Image printing processing method, device, equipment and storage medium

Abstract

The present invention discloses an image printing processing method, device, equipment and storage medium, which relates to the field of inkjet printing technology, and solves the problem of low efficiency of image detection and adjustment process in the prior art. The method comprises: obtaining a CMYK image to be printed; using a target CNN model to classify, identify and locate the CMYK image to be printed, and obtain N types of first feature images and the positioning position of each type of first feature image in the N types of first feature images; printing each type of first feature image in the N types of first feature images and sampling each type of first feature image after printing based on the positioning position to obtain N types of second feature images; detecting each type of second feature image in the N types of second feature images to determine target adjustment information; adjusting the CMYK image to be printed based on the target adjustment information and printing based on the adjusted CMYK image, so as to improve the image printing processing efficiency of the image detection and adjustment process.

Description

Technical Field

[0001] The present invention relates to the field of inkjet printing technology, and in particular to an image printing processing method, device, equipment and storage medium. Background Art

[0002] In the actual printing process in the field of inkjet printing, it is often necessary to make some adjustments to the image or the image printing process. For example, image parameters are adjusted after problems are found in image detection. Such a detection and adjustment process can improve image printing accuracy and thus improve the overall printing efficiency of image printing.

[0003] One of the traditional methods for image detection and adjustment is to use pre-designed reference patterns and patterns to be calibrated as calibration patterns. Normal production must be stopped to print calibration patterns. After printing, the scale in the printed calibration image must be manually checked to determine whether image parameters need to be adjusted. This method can only achieve offline detection. This method prints some non-target printed images on the substrate, wastes the substrate, increases costs, and destroys the automation process of the subsequent process.

[0004] In the prior art, the printed image can also be calibrated directly, but generally the machine is shut down after printing, and then an experienced technician finds some specific printed image areas, determines the parameter adjustment amount or changes the printing method through manual observation, and then restarts the machine. This detection and adjustment method is heavily dependent on human experience and has high work intensity and low efficiency. Summary of the invention

[0005] The object of the present invention is to provide an image printing processing method, device, equipment and storage medium for realizing inkjet printing image detection to improve image printing efficiency.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an image printing processing method, comprising:

[0008] Get the CMYK image to be printed;

[0009] The target CNN model is used to classify, identify and locate the CMYK image to be printed, and obtain N types of first characteristic images and the positions of the N types of first characteristic images; wherein N is a positive integer greater than 2, and the N types of first characteristic images at least include a first characteristic image of a color matching class, a first characteristic image of a broken hole class, a first characteristic image of a color uniformity class, and a first characteristic image of a driving waveform class;

[0010] Printing the CMYK image to be printed, and sampling the image areas corresponding to the positions of the N types of first characteristic images on the printed image to obtain N types of second characteristic images;

[0011] Detecting each type of second feature image in the N types of second feature images to determine target adjustment information;

[0012] The to-be-printed CMYK image is adjusted based on the target adjustment information, and printing is performed based on the adjusted CMYK image.

[0013] Compared with the prior art, the image printing processing method provided by the present invention directly inputs the acquired CMYK image to be printed into the target CNN model, thereby avoiding the problem of increased time cost caused by the color mode conversion of the image in RGB format in the traditional technology, and improving the image processing efficiency. Furthermore, before printing, the target CNN model is used to classify, identify and locate the CMYK image to be printed, which can save the process of manually designing reference patterns in advance, greatly reducing the time of image printing processing, and improving the operability and efficiency of image classification and positioning. In addition, when sampling is performed after printing the image, according to the positioning position of each type of first feature image, it is possible to accurately sample only the corresponding area of ​​each type of first feature image, without sampling the entire image, which can reduce the image sampling area and thus improve the sampling efficiency. After sampling, classification detection is performed to determine the target adjustment information, and printing is completed after adjustment according to the target adjustment information, ultimately improving the image printing processing efficiency of the image detection process as a whole.

[0014] In a second aspect, the present invention further provides an image printing processing device, comprising:

[0015] An acquisition module, used to acquire a CMYK image to be printed;

[0016] A classification module, used for classifying, identifying and locating the CMYK image to be printed by using a target CNN model, and obtaining N types of first characteristic images and the positioning position of each type of first characteristic image in the N types of first characteristic images; wherein N is a positive integer greater than 2, and the N types of first characteristic images at least include a first characteristic image of a color matching type, a first characteristic image of a broken hole type, a first characteristic image of a color uniformity type, and a first characteristic image of a driving waveform type;

[0017] A processing module, used for performing printing processing on the CMYK image to be printed, and performing sampling processing on each type of first characteristic image after printing based on the positioning position to obtain N types of second characteristic images;

[0018] A detection module, used to detect each type of second feature image in the N types of second feature images to determine target adjustment information;

[0019] A printing module is used to adjust the CMYK image to be printed based on the target adjustment information, and print based on the adjusted CMYK image.

