixels with different parameter values ​​is different. Compared with using the same blur radius for all target pixels, the target blur radius can be made to better match the actual situation of the target pixel point.

[0010] Optionally, the difference degree value is determined by: , G is the difference between the target pixel and other pixels in the neighborhood, norm is the normalization function, U is the number of other pixels in the neighborhood of the target pixel, is the gray value of the u-th pixel in the neighborhood of the target pixel, is the gray value of the target pixel, is the average gray value of all pixels in the neighborhood of the target pixel. To take the absolute value.

[0011] In this way, by comparing the grayscale value of the target pixel with the grayscale values ​​of other pixels in the neighborhood of the target pixel, the determined difference degree value can better characterize the degree of difference in pixel value between the target pixel and other pixels in the neighborhood; and the difference degree value will be based on the average grayscale value of all pixels in the neighborhood of the target pixel and the grayscale value of the target pixel as the denominator, and the difference degree value can also characterize the probability that the target pixel belongs to rust.

[0012] Optionally, the surface grayscale image is enhanced by unsharp mask (USM) using a target blur radius corresponding to the target pixel point to obtain an enhanced image, including: using the target blur radius corresponding to the target pixel point to blur the pixel point corresponding to the target pixel point in the surface grayscale image to obtain a blurred image after blurring the surface grayscale image; using the surface grayscale image to perform difference processing on the blurred image to obtain a difference image, and multiplying the pixel value of the pixel point in the difference image by a preset coefficient to obtain a weighted difference image; and performing image superposition processing on the weighted difference image and the surface grayscale image to obtain the enhanced image.

[0013] In this way, the contrast between the pixel points with a higher probability of being located at the rust and the pixel points with a lower probability of being located at the rust in the obtained enhanced image is enhanced, so that the rust area in the scraper can be determined more accurately according to the enhanced image.

[0014] Optionally, the target blur radius corresponding to the target pixel point is used to blur the pixel point corresponding to the target pixel point in the surface grayscale image, including: blurring the pixel point corresponding to the target pixel point in the surface grayscale image according to the target blur radius corresponding to the target pixel point and the target blur algorithm; wherein the target blur algorithm includes a Gaussian blur algorithm, a mean blur algorithm and a median blur algorithm.

[0015] Optionally, threshold segmentation is performed on the enhanced image to obtain the rust area in the enhanced image, including: obtaining a target threshold for threshold segmentation of the enhanced image; and taking an area composed of pixel points in the enhanced image whose grayscale values ​​are less than the target threshold as the rust area in the enhanced image.

[0016] In this way, the area composed of pixels whose grayscale values ​​in the enhanced image are less than the target threshold is taken as the rust area in the enhanced image, thereby effectively realizing the determination of the rust area on the scraper surface.

[0017] Optionally, the target threshold for performing threshold segmentation on the enhanced image is obtained in the following manner: the target threshold for performing threshold segmentation on the enhanced image is determined using the Otsu method.

[0018] Optionally, the surface image of the scraper to be detected is obtained in the following manner: an initial image including the scraper to be detected is obtained by an image acquisition device; the initial image is input into a pre-trained scraper area segmentation model to obtain the surface image of the scraper to be detected output by the scraper area segmentation model; the scraper area segmentation model is used to output the area image where the scraper to be detected is located in the initial image as a surface image.

[0019] The technical solution provided by the embodiments of the present application may include the following beneficial effects: obtaining a surface image of the scraper to be detected, adjusting a preset blur radius according to the parameter value of a target pixel in the surface image, and obtaining a target blur radius corresponding to the target pixel; since the parameter value of the target pixel is determined according to the pixel values ​​of other pixels in the neighborhood of the target pixel in the red channel, the pixel value of rust on the red channel is larger, and therefore the target blur radius can better match the target pixel; graying the surface image to obtain a surface grayscale image, and using the target blur radius corresponding to the target pixel to perform unsharp mask (USM) enhancement processing on the surface grayscale image to obtain an enhanced image, in which the features of rust in the enhanced image are more obvious, and threshold segmentation is performed on the enhanced image, so that the rust area in the enhanced image can be obtained more accurately, thereby reducing the difficulty of detecting rust on the surface of the scraper, thereby reducing the difficulty of detecting damage to the scraper caused by rust.

