OnePass inkjet printing defect detection method

Through image processing technology and algorithms, the nozzle defects and nozzle splicing errors of OnePass inkjet printers are automatically detected, which solves the problem of printing quality degradation after long-term work, and realizes high-precision, fast and automated detection, reducing the Inspection costs, optimized maintenance and performance of inkjet printers.

CN118799318BActive Publication Date: 2024-11-19ZHEJIANG UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411276734.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-12
Publication Date
2024-11-19
Estimated Expiration
2044-09-12

AI Technical Summary

Technical Problem

OnePass inkjet printers are prone to nozzle blockage, oblique spray, satellite ink drops and ink leakage after long-term work, resulting in a decrease in printing quality. The existing detection methods are inefficient and costly, especially in the face of large-scale nozzle splicing misalignment. It is difficult to achieve accurate and rapid detection.

Method used

Using image processing technology and various algorithms, by obtaining electronic scanning diagrams, identifying the ink dot position of the nozzle, building a point set corresponding model, extracting abnormal points, automatically detecting nozzle defects and quantifying splicing errors between nozzles, realizing the automation of inkjet printers Inspection and maintenance.

Benefits of technology

It realizes high-precision, fast and automated inspection of OnePass inkjet printers, reduces inspection costs, meets the real-time requirements of high-speed printing processes, can provide high confidence detection results in extreme cases, and optimizes the maintenance of inkjet printers. and performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118799318B_ABST
    Figure CN118799318B_ABST
Patent Text Reader

Abstract

A OnePass inkjet printing defect detection method belongs to the technical field of inkjet printing defect detection. The present invention models the printing point set through the integration of image processing technology and multiple algorithms to identify all nozzle defects existing in the current printing process and calculate the splicing errors between each nozzle; compared with the prior art, the method of the present invention does not rely on the expensive piezoelectric sensing system. The proposed method is not only simple and easy to implement, but also can meet the strict requirements of detection accuracy and real-time performance in the high-speed printing process; in the face of extreme situations in actual applications, such as large-scale failures of print heads or large-scale abnormal factors, it can still provide high-confidence defect detection results; this has important practical value for the maintenance and optimization of inkjet printers.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of inkjet printing defect detection, and in particular relates to a OnePass inkjet printing defect detection method. Background Art

[0002] Inkjet printing refers to the process of ejecting ink droplets onto a print medium through the nozzles on the print head to obtain images or text. OnePass printing is different from traditional scanning printing and is also called Single Pass printing or single-pass printing. This technology uses high-precision and high-stability print heads, integrates advanced inkjet control technology, and adopts a fixed-position multi-array print head structure. This enables it to complete the ejection of multiple colors of ink in a single inkjet process, achieving high-efficiency and high-resolution printing.

[0003] After the OnePass inkjet printer has been working for a long time, the air inside the nozzle has not been completely exhausted, so that the solidified ink blocks the inkjet hole, affecting the straightness of the inkjet and causing oblique spraying. In addition, the shape and volume of the ink droplets may vary, which may eventually cause the nozzle to be blocked. The mismatch between the driving waveform and the inkjet pulse may cause the tail of the ink droplet to break due to the necking effect, forming satellite ink droplets. Fluctuations in air pressure may also induce ink leakage. In addition, due to its high-precision positioning characteristics, the OnePass printer often causes multiple nozzles to be misaligned when spliced. This is caused by a variety of factors such as loose fasteners caused by thermal expansion and contraction or jitter and deviation of the mechanical structure during printing. The above inkjet defects may have an adverse effect on printing quality and color reproduction.

[0004] Traditional inkjet defect detection methods usually rely on manual monitoring, which is not only inefficient but also accompanied by high labor costs. The mainstream methods currently used are through high-speed cameras or piezoelectric self-sensing methods. Although these methods have made significant contributions to improving the research and application of inkjet printing technology, they still have certain limitations in equipment experimental requirements and scope of application. Especially when faced with large nozzle splicing misalignment, existing technologies may not be able to achieve accurate and rapid detection of errors. Summary of the invention

[0005] In view of the above problems existing in the prior art, the purpose of the present invention is to provide a OnePass inkjet printing defect detection method, which can automatically analyze and detect all nozzle failure conditions and quantify the splicing errors between nozzles, thereby providing a new solution for optimizing printer maintenance.