[0020] In a third aspect, the present invention also provides an electronic device, comprising: a processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate via the communication bus; the memory stores a computer program that can be run by the processor; when the processor runs the computer program, it executes the above-mentioned image printing processing method.

[0021] In a fourth aspect, the present invention further provides a computer storage medium, wherein instructions are stored in the computer storage medium, and when the instructions are executed, the above-mentioned image printing processing method is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0023] Figure 1 A schematic flow chart of an image printing processing method provided by an embodiment of the present invention;

[0024] Figure 2 A schematic diagram of the structure of a target CNN model provided by one embodiment of the present invention;

[0025] Figure 3 A schematic diagram of the structure of an image printing processing device provided by an embodiment of the present invention;

[0026] Figure 4 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, words such as "first" and "second" are used to distinguish the same items or similar items with basically the same functions and effects. For example, the first threshold and the second threshold are only used to distinguish different thresholds, and their order is not limited. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit them to be different.

[0028] It should be noted that, in the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0029] In the present invention, "at least one" means one or more, "more than one" means two or more. "And / or" describes the association relationship of the associated objects, indicating that three types of relationships may exist.

[0030] like Figure 1 As shown, an embodiment of the present invention provides an image printing processing method, which may include:

[0031] Step 110: Obtain a CMYK image to be printed;

[0032] CMYK, also known as the printing color mode, is a color mode that relies on reflection. Similar to RGB, CMYK is the first letter of the names of four printing inks, among which C represents cyan, M represents magenta, Y represents yellow, and K represents black.

[0033] Step 120: using the target CNN model to classify, identify and locate the CMYK image to be printed, and obtain N types of first feature images and positions of the N types of first feature images; wherein N is a positive integer greater than 2, and the N types of first feature images at least include a first feature image of a color matching class, a first feature image of a broken hole class, a first feature image of a color uniformity class, and a first feature image of a driving waveform class;

[0034] There are also methods in the prior art that use machine learning models to identify input images, and the input images are RGB images, but the CNN model is more complicated and time-consuming to process RGB images. In the embodiment of the present invention, the CMYK image to be printed is directly input into the target CNN model, and the data processing speed is improved to a certain extent.

[0035] Specifically, Figure 2 As shown in the figure, the target CNN model includes at least an input layer (Input), three convolutional layers (Convolutinon), an activation function (Relu), two pooling layers (Pooling), a flattening layer (Flatten Layer), a fully connected layer (Full Connected Layer) and an output layer (Output).

[0036] use Figure 2 The target CNN model in step 120 is specifically described as follows:

[0037] Input the CMYK image to be printed into the input layer;

[0038] The input layer preprocesses the CMYK image to be printed and outputs the preprocessed image to the first convolutional layer of the three convolutional layers;

[0039] The first convolutional layer performs the first feature extraction on the preprocessed image, and uses the activation function Relu to output the image after the first feature extraction to the first pooling layer of the two pooling layers;

[0040] The first pooling layer performs the first downsampling on the image after the first feature extraction and outputs it to the second convolution layer of the three convolution layers;

[0041] The second convolutional layer performs a second feature extraction on the received image, activates it using the activation function Relu, and outputs it to the second pooling layer;

[0042] The image received by the second pooling layer is downsampled for the second time and output to the third convolutional layer. The third convolutional layer performs a third feature extraction on the received image until the third convolutional layer uses the activation function Relu to output the image after the third feature extraction to the flattening layer.

[0043] The flattening layer reduces the dimension of the image after the third feature extraction and outputs the reduced-dimensional image to the fully connected layer;

[0044] The fully connected layer classifies the reduced-dimensional image and outputs it to the output layer, which uses softmax as the target activation function to output N types of first feature images. The first feature image of the color matching class is suitable for color matching detection, the first feature image of the broken hole class is suitable for broken hole (broken needle) detection, the first feature image of the color uniformity class is suitable for color uniformity detection, and the first feature image of the driving waveform class is suitable for waveform detection.

[0045] The target CNN model is trained by the initial CNN model constructed by the classified image sample set. The training process can refer to the relevant technology and will not be described here one by one.