[0020] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0022] Figure 1 The present invention is a flow chart of a method for online detection of scraper damage for printing according to an exemplary embodiment. DETAILED DESCRIPTION

[0023] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0024] First, the application scenario of the embodiment of the present application is briefly introduced. In the application scenario of the present application, the printing scraper is usually made of metal. In order to determine whether the printing scraper is damaged by rust, the rust on the surface of the scraper can be detected.

[0025] When detecting rust on the surface of metal, the related technology mainly detects rust on the metal surface by collecting a large number of rusty metal surface images. However, it is difficult to collect rusty surface images of the scraper in all cases, which makes it difficult to detect rust on the scraper using the related technology, and it is also difficult to detect damage to the scraper caused by rust.

[0026] In view of the above technical problems, the present application provides an online detection method for scraper damage for printing. Figure 1 FIG. 1 is a flow chart of a method for detecting online scraper damage for printing according to an exemplary embodiment. Figure 1 As shown, the method includes the following steps.

[0027] In step S101, a surface image of the scraper to be detected is obtained, and for a target pixel in the surface image, a parameter value of the target pixel is determined according to the pixel values ​​of other pixels in the neighborhood of the target pixel in the red channel.

[0028] The surface image of the scraper to be inspected can be acquired by an image acquisition device. The surface image is a visible light image. The visible light image includes three channels, namely red, green and blue.

[0029] When there is rust in the surface image, since the rust mainly appears in red or reddish brown, and the color of the scraper itself made of metal is silvery white, the pixel value of the pixel point of the rust in the surface image in the red channel is larger than the pixel value of the pixel point of the scraper itself in the surface image in the red channel.

[0030] There is usually at least one other rust pixel around the rust pixel. Therefore, the parameter value of the target pixel in the surface image of the scraper can be determined based on this feature of the rust. The target pixel can be any pixel in the surface image.

[0031] Parameter value of the target pixel , norm is the normalization function; G is the difference between the target pixel and other pixels in the neighborhood; U is the number of other pixels in the neighborhood of the target pixel. is the pixel value of the u-th pixel in the red channel within the neighborhood of the target pixel. is the pixel value of the u-th pixel in the green channel within the neighborhood of the target pixel. It is the pixel value of the u-th pixel in the blue channel within the neighborhood of the target pixel.

[0032] There is usually a certain difference in the degree of rust at different positions in the rust area in the surface image, which makes different pixels in the rust area have certain differences in pixel values. The colors of the positions of the scraper surface image that are the color of the scraper itself are relatively close. Therefore, the pixel point with a greater difference in pixel value with other pixel points in the neighborhood in the surface image is more likely to be the pixel point where the rust area in the surface image is located.

[0033] The neighborhood range, for example, may refer to a range of sizes such as 7×7, 9×9, and 11×11 centered on the pixel point. By comparing the pixel values ​​of other pixels in the neighborhood range of the target pixel point in the red channel with the sum of the pixel values ​​of other pixels in the neighborhood range of the target pixel point in the three channels, the parameter value of the target pixel point can be easily controlled between 0 and 1.

[0034] The degree of difference in pixel value between the target pixel and other pixels in the neighborhood can be determined by comparing the pixel value of the target pixel with the pixel values ​​of other pixels in the neighborhood. The compared pixel value can be the pixel value in a certain channel or the pixel value after grayscale processing of the visible light image.

[0035] Here, since the parameter value of the target pixel is determined based on the pixel values ​​of other pixels in the neighborhood of the target pixel in the red channel, the rust in the printing scraper has a larger pixel value in the red channel of the surface image, and there are usually other pixels covered by the rust around the target pixel where the rust is located. Therefore, the parameter value of the target pixel can represent the probability that the target pixel in the surface image belongs to the rust in the scraper.