[0006] The present invention provides the following technical solution: a OnePass inkjet printing defect detection method, comprising the following steps:

[0007] S1: Obtaining an electronic scan image for detection, wherein the electronic scan image is obtained by scanning a test image printed on a OnePass inkjet printer, wherein the test image is a test page in which preset pixel positions correspond to the nozzle numbers of the printer; dividing the electronic scan image into regions to obtain a scan sub-image of each nozzle region;

[0008] S2: performing image processing on the scanned sub-image to obtain the position information of all ink points in the image, and marking the ink points with satellite ink droplets and ink leakage defect features, including impurity points and missing points;

[0009] S3: constructing an optimal point set correspondence model based on the collected ink point position information and the existing actual nozzle position information in the scanning sub-image, and outputting the point set correspondence relationship and the optimal transformation parameters;

[0010] S4: classifying the abnormal points existing in the point set correspondence model, extracting the ink points with blockage and oblique spraying defects, and determining the numbers of all defective nozzles in the current printhead in combination with the ink point correspondence conditions marked in S2 and outputting them; the abnormal points are ink points that have no corresponding relationship in the model results;

[0011] S5: After matching and normalizing the transformation parameters of all adjacent nozzles one by one, the stitching error between the nozzles is quantified and obtained to evaluate the current printer performance.

[0012] Furthermore, the specific process of S2 is as follows:

[0013] S2.1: Perform special grayscale processing on the scanned sub-image to obtain a grayscale image;

[0014] S2.2: Binarize the grayscale image and perform contour recognition, then screen all contours and mark the contours with relevant defect features;

[0015] S2.3: Calculate the centroid position of each contour and extract the position information of all ink points.

[0016] Furthermore, the specific process of S3 is as follows:

[0017] S3.1: Assume that the actual nozzle positions in the scanned sub-image constitute a set A, and all the ink point positions collected in the actual printed image constitute a set B. Construct a distribution matrix K to represent the correspondence of the point sets; construct augmented nodes to handle abnormal points;

[0018] S3.2: Construct an objective function to optimize the matching of the point set and its transformation parameters.

[0019] Furthermore, the specific process of S3.2 is as follows:

[0020] S3.2.1: Normalize the K rows and columns of the iterated allocation matrix by the Sinkhorn algorithm;

[0021] S3.2.2: Update the transformation relationship S, R, t using the least squares method;

[0022] S3.2.3: Repeat steps S3.2.1 and S3.2.2 until the objective function converges.

[0023] Furthermore, the specific process of S4 is as follows:

[0024] S4.1: Extract all outliers in point set A and compare their positions with those in point set B;

[0025] S4.2: Output the nozzle number with specific defects according to different abnormal corresponding situations and special marking points.

[0026] By adopting the above technology, compared with the prior art, the beneficial effects of the present invention are as follows:

[0027] The present invention models the printing point set through the integration of image processing technology and multiple algorithms to identify all nozzle defects existing in the current printing process and calculate the splicing errors between each nozzle. Compared with the prior art, the method of the present invention does not rely on the expensive piezoelectric sensing system. The proposed method is not only simple and easy to implement, but also can meet the strict requirements of detection accuracy and real-time performance in the high-speed printing process. In the face of extreme situations in actual applications, such as large-scale failures of print heads or large-scale abnormal factors, it can still provide high-confidence defect detection results. This has important practical value for the maintenance and optimization of inkjet printers. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a flow chart of the method of the present invention;

[0029] Figure 2 A schematic diagram of the structure of the OnePass inkjet printer of the present invention;

[0030] Figure 3 3a is a schematic diagram of a test page of the present invention, wherein 3a is a schematic diagram of a monochrome overall printing test page, 3b is a schematic diagram of a single nozzle sub-image, and 3c is a schematic diagram of the positions of some nozzles;

[0031] Figure 4 It is a schematic diagram of the structure of satellite ink droplets and nozzle ink leakage in the image;

[0032] Figure 5 This is a schematic diagram of the characteristic detection results of nozzle blockage defects and oblique spray defects;

[0033] Figure 6This is a structural diagram of the nozzle splicing error. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments of the specification. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0035] On the contrary, the present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention as defined by the claims. Further, in order to make the public have a better understanding of the present invention, some specific details are described in detail in the detailed description of the present invention below. Those skilled in the art can fully understand the present invention without the description of these details.

[0036] See also Figure 1-6 , a OnePass inkjet printing defect detection method, wherein the structure of the OnePass inkjet printer is as follows Figure 2 As shown in the figure. It adopts a multi-array nozzle structure, and the CMYK four-color nozzles are configured in equal amounts to form a nozzle group. During the printing process, the nozzle group is fixed, and the paper moves at high speed along the roller driven by the motor.