[0046] Step 130: Printing the CMYK image to be printed, and sampling the image areas corresponding to the positions of the N types of first characteristic images on the printed image to obtain N types of second characteristic images;

[0047] It is understandable that after the target CNN model is used to classify, identify and locate the CMYK image to be printed, each type of feature image obtained is a part of the entire image to be printed. In the traditional technology, the entire image area of ​​the printed image is sampled during sampling processing, while in the present invention, only each type of first feature image can be sampled according to the position of the N types of first feature images on the printed image, so the image data processing amount in the sampling processing step is greatly reduced. The advantages of this are, first, to increase the sampling rate of the sampling processing step, and second, to improve the overall processing efficiency of the entire detection process.

[0048] Step 140: Detect each type of second feature image in the N types of second feature images to determine target adjustment information;

[0049] Specifically, the target adjustment information may be a parameter to be adjusted and a parameter value, or the target adjustment information may be an adjustment strategy.

[0050] It is understandable that the first characteristic image of the color matching class can be used to detect whether the position of each color in the image is accurate. If the position is inaccurate, the position is corrected.

[0051] The first characteristic image of the broken hole type can be used to determine whether there is a broken hole in the nozzle. The broken hole can be a nozzle with an ink ejection amount of 0, or a nozzle with an ink ejection amount less than a preset ink amount.

[0052] The first characteristic image of the color uniformity class can be used to determine the color uniformity of the image.

[0053] The first characteristic image of the driving waveform can be used to determine whether the driving waveform needs to be adjusted.

[0054] Step 150: Adjust the to-be-printed CMYK image based on the target adjustment information, and print based on the adjusted CMYK image.

[0055] From the above content, it can be seen that in the embodiment of the present invention, firstly, the acquired CMYK image to be printed is directly input into the target CNN model, which avoids the problem of increased time cost caused by the color mode conversion of the image in RGB format in the traditional technology, thereby improving the image processing efficiency; secondly, the target CNN model is used to classify, identify and locate the CMYK image to be printed before printing, which can save the process of manually designing reference patterns in advance, greatly reduce the time of image printing processing, and improve the operability and efficiency of image classification; thirdly, when sampling after the printed image is completed, only the designated sampling of the corresponding area of ​​each first feature image can be performed according to the positioning position, instead of sampling the entire image, which can reduce the sampling area and thus reduce the amount of data processing in the sampling. The subsequent classification and detection steps can also reduce the amount of data processing, thereby improving the overall printing processing efficiency of the image printing process in the inkjet printing field.

[0056] It should be noted that step 120, i.e., the process of classifying, identifying and locating the CMYK image to be printed using the target CNN model, can be processed offline or online. If offline processing is performed, that is, step 120 is performed separately before starting printing, and then the data obtained after processing is saved to the database. When actually printing, the data is directly searched in the database according to the printed file, and then directly called, which improves performance. If the processing is performed online, the obtained data is also saved to the database, so that when printing the same file, it can be directly called to improve performance. That is to say, in an embodiment of the present invention, the target CNN supports both offline mode and online mode.

[0057] Regarding the beneficial effects of the present solution, it should also be noted that the background technology mentions that the prior art uses machine learning methods to replace the process of manual detection and adjustment, but these solutions all perform classification, identification, positioning and adjustment after printing the image. The defects of this are: 1) The prior art is based on a large amount of printed image data when training the model, and the present application is trained for the CMYK image to be printed, which is more time-saving; 2) The prior art directly samples the printed image for analysis, but deviations will occur during the printing process, and it is difficult for CNN to determine whether it is the deviation in the printing process or the effect required by the image itself, so it is very easy to cause misjudgment, while the present application first analyzes the classified feature image, which is based on the classification of the correct effect, and then detects and analyzes the actual printing effect. In this way, if the sampled data has deviations, it is the place that needs to be adjusted, avoiding the possibility of misjudgment and improving image printing efficiency; 3) The prior art directly samples the printed image for analysis, but if the original print file uses a spot color channel, such as CMYK S1S2, it is impossible to obtain the S1 and S2 channel adjustment information by analyzing the printed image, because the spot color has been integrated into CMYK after printing. It should be noted that spot color refers to specific color ink used to achieve specific effects, such as CMYKS1S2, S1 light blue, S2 light red, etc. Once this color is printed, it cannot be separated when visually captured. The visually captured image can only be converted to CMYK channels, and the S1 and S2 channel information is lost.