[0036] When there is dust on the surface of the scraper, there is a difference between the pixel values ​​of the dust pixels and the rust pixels in the red channel. Therefore, the parameter value of the target pixel is determined according to the pixel values ​​of other pixels in the neighborhood of the target pixel in the red channel. This can also avoid the influence of the dust on the scraper surface on the detection of the rust area.

[0037] In one embodiment, the difference between the target pixel and other pixels in the neighborhood can be determined in the following manner: , G is the difference between the target pixel and other pixels in the neighborhood, norm is the normalization function, U is the number of other pixels in the neighborhood of the target pixel, is the gray value of the u-th pixel in the neighborhood of the target pixel, is the gray value of the target pixel, is the average gray value of all pixels in the neighborhood of the target pixel. To take the absolute value.

[0038] In the surface image of the scraper, the color of the scraper itself is mainly silver, while the color of the rust in the surface image of the scraper is mainly red or reddish brown, and there are usually other pixels located on the rust around the target pixel located on the rust, so that the smaller the grayscale value of the target pixel in the surface image of the scraper, the greater the probability that the target pixel is located on the rust existing on the surface of the scraper, and the smaller the grayscale values ​​of other pixels around the target pixel, the greater the probability that the target pixel is located on the rust existing on the surface of the scraper.

[0039] The grayscale value of a pixel in the surface image of the scraper can be determined based on the pixel values ​​in the red, green and blue channels. For example, the average of the pixel values ​​in the red, green and blue channels can be used as the grayscale value of the pixel in the surface image.

[0040] In this way, by comparing the grayscale value of the target pixel with the grayscale values ​​of other pixels in the neighborhood of the target pixel, the determined difference degree value can better characterize the degree of difference in pixel value between the target pixel and other pixels in the neighborhood; and the difference degree value will be based on the average grayscale value of all pixels in the neighborhood of the target pixel and the grayscale value of the target pixel as the denominator, and the difference degree value can also characterize the probability that the target pixel belongs to rust.

[0041] In step S102, the preset blur radius is adjusted according to the parameter value of the target pixel point to obtain the target blur radius corresponding to the target pixel point.

[0042] The blur radius of a pixel can be used to blur the pixel. Since the pixel values ​​of pixels within different radii centered on the pixel are different, the blur radius used when blurring the pixel is different and the pixel values ​​of the referenced pixels may be different, resulting in different blurring effects on the pixels.

[0043] In the embodiment of the present application, when the preset blur radius is adjusted according to the parameter value of the target pixel point to obtain the target blur radius corresponding to the target pixel point, since the parameter value of the target pixel point can characterize the probability that the target pixel point belongs to rust, and it is difficult to accurately distinguish between the pixel points of rust in the scraper and the pixel points of the scraper itself only by relying on the parameter value of the target pixel, therefore, in order to better distinguish between the pixel points of rust in the scraper and the pixel points of the scraper itself, the details of the target pixel points with larger parameter values ​​can be retained, while the details of the target pixel points with larger parameter values ​​can be weakened. For this reason, the target blur radius corresponding to the target pixel point can be negatively correlated with the parameter value of the target pixel point.

[0044] In one embodiment, adjusting the preset blur radius according to the parameter value of the target pixel point to obtain the target blur radius corresponding to the target pixel point includes: , where Z is the target blur radius corresponding to the target pixel, P is the preset blur radius, exp is an exponential function with a natural constant as the base, and V is the parameter value of the target pixel. To round up.

[0045] By multiplying the inverse of the exponential function of the parameter value of the target pixel point by the preset blur radius, the inverse of the exponential function of the parameter value of the target pixel point is between 0 and 1, which can ensure that the target blur radius obtained after adjusting the preset blur radius is greater than 0, and the target blur radius obtained after adjusting the preset blur radius is less than or equal to the preset blur radius.