[0037] The specific steps of the defect detection method are as follows:

[0038] S1: Obtain an electronic scan image. The scan image is obtained by scanning a test image printed on a OnePass inkjet printer. The test image is a test page in which the preset pixel positions correspond to the nozzle numbers of the printer. Divide the scan image into regions and obtain a scan sub-image of each nozzle region. The specific process is as follows:

[0039] S1.1: The test page generated by the code is as follows Figure 3 shown. Figure 3 a shows a monochrome overall print test page. In order to ensure clear positioning and capture of the nozzles in each printhead, the odd-numbered and even-numbered printheads are staggered and divided by black solid lines. Figure 3 b The black frame is a single nozzle sub-image. Figure 3 c is a partial nozzle position map, where one nozzle corresponds to one pixel.

[0040] S1.2: Print using OnePass inkjet printer, the nozzle model is Samba G3L, each color consists of 8 nozzles, and the printing accuracy is 1200dpi. In order to further improve the calculation accuracy, each sub-image area is scanned at 9600dpi.

[0041] S2: Perform image processing on the scanned sub-image to obtain the position information of all ink points in the image, and mark the ink points with satellite ink droplets and ink leakage defect features, including a large number of impurity points and missing points.

[0042] S2.1: Perform special grayscale processing on the scanned sub-image to reduce non-target impurities in the image and make the image easier to identify. According to the inherent characteristics between the CMYK color mode and the RGB color space, let:

[0043]

[0044] Among them I C ,I M ,I Y ,I K They represent the grayscale results of the four color data of CMYK. R ,I G ,I B Respectively represent the pixel values ​​of the R channel, G channel and B channel of the scanned image.

[0045] S2.2: Binarize the grayscale image and perform contour recognition. It is known that ink droplets generally appear circular in an ideal state. Figure 4 As shown in the figure, satellite droplets appear as scattered points or burr-like structures attached to the edge of the main droplet in the image (rectangular frame selection). Nozzle leakage appears as abnormally low grayscale values ​​in a specific area in the image, which is caused by repeated deposition of ink at the same location (circular frame selection). Traverse all contours and mark the contours with relevant features in the figure.

[0046] S2.3: For each contour , calculate its centroid position by the following method ( ):

[0047]

[0048]

[0049] in For outline In the region Gray value at the point, are the horizontal and vertical pixel coordinates in the scanned sub-image. All centroid information and special annotations are recorded and normalized to form a set B of all ink point positions collected in the actual printed image. At this time, there are a large number of impurity points and missing points in point set B.

[0050] S3: Based on the collected ink point location information and the existing actual nozzle location information in the scanned sub-image, the optimal point set corresponding model is constructed, and the relevant parameters of the current corresponding nozzle transformation are solved by the least squares method. The specific process is as follows:

[0051] S3.1: Assume that the actual nozzle positions in the scanned subgraph form a set A, represented by A= , all the ink point positions collected in the actual printed image constitute a set B, expressed as B = , M represents the number of nozzles in the sub-image, N represents the number of ink points collected, i and j represent the point numbers in the A and B point sets. Define a binary assignment matrix K with a dimension of (M+1)*(N+1), where the M*N sub-matrices inside K define the corresponding relationship between A and B. If Corresponds to ,but =1, otherwise = 0. The extra rows and columns are used to handle outliers. In order to ensure the one-to-one correspondence between points, the matrix K must be a doubly random matrix. An example of the matrix is ​​shown in Table 1. In this case, points x1, x2, and x3 correspond to points y1, y4, y2, and y3 is an outlier.

[0052] Table 1 Matrix example

[0053]

[0054] S3.2: Construct an objective function to continuously optimize the matching of the point set and its transformation parameters.

[0055] Among them, the objective function in step S3.2 is:

[0056]

[0057] R∈ SO(2) is the rotation matrix of the current nozzle, is the displacement deflection,

[0058] S represents the expansion and contraction change of the image size caused by the error between the paper and the motor.

[0059] Let the X direction be the paper feeding direction and the Y direction be the nozzle arrangement direction:

[0060]

[0061] Represents a regularized variable to prevent excessive scaling changes from affecting other parameters. , is the smoothing weight parameter. is the random error threshold, which represents how much noise the model can tolerate. The last entropy term makes the allocation matrix gradually and continuously improve during the optimization process. is the control parameter, as With the increasing iteration of , the model tends to regard unmatched points as outliers, reducing the overall matching cost and thus obtaining the global optimal solution.