[0058] In an optional implementation, the N types of second characteristic images include a color-coded second characteristic image; the color-coded second characteristic image includes M sub-characteristic images; wherein M is a positive integer greater than 2;

[0059] Step 140 may specifically include the following steps:

[0060] 1) Obtain the position weight of each nozzle on the nozzle;

[0061] 2) determining the first position of the color to be detected in each sub-feature image;

[0062] 3) determining a second position of a reference color of the color to be detected in each sub-feature image;

[0063] 4) determining a position difference between the first position and the second position as an initial position correction value;

[0064] 5) Substitute the initial position correction value and the position weight of the nozzle corresponding to the color to be detected on each sub-feature image into the formula:

[0065]

[0066] The target position correction value of the color to be detected is calculated; wherein Δx1 is the initial position correction value of the color to be detected in the first sub-feature image, and W1 is the position weight of the nozzle corresponding to the color to be detected in the first sub-feature image; Δx2 is the initial position correction value of the color to be detected in the second sub-feature image, and W2 is the position weight of the nozzle corresponding to the color to be detected in the second sub-feature image; Δx M is the initial position correction value of the color to be detected in the Mth sub-feature image, W M is the position weight of the nozzle corresponding to the color to be detected in the Mth sub-feature image; ΔX is the target position correction value.

[0067] It is understandable that the principle of overprinting is based on color superposition, and rich color effects can be achieved by accurately controlling the printing order and position of each color. In the printing process, black is usually ranked first in the color sequence to ensure the accuracy of overprinting. Therefore, in the embodiment of the present invention, black is used as a reference, and the other three colors (cyan, magenta and yellow) in CMYK are detected as the colors to be detected.

[0068] For example, if the color to be detected is yellow, the second characteristic image of the color matching class includes, for example, three sub-characteristic images; a machine vision algorithm can be used to first determine the initial position difference (initial position correction value) Δx1 between yellow and black in the first sub-characteristic image, and determine the nozzle and nozzle position weight W1 corresponding to yellow in the first sub-characteristic image; secondly, in the second sub-characteristic image, determine the initial position difference (initial position correction value) Δx2 between yellow and black, and determine the nozzle and nozzle position weight W2 corresponding to yellow in the second sub-characteristic image; thirdly, in the third sub-characteristic image, determine the initial position difference (initial position correction value) Δx3 between yellow and black, and determine the nozzle and nozzle position weight W3 corresponding to yellow in the second sub-characteristic image; finally, according to formula (1), the target position correction value of yellow is ΔX. Based on this target position correction value, the position parameter of yellow is adjusted in the CMYK image to be printed to improve the color matching accuracy.

[0069] It should be noted that in the prior art, since there is only one reference image before printing, when the printed image is compared with the reference image for detection, only one comparison can be performed to obtain a position difference. However, if the reference image itself may have errors, in this case, no matter how the comparison and detection are performed, the error of the correction value obtained will be large. However, in the implementation of the present invention, the reference image is a plurality of sub-feature images, each of which can obtain an initial position correction value. The plurality of sub-feature images can play a self-correction process according to the above process, which can reduce the error caused by a reference image with a high probability. If the error is reduced, the image printing accuracy can be improved. If the accuracy is improved, the number of detection and adjustment can be reduced, and finally the overall image printing efficiency is improved.

[0070] In a specific implementation, the second characteristic image of the color-trapping class and the first characteristic image of the color-trapping class are used as the characteristic images of the color-trapping class, and the characteristic images of the color-trapping class can be specifically divided into a straight pattern, a |-shaped pattern, and a cross-shaped pattern. The straight pattern can be used for color-trapping detection in the y direction, the |-shaped pattern can be used for color-trapping detection in the x direction, and the cross-shaped pattern can be used for color-trapping detection in both the y and x directions.

[0071] In an optional implementation, the N types of second characteristic images include broken hole type second characteristic images;

[0072] Optionally, the algorithm of the above steps may adopt a color matching algorithm.

[0073] Step 140 may use a color matching algorithm to perform the following calculation:

[0074] Determine a first color value corresponding to the position to be detected in the second characteristic image of the broken hole type;

[0075] Determine a second color value corresponding to the position to be detected in the first characteristic image of the broken hole type;

[0076] Based on the color difference between the first color value and the second color value, determining whether the nozzle hole corresponding to the position to be detected is a broken hole;

[0077] If the nozzle hole corresponding to the position to be detected is a broken hole, the broken hole compensation strategy in the target adjustment information is adopted; the broken hole compensation strategy is to use the normal nozzle holes adjacent to the broken hole to perform ink volume compensation operation on the broken hole.

[0078] In a specific implementation, if the color channel difference between the first color value and the second color value is greater than a preset difference, the position to be detected is determined to be a broken hole. For example, if the value of the M channel in the first color value is 255, the value of the M channel in the second color value is 20, and the difference is 235, which is greater than the threshold value of 200, then it is considered to be a broken hole.