[0046] The existence of the rounding-up operator can ensure that the target blur radius obtained after adjusting the preset blur radius is a positive number; the preset blur radius can be set in advance according to actual needs, for example, the preset blur radius can be a positive integer such as 3 and 4.

[0047] In this way, since the target blur radius corresponding to the target pixel point is obtained by adjusting the preset blur radius according to the parameter value of the target pixel point, the target blur radius corresponding to target pixels with different parameter values ​​is different. Compared with using the same blur radius for all target pixels, the target blur radius can be made to better match the actual situation of the target pixel point.

[0048] And because the target blur radius corresponding to the target pixel is negatively correlated with the parameter value of the target pixel, the details of the pixels with a higher probability of rust in the grayscale image can be effectively retained.

[0049] In step S103, the surface image is grayed to obtain a surface gray image, and the surface gray image is enhanced by unsharp masking (USM) using a target blur radius corresponding to a target pixel point to obtain an enhanced image, and the enhanced image is threshold segmented to obtain a rust area in the enhanced image.

[0050] USM (Unsharp Masking) is an image enhancement technology that can improve image clarity by enhancing the edges of an image. By performing unsharp masking USM enhancement processing on a surface grayscale image, an enhanced image is obtained, which can better reflect rust in the enhanced image so that the rust area can be determined from the enhanced image.

[0051] Since the unsharp mask (USM) enhancement processing of the surface grayscale image is performed according to the target blur radius corresponding to the target pixel point, the target blur radius is determined according to the parameter value of the target pixel point, and the target parameter value can characterize the probability that the target pixel point belongs to rust, the image enhancement of the surface grayscale image can be better achieved according to the characteristics of rust.

[0052] In one embodiment, an unsharp mask (USM) enhancement process is performed on a surface grayscale image using a target blur radius corresponding to a target pixel point to obtain an enhanced image, including: using a target blur radius corresponding to a target pixel point to blur a pixel point corresponding to a target pixel point in a surface grayscale image to obtain a blurred image after blurring the surface grayscale image; using the surface grayscale image to perform a differential process on the blurred image to obtain a differential image, and multiplying the pixel values ​​of the pixel points in the differential image by a preset coefficient to obtain a weighted differential image; and performing an image superposition process on the weighted differential image and the surface grayscale image to obtain the enhanced image.

[0053] Since the blurring of the surface grayscale image is performed based on the target blurring radius corresponding to the target pixel point in the surface image, the details of the pixel points with a higher probability of being rusted in the surface grayscale image can be retained in the blurred image, and the blurred image is differentially processed using the surface grayscale image to obtain a differential image. In the differential image, the contrast between the pixel points with a higher probability of being rusted and the similar points with a lower probability of being rusted can be improved. Therefore, in the obtained enhanced image, the rust in the scraper can be further enhanced so as to determine the rust present in the scraper.

[0054] The preset coefficient can be selected according to actual needs. For example, the preset coefficient can be set to 1.2. Alternatively, a terminal device for online detection of scraper damage for printing can display a sliding control. When the sliding control hovers at different positions, it corresponds to preset coefficients with different values. The range of the preset coefficient can be limited to between 0.5 and 2. Users can achieve adaptive adjustment of the preset coefficient by hovering the sliding control at different positions.

[0055] In this way, the contrast between the pixel points with a higher probability of being located at the rust and the pixel points with a lower probability of being located at the rust in the obtained enhanced image is enhanced, so that the rust area in the scraper can be determined more accurately according to the enhanced image.

[0056] In one embodiment, a target blur radius corresponding to the target pixel is used to blur the pixel corresponding to the target pixel in the surface grayscale image, including: blurring the pixel corresponding to the target pixel in the surface grayscale image according to the target blur radius corresponding to the target pixel and a target blur algorithm; wherein the target blur algorithm includes a Gaussian blur algorithm, a mean blur algorithm and a median blur algorithm.