[0062] The specific process of step S3.2 is as follows:

[0063] S3.2.1: Normalize the rows and columns of the soft assignment matrix K of each iteration by the Sinkhorn algorithm:

[0064] , i = 1, 2, ..., M + 1 and j = 1, 2, ..., N + 1

[0065] , i=1,2,…,M and j=1,2,…,N+1

[0066] , i=1,2,…,M+1 and j=1,2,…,N;

[0067] Indicates row or column;

[0068] S3.2.2: Use least squares to update the transformation relationship S, R, t. The goal is to:

[0069] ;

[0070] is the objective function after K matrix iteration;

[0071] S3.2.3: Repeat steps S3.2.1 and S3.2.2 until the objective function converges, and output the point set correspondence and the optimal transformation parameters S, R, t.

[0072] S4: Classify the abnormal points in the point set correspondence model, extract the ink points with blockage and oblique spraying defects, and determine the nozzle numbers of all defective nozzles in the current print head in combination with the correspondence of the specially marked ink points in S2 and output them.

[0073] S4.1: Extract all outliers in point set A and compare their positions with those in B. ) At this time or Points are considered as outliers. In the final matching results, outliers that are not matched with any point are represented as or =1, o,u represent the outlier numbers of point sets A and B respectively. Extract all outliers in A , calculate the optimal transformation Ideal location for , and point Centered on is the radius ( is the minimum distance between adjacent points in point set A) to search for abnormal points within the circle .

[0074] S4.2: If not found in the range , then it is believed that The corresponding nozzle has a clogging defect. There are multiple , then comprehensively consider and The Euclidean distance and relative position in space are more inclined to select the ones that are close in the nozzle direction (Y axis). The point is taken as the oblique spray landing point and recorded The corresponding nozzle has an oblique spray defect. Figure 4 The results of the two defect feature detections are shown. The nozzle numbers with specific defects are output according to different abnormal correspondence situations and special marking points. Automatic analysis and detection of all nozzle failures of OnePass inkjet printers are achieved, which increases convenience and has higher accuracy.

[0075] S5: Nozzle splicing error detection, the error is roughly expressed as follows Figure 5 shown.

[0076] In this embodiment, a nozzle-by-nozzle matching method is adopted, in which the point set A is obtained based on the actual nozzle position information. Through normalization processing, the upper right corner nozzle of each nozzle is uniformly positioned as the starting point, so as to ensure that each set of parameters obtained in the end is based on this point as a reference. In order to highlight the deviation at the splicing point, these parameters are further used to quantify the angle and displacement errors between nozzles.

[0077] Assume that the first nozzle = (a, b), , the second nozzle = (c, d), , the ideal splicing center between the nozzles is ( , ). Then the relative rotation between them is , the relative displacement is:

[0078] ;

[0079] and Together they define the stitching error between print heads.

[0080] By using a similar approach, we can systematically detect and quantify the stitching errors of all print heads on this printer.

[0081] In summary, this embodiment provides a method for detecting inkjet defects of OnePass inkjet printers, including automatic analysis and detection of all nozzle failure conditions and quantification of splicing errors between nozzles. This real-time performance evaluation mechanism provides instant status feedback for the inkjet printing process. With its high precision, economy and computational efficiency, this method can still provide high-confidence defect detection results in the face of extreme situations in practical applications, such as large-scale failures of nozzles or interference from large-scale abnormal factors. This has important practical value for the maintenance and optimization of inkjet printers.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A OnePass inkjet printing defect detection method, characterized in that: The following steps are involved: S1: Obtaining an electronic scan image for detection, wherein the electronic scan image is obtained by scanning a test image printed on a OnePass inkjet printer, wherein the test image is a test page in which preset pixel positions correspond to the nozzle numbers of the printer; dividing the electronic scan image into regions to obtain a scan sub-image of each nozzle region; S2: performing image processing on the scanned sub-image to obtain the position information of all ink points in the image, and marking the ink points with satellite ink droplets and ink leakage defect features, including impurity points and missing points; S3: Based on the collected ink point position information and the existing actual nozzle position information in the scanning sub-image, an optimal point set correspondence model is constructed, and the point set correspondence relationship and the optimal transformation parameters are output; the specific process of S3 is as follows: S3.1: Assume that the actual nozzle positions in the scanned sub-image constitute a set A, and all the ink point positions collected in the actual printed image constitute a set B. Construct a distribution matrix K to represent the correspondence of the point sets; construct augmented nodes to handle abnormal points; S3.2: Construct an objective function to optimize the matching of the point set and its transformation parameters; The objective function is: ; R∈ SO(