[0079] In an optional implementation, step 140 may use a color matching algorithm to perform the following calculation:

[0080] Determine a first grayscale gradient change value of the second characteristic image of the broken hole type along the horizontal direction;

[0081] Determine a second grayscale gradient change value of the first characteristic image of the broken hole type along the horizontal direction;

[0082] Determine the broken holes among all the nozzle holes corresponding to the second characteristic image of the broken hole type based on the first gray gradient change value and the second gray gradient change value;

[0083] When there is a broken hole, the above-mentioned broken hole compensation strategy is adopted.

[0084] In a specific implementation, when the first grayscale gradient change value undergoes a sudden change compared to the second grayscale gradient change value, the position where the sudden change occurs is characterized as the position of the broken hole.

[0085] In an optional implementation, the N types of second characteristic images include a color uniformity type second characteristic image;

[0086] Step 140 may specifically include the following steps:

[0087] Determine a first color grayscale value of a second characteristic image of a color uniformity class;

[0088] Determine a second color grayscale value of the first characteristic image of the color uniformity class;

[0089] Determine a color grayscale change value of a second characteristic image of a color uniformity class based on the first color grayscale value and the second color grayscale value;

[0090] If the color grayscale change value is greater than or equal to the preset color grayscale value, edge detection is performed on the image area corresponding to the color grayscale change value to obtain the image area to be adjusted in the target adjustment information.

[0091] In an optional implementation, the N types of second characteristic images include a driving waveform type second characteristic image;

[0092] Step 140 may specifically use a machine learning algorithm to perform the following steps:

[0093] Based on the first characteristic image of the driving waveform class, identifying satellite points in the second characteristic image of the driving waveform class, and calculating the number of satellite points of all satellite points;

[0094] If the number of satellite points is greater than or equal to the preset number, the driving waveform voltage adjustment value in the target adjustment information is determined.

[0095] Alternatively, step 140 may specifically include the following steps:

[0096] Based on the roundness of ink dots of the first characteristic image of the driving waveform class, a roundness detection is performed on the roundness of ink dots of the second characteristic image of the driving waveform class to determine the driving waveform slope adjustment value in the target adjustment information;

[0097] Alternatively, step 140 may specifically include the following steps:

[0098] Based on the reference straight line, the straightness of each row of ink dots of the second characteristic image of the driving waveform is detected, and the peak spacing adjustment value of the driving waveform in the target adjustment information is determined;

[0099] The image printing processing device provided by the present invention is described below. The image printing processing device described below and the image printing processing method described above can be referred to each other.

[0100] like Figure 3 As shown, an embodiment of the present invention further provides an image printing processing device, which is used to implement the image printing processing method in any of the above embodiments. The image printing processing device may include:

[0101] An acquisition module 310 is used to acquire a CMYK image to be printed;

[0102] The classification module 320 is used to classify, identify and locate the CMYK image to be printed using a target CNN model to obtain N types of first feature images and positions of the N types of first feature images; wherein N is a positive integer greater than 2, and the N types of first feature images at least include a first feature image of a color matching class, a first feature image of a broken hole class, a first feature image of a color uniformity class and a first feature image of a driving waveform class;

[0103] The processing module 330 is used to print the CMYK image to be printed, and perform sampling processing on the image area corresponding to the position of the N types of first characteristic images on the printed image to obtain N types of second characteristic images;

[0104] A detection module 340 is used to detect each type of second feature image in the N types of second feature images to determine target adjustment information;

[0105] The printing module 350 is used to adjust the to-be-printed CMYK image based on the target adjustment information, and to print based on the adjusted CMYK image.

[0106] In an optional implementation, the target CNN model includes at least an input layer, three convolutional layers, an activation function, two pooling layers, a flattening layer, a fully connected layer, and an output layer; the classification module 320 is specifically used for:

[0107] The target CNN model is used to classify and identify the CMYK images to be printed, and N types of first feature images are obtained, including:

[0108] Input the CMYK image to be printed into the input layer;

[0109] The input layer preprocesses the CMYK image to be printed and outputs the preprocessed image to the first convolutional layer of the three convolutional layers;

[0110] The first convolutional layer performs the first feature extraction on the preprocessed image, and uses the activation function to output the image after the first feature extraction to the first pooling layer of the two pooling layers;

[0111] The first pooling layer performs the first downsampling on the image after the first feature extraction and outputs it to the second convolution layer of the three convolution layers;

[0112] Until the third convolutional layer of the three convolutional layers uses the activation function to output the image after the third feature extraction to the flattening layer;

[0113] The flattening layer reduces the dimension of the image after the third feature extraction and outputs the reduced-dimensional image to the fully connected layer;

[0114] The flattening layer reduces the dimension of the image after the third feature extraction and outputs the reduced-dimensional image to the fully connected layer;

[0115] The fully connected layer classifies the reduced-dimensional image and outputs it to the output layer, which uses softmax as the target activation function to output the first feature image of N categories.