[0057] The Gaussian blur algorithm can reduce the noise in the image while blurring the image. The Gaussian blur takes into account the spatial relationship between pixels and smooths out the influence of noise by weighted averaging the surrounding pixel values, thereby improving the visual quality of the image. The mean blur algorithm reduces the noise in the image by simply averaging the pixels around the target pixel, so as to retain the details of the pixels with a higher probability of being rusted in the surface grayscale image. The median blur algorithm can effectively retain the edge information of the pixels in the surface grayscale image by selecting the median of other pixels around the pixel, and can reduce the noise in the surface grayscale image.

[0058] When using the Gaussian blur algorithm to blur the pixel points corresponding to the target pixel points in the surface grayscale image, a smaller Gaussian standard deviation such as 1 or 2 can be used to retain the details in the surface grayscale image.

[0059] In this way, the pixel points corresponding to the target pixel points in the surface grayscale image are blurred using any one of the Gaussian blur algorithm, mean blur algorithm and median blur algorithm and the target blur radius corresponding to the pixel points. Since the target blur radius is determined based on the parameter value of the pixel point corresponding to the pixel point in the surface image, it is easy to retain the details of the rust in the obtained enhanced image and weaken the details of the pixel points of the scraper itself.

[0060] In one embodiment, threshold segmentation is performed on the enhanced image to obtain the rust area in the enhanced image, including: obtaining a target threshold for threshold segmentation of the enhanced image; and taking an area composed of pixels in the enhanced image whose grayscale values ​​are less than the target threshold as the rust area in the enhanced image.

[0061] The color of the scraper itself is silver on the surface of the scraper, while the color of the rust that may exist on the surface of the scraper is red or reddish brown. In the enhanced image, the pixel value of the pixel located at the rust is lower than the pixel value of the pixel located at the scraper.

[0062] When the grayscale value of the pixel in the enhanced image is greater than or equal to the target threshold, the grayscale value of the pixel is closer to the color of the scraper itself after grayscale processing, and the probability that the pixel is located on the rust on the scraper surface is low. The pixel in the enhanced image with a grayscale value greater than or equal to the target threshold can be used as the pixel of the scraper itself.

[0063] On the contrary, when the grayscale value of the pixel in the enhanced image is less than the target threshold, the difference between the grayscale value of the pixel and the color of the scraper itself after grayscale processing is large, and the probability that the pixel is located on the rust on the scraper surface is high. The pixel with a grayscale value less than the target threshold in the enhanced image can be regarded as the pixel where the rust is located.

[0064] Among them, the target threshold can be pre-specified according to actual needs, for example, it can be set between 120 and 140. The target threshold can also be determined after multiple actual tests on the surface image of the scraper with rust. It can also be determined based on the pixel values ​​of the pixel points of the rust and other areas except the rust in the surface image of the scraper with rust.

[0065] In this way, the area composed of pixels whose grayscale values ​​in the enhanced image are less than the target threshold is taken as the rust area in the enhanced image, thereby effectively realizing the determination of the rust area on the scraper surface.

[0066] For example, when the rust area on the scraper surface is determined, the position information of the rust area on the scraper surface and the area of ​​the rust area can be sent to the rust removal mechanism, so that the rust removal mechanism adopts a corresponding rust treatment method according to the area of ​​the rust area, and goes to the position indicated by the position information to remove rust from the rust area; different rust treatment methods may include mechanical grinding, sand blasting, and spraying rust removers, etc.

[0067] In one embodiment, the target threshold for performing threshold segmentation on the enhanced image is obtained in the following manner: the target threshold for performing threshold segmentation on the enhanced image is determined using the Otsu method.

[0068] The Otsu method determines the target threshold based on minimizing the intra-class variance or maximizing the inter-class variance, which can determine a more accurate target threshold and make the obtained target threshold have a better segmentation effect.

[0069] In this way, since the obtained target threshold is more accurate, it is helpful to more accurately determine the rust area existing on the surface of the scraper.

[0070] Through the online detection method for damage to a printing scraper provided in an embodiment of the present application, a surface image of the scraper to be detected is obtained, and the preset blur radius is adjusted according to the parameter value of the target pixel in the surface image to obtain the target blur radius corresponding to the target pixel; since the parameter value of the target pixel is determined according to the pixel values ​​of other pixels in the neighborhood of the target pixel in the red channel, the pixel value of rust on the red channel is larger, and therefore the target blur radius can better match the target pixel; the surface image is grayed to obtain a surface grayscale image, and the surface grayscale image is enhanced by unsharp masking (USM) using the target blur radius corresponding to the target pixel to obtain an enhanced image, in which the features of the rust in the enhanced image are more obvious, and the enhanced image is thresholded, so that the rust area in the enhanced image can be obtained more accurately, thereby reducing the difficulty of detecting rust on the surface of the scraper, thereby reducing the difficulty of detecting damage to the scraper caused by rust.

[0071] In one embodiment, the surface image of the scraper to be detected is obtained in the following manner: an initial image including the scraper to be detected is obtained by an image acquisition device; the initial image is input into a pre-trained scraper region segmentation model to obtain the surface image of the scraper to be detected output by the scraper region segmentation model; the scraper region segmentation model is used to output the region image where the scraper to be detected is located in the initial image as a surface image.

[0072] The initial image of the scraper to be inspected captured by the image acquisition device may include other parts besides the scraper. For example, the initial image may include the background area of ​​the environment where the scraper is located. This background area may affect the detection device's judgment on the rust area.

[0073] In order to avoid the judgment of the rust area by the background area that may exist in the initial image, so as to obtain more accurate rust detection results and better realize the detection of damage caused by rust on the scraper, the initial image including the scraper can be annotated, and the pixel values ​​of the pixel points in other areas of the initial image except the scraper are marked as 0 to obtain the scraper area image of the area where the scraper is located in the initial image.

[0074] The initial image including the scraper captured by the image acquisition device can be used as the input of the pre-constructed network model, and the scraper area image can be used as the output of the pre-constructed network model. The pre-constructed network model is trained, and the trained network model is used as the scraper area segmentation model, so that the scraper area image of the area where the scraper is located in the initial image can be extracted using the scraper area segmentation model.

[0075] The model structure of the pre-built network model may be, for example, FCN (Fully Convolutional Network), MASK-R-CNN (Mask Region-based Convolutional Neural Network), and U-Net (U-type network).

[0076] In this way, the initial image including the scraper to be inspected is input into the pre-trained scraper area segmentation model to obtain the surface image of the scraper to be inspected, which can avoid the influence of pixels in other areas except the scraper in the initial image, thereby more accurately determining the rust area on the scraper surface.

[0077] It should be understood that the features of some embodiments of the various applications described herein may be combined with each other, unless specifically stated otherwise.

[0078] Although terms such as "first", "second", and "third" may be used herein to describe various components, parts, regions, layers, or sections, these components, parts, regions, layers, or sections are not limited to these terms. Instead, these terms are only used to distinguish one component, part, region, layer, or section from another component, part, region, layer, or section. Therefore, without departing from the teachings of the examples described herein, the first component, part, region, layer, or section mentioned in the examples may also be referred to as the second component, part, region, layer, or section.

[0079] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description herein, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0080] In addition, the word "exemplary" is used herein to mean serving as an example, instance, or diagram. Any aspect or design described herein as "exemplary" is not necessarily to be understood as being advantageous over other aspects or designs. Instead, the use of the word exemplary is intended to present concepts in a concrete manner. As used herein, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or".

[0081] Likewise, although the present application has been shown and described with respect to one or more implementations, equivalent variations and modifications will occur to those skilled in the art upon reading and understanding this specification and the accompanying drawings. With particular regard to the various functions performed by the components (e.g., elements, resources, etc.) described above, unless otherwise noted, the terms used to describe such components are intended to correspond to any component (functionally equivalent) that performs the specific functions of the described components, even if not structurally equivalent to the disclosed structures.