[0116] In an optional implementation, the N types of second characteristic images include a color-coded second characteristic image; the color-coded second characteristic image includes M sub-characteristic images; wherein M is a positive integer greater than 2; the detection module 340 is specifically configured to:

[0117] Get the position weight of each nozzle on the nozzle;

[0118] Determine the first position of the color to be detected in each sub-feature image;

[0119] Determine a second position of the reference color of the color to be detected in each sub-feature image;

[0120] determining a position difference between the first position and the second position as an initial position correction value;

[0121] Substitute the initial position correction value and the position weight of the nozzle corresponding to the color to be detected on each sub-feature image into the formula:

[0122]

[0123] The target position correction value of the color to be detected is calculated; wherein Δx1 is the initial position correction value of the color to be detected in the first sub-feature image, and W1 is the position weight of the nozzle corresponding to the color to be detected in the first sub-feature image; Δx2 is the initial position correction value of the color to be detected in the second sub-feature image, and W2 is the position weight of the nozzle corresponding to the color to be detected in the second sub-feature image; Δx M is the initial position correction value of the color to be detected in the Mth sub-feature image, W M is the position weight of the nozzle corresponding to the color to be detected in the Mth sub-feature image; ΔX is the target position correction value.

[0124] In an optional implementation, the N types of second characteristic images include broken hole type second characteristic images; the detection module 340 is specifically used for:

[0125] Determine a first color value corresponding to the position to be detected in the second characteristic image of the broken hole type;

[0126] Determine a second color value corresponding to the position to be detected in the first characteristic image of the broken hole type;

[0127] Based on the difference between the first color value and the second color value, determining whether the nozzle hole corresponding to the position to be detected is a broken hole;

[0128] If the nozzle hole corresponding to the position to be detected is a broken hole, the broken hole compensation strategy in the target adjustment information is adopted; the broken hole compensation strategy is to use the normal nozzle holes adjacent to the broken hole to perform ink volume compensation operation on the broken hole.

[0129] In an optional implementation, the detection module 340 is specifically configured to:

[0130] Determine a first grayscale gradient change value of the second characteristic image of the broken hole type along the horizontal direction;

[0131] Determine a second grayscale gradient change value of the first characteristic image of the broken hole type along the horizontal direction;

[0132] Determine the broken holes among all the nozzle holes based on the first grayscale gradient change value and the second grayscale gradient change value;

[0133] When there is a broken hole in the nozzle, a broken hole compensation strategy is adopted.

[0134] In an optional implementation, the N types of second characteristic images include a color uniformity type second characteristic image; the detection module 340 is specifically used to:

[0135] Determine a first color grayscale value of a second characteristic image of a color uniformity class;

[0136] Determine a second color grayscale value of the first characteristic image of the color uniformity class;

[0137] Determine a color grayscale change value of a second characteristic image of a color uniformity class based on the first color grayscale value and the second color grayscale value;

[0138] If the color grayscale change value is greater than or equal to the preset color grayscale value, edge detection is performed on the image area corresponding to the color grayscale change value to obtain the image area to be adjusted in the target adjustment information.

[0139] In an optional implementation, the N types of second characteristic images include a driving waveform type second characteristic image; the detection module 340 is specifically used for:

[0140] Based on the first characteristic image of the driving waveform class, identifying satellite points in the second characteristic image of the driving waveform class, and calculating the number of satellite points of all satellite points;

[0141] If the number of satellite points is greater than or equal to the preset number, determining the driving waveform voltage adjustment value in the target adjustment information;

[0142] or,

[0143] Detecting each of the N types of second characteristic images to determine target adjustment information, including: performing roundness detection on the roundness of ink dots of the second characteristic image of the driving waveform based on the roundness of ink dots of the first characteristic image of the driving waveform to determine the driving waveform slope adjustment value in the target adjustment information;

[0144] or,

[0145] Each of the N types of second characteristic images is detected to determine the target adjustment information, including: based on the reference straight line, the straightness of each row of ink dots of the driving waveform type second characteristic image is detected, and the peak spacing adjustment value of the driving waveform in the target adjustment information is determined.