[0082] Additionally, while particular features of the present application may have been disclosed with respect to only one of several implementations, such features may be combined with one or more other features of other implementations as may be desired and advantageous for any given or particular application.

[0083] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary technical means in the art that are not disclosed in the present application, and the specification and embodiments are only considered as exemplary.

[0084] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. A method for detecting damage of a printing scraper on-line, characterized in that: include: Obtain a surface image of the scraper to be detected, and for a target pixel in the surface image, determine a parameter value of the target pixel according to pixel values ​​of other pixels in a neighborhood of the target pixel in a red channel; The preset blur radius is adjusted according to the parameter value of the target pixel point to obtain the target blur radius corresponding to the target pixel point; The surface image is grayed to obtain a surface gray image, and the surface gray image is enhanced by unsharp masking (USM) using the target blur radius corresponding to the target pixel point to obtain an enhanced image, and the enhanced image is segmented by threshold value to obtain the rust area in the enhanced image; Among them, the parameter value of the target pixel , norm is the normalization function; G is the difference between the target pixel and other pixels in the neighborhood; U is the number of other pixels in the neighborhood of the target pixel. is the pixel value of the u-th pixel in the red channel within the neighborhood of the target pixel. is the pixel value of the u-th pixel in the green channel within the neighborhood of the target pixel, is the pixel value of the u-th pixel in the blue channel within the neighborhood of the target pixel; Target blur radius , P is the preset blur radius, exp is an exponential function with a natural constant as the base, and V is the parameter value of the target pixel. To round up; Difference value , is the gray value of the u-th pixel in the neighborhood of the target pixel, is the gray value of the target pixel, is the average gray value of all pixels in the neighborhood of the target pixel. To take the absolute value; The target blur radius corresponding to the target pixel point is used to perform unsharp mask (USM) enhancement processing on the surface grayscale image to obtain an enhanced image, including: using the target blur radius corresponding to the target pixel point to blur the pixel point corresponding to the target pixel point in the surface grayscale image to obtain a blurred image after the surface grayscale image is blurred; using the surface grayscale image to perform difference processing on the blurred image to obtain a difference image, and multiplying the pixel value of the pixel point in the difference image by a preset coefficient to obtain a weighted difference image; and performing image superposition processing on the weighted difference image and the surface grayscale image to obtain the enhanced image.

2. The method for detecting damage of a printing scraper according to claim 1, characterized in that: Using the target blur radius corresponding to the target pixel, blurring is performed on the pixel corresponding to the target pixel in the surface grayscale image, including: According to the target blur radius and the target blur algorithm corresponding to the target pixel point, blurring is performed on the pixel point corresponding to the target pixel point in the surface grayscale image; The target fuzzy algorithm includes a Gaussian fuzzy algorithm, a mean fuzzy algorithm and a median fuzzy algorithm.

3. The method for online detection of scraper damage for printing according to claim 1, characterized in that: Perform threshold segmentation on the enhanced image to obtain the rust area in the enhanced image, including: Obtaining a target threshold for performing threshold segmentation on the enhanced image; The area composed of pixels whose grayscale values ​​in the enhanced image are less than the target threshold is taken as the rust area in the enhanced image.

4. The method for detecting damage of a printing scraper according to claim 3, characterized in that: The target threshold for threshold segmentation of the enhanced image is obtained as follows: The target threshold for threshold segmentation of the enhanced image is determined using Otsu's method.

5. The method for detecting damage of a printing scraper according to claim 1, characterized in that: The surface image of the scraper to be inspected is obtained by: Acquire an initial image including the scraper to be inspected by an image acquisition device; The initial image is input into a pre-trained scraper region segmentation model to obtain a surface image of the scraper to be detected output by the scraper region segmentation model; the scraper region segmentation model is used to output the region image where the scraper to be detected is located in the initial image as a surface image.

Images