[0146] like Figure 4 As shown, an embodiment of the present invention further provides an electronic device, which may include: a processor (processor) 410, a communication interface (CommunicationsInterface) 420, a memory (memory) 430 and a communication bus, wherein the processor 410, the communication interface 420 and the memory 430 communicate with each other through the communication bus. The memory 430 stores a computer program that can be run by the processor 410; when the processor 410 runs the computer program, it can execute the image printing processing method in any of the above embodiments.

[0147] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0148] In yet another aspect, the present invention further provides a non-transitory computer-readable storage medium, in which instructions are stored. When the instructions are executed, the image printing processing method in any of the above embodiments is implemented.

[0149] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art may understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple situations. A single processor or other unit may implement several functions listed in the claims. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0150] Although the present invention has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present invention. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present invention as defined by the appended claims and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present invention. Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, the present invention is intended to include such modifications and variations if they fall within the scope of the claims of the present invention and their equivalents.

Claims

1. An image printing processing method, characterized in that: include: Get the CMYK image to be printed; The target CNN model is used to classify, identify and locate the CMYK image to be printed, and obtain N types of first characteristic images and the positions of the N types of first characteristic images; wherein N is a positive integer greater than 2, and the N types of first characteristic images at least include a first characteristic image of a color matching class, a first characteristic image of a broken hole class, a first characteristic image of a color uniformity class, and a first characteristic image of a driving waveform class; Printing the CMYK image to be printed, and sampling the image areas corresponding to the positions of the N types of first characteristic images on the printed image to obtain N types of second characteristic images; Detecting each type of second feature image in the N types of second feature images to determine target adjustment information; Adjusting the to-be-printed CMYK image based on the target adjustment information, and printing based on the adjusted CMYK image; The N types of second characteristic images include a color-coded second characteristic image; the color-coded second characteristic image includes M sub-characteristic images; wherein M is a positive integer greater than 2; The detecting each type of the N types of second characteristic images to determine the target adjustment information includes: Get the position weight of each nozzle on the nozzle; Determine the first position of the color to be detected in each sub-feature image; Determine a second position of the reference color of the color to be detected in each sub-feature image; determining a position difference between the first position and the second position as an initial position correction value; Substitute the initial position correction value and the position weight of the nozzle corresponding to the color to be detected on each sub-feature image into the formula: The target position correction value of the color to be detected is calculated; wherein Δx1 is the initial position correction value of the color to be detected in the first sub-feature image, and W1 is the position weight of the nozzle corresponding to the color to be detected in the first sub-feature image; Δx2 is the initial position correction value of the color to be detected in the second sub-feature image, and W2 is the position weight of the nozzle corresponding to the color to be detected in the second sub-feature image; Δx M is the initial position correction value of the color to be detected in the Mth sub-feature image, W M is the position weight of the nozzle corresponding to the color to be detected in the Mth sub-feature image; ΔX is the target position correction value.

2. The image printing processing method according to claim 1, characterized in that: The N types of second characteristic images include a broken hole type second characteristic image; The detecting each type of the N types of second characteristic images to determine the target adjustment information includes: Determine a first color value corresponding to the position to be detected in the second characteristic image of the broken hole type; Determine a second color value corresponding to the position to be detected in the first characteristic image of the broken hole type; Based on the difference between the first color value and the second color value, determining whether the nozzle hole corresponding to the position to be detected is a broken hole; If the nozzle hole corresponding to the position to be detected is a broken hole, the broken hole compensation strategy in the target adjustment information is adopted; the broken hole compensation strategy is to use the normal nozzle hole adjacent to the broken hole to perform ink volume compensation operation on the broken hole.

3. The image printing processing method according to claim 2, characterized in that: The detecting each type of the N types of second characteristic images to determine the target adjustment information includes: Determine a first gray gradient change value of the second characteristic image of the broken hole type along the horizontal direction; Determine a second grayscale gradient change value of the first characteristic image of the broken hole type along the horizontal direction; Based on the first grayscale gradient change value and the second grayscale gradient change value, determining the broken holes among all the nozzle holes corresponding to the second characteristic image of the broken hole type; When a broken hole exists, the broken hole compensation strategy is adopted.

4. The image printing processing method according to claim 1, characterized in that: The N types of second characteristic images include a color uniformity type second characteristic image; The detecting each type of the N types of second characteristic images to determine the target adjustment information includes: Determine a first color grayscale value of the second characteristic image of the color uniformity class; Determine a second color grayscale value of the first characteristic image of the color uniformity class; Determining a color grayscale change value of the second characteristic image of the color uniformity class based on the first color grayscale value and the second color grayscale value; If the color grayscale change value is greater than or equal to the preset color grayscale value, edge detection is performed on the image area corresponding to the color grayscale change value to obtain the image area to be adjusted in the target adjustment information.

5. The image printing processing method according to claim 1, characterized in that: The N types of second characteristic images include a driving waveform type second characteristic image; The detecting each type of the N types of second characteristic images to determine the target adjustment information includes: Based on the first characteristic image of the driving waveform class, identifying satellite points in the second characteristic image of the driving waveform class, and calculating the number of satellite points of all satellite points; If the number of satellite points is greater than or equal to a preset number, determining a driving waveform voltage adjustment value in the target adjustment information; or, The detecting each of the N types of second characteristic images to determine the target adjustment information includes: based on the roundness of ink dots of the first characteristic image of the driving waveform, performing roundness detection on the roundness of ink dots of the second characteristic image of the driving waveform to determine the driving waveform slope adjustment value in the target adjustment information; or, The detecting of each type of the N types of second characteristic images to determine the target adjustment information includes: based on a reference straight line, detecting the straightness of each row of ink dots of the driving waveform type second characteristic image, and determining the peak spacing adjustment value of the driving waveform in the target adjustment information.

6. The image printing method according to claim 1, characterized in that: The target CNN model includes at least an input layer, three convolutional layers, an activation function, two pooling layers, a flattening layer, a fully connected layer and an output layer; The method of using the target CNN model to classify, identify and locate the CMYK image to be printed to obtain the N types of first feature images includes: Inputting the to-be-printed CMYK image into the input layer; The input layer preprocesses the CMYK image to be printed, and outputs the preprocessed image to the first convolution layer of the three convolution layers; The first convolution layer performs a first feature extraction on the preprocessed image, and outputs the image after the first feature extraction to the first pooling layer of the two pooling layers using the activation function; The first pooling layer performs a first downsampling on the image after the first feature extraction, and outputs the image to the second convolutional layer of the three convolutional layers; Until the third convolution layer of the three convolution layers outputs the image after the third feature extraction to the flattening layer using the activation function; The flattening layer reduces the dimension of the image after the third feature extraction, and outputs the reduced-dimensional image to the fully connected layer; The fully connected layer classifies the reduced-dimensional image and outputs it to the output layer, and the output layer outputs the N-category first feature image with softmax as the target activation function.

7. An image printing processing device, characterized in that: include: An acquisition module, used to acquire a CMYK image to be printed; A classification module, used for classifying, identifying and locating the CMYK image to be printed by using a target CNN model, and obtaining N types of first characteristic images and positions of the N types of first characteristic images; wherein N is a positive integer greater than 2, and the N types of first characteristic images at least include a first characteristic image of a color matching class, a first characteristic image of a broken hole class, a first characteristic image of a color uniformity class, and a first characteristic image of a driving waveform class; A processing module, used for printing the CMYK image to be printed, and performing sampling processing on the image area corresponding to the position of the N types of first characteristic images on the printed image to obtain N types of second characteristic images; A detection module, used to detect each type of second feature image in the N types of second feature images to determine target adjustment information; A printing module, used for adjusting the to-be-printed CMYK image based on the target adjustment information, and printing based on the adjusted CMYK image; The N types of second characteristic images include a color-coded second characteristic image; the color-coded second characteristic image includes M sub-characteristic images; wherein M is a positive integer greater than 2; and the detection module is specifically used for: Get the position weight of each nozzle on the nozzle; Determine the first position of the color to be detected in each sub-feature image; Determine a second position of the reference color of the color to be detected in each sub-feature image; determining a position difference between the first position and the second position as an initial position correction value; Substitute the initial position correction value and the position weight of the nozzle corresponding to the color to be detected on each sub-feature image into the formula: The target position correction value of the color to be detected is calculated; wherein Δx1 is the initial position correction value of the color to be detected in the first sub-feature image, and W1 is the position weight of the nozzle corresponding to the color to be detected in the first sub-feature image; Δx2 is the initial position correction value of the color to be detected in the second sub-feature image, and W2 is the position weight of the nozzle corresponding to the color to be detected in the second sub-feature image; Δx M is the initial position correction value of the color to be detected in the Mth sub-feature image, W M is the position weight of the nozzle corresponding to the color to be detected in the Mth sub-feature image; ΔX is the target position correction value.

8. An electronic device, characterized in that: include: A processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate via the communication bus; The memory stores a computer program that can be executed by the processor; when the processor executes the computer program, the image printing processing method according to any one of claims 1 to 6 is executed.

9. A computer storage medium, characterized in that The computer storage medium stores instructions, and when the instructions are executed, the image printing processing method according to any one of claims 1 to 6 is implemented.

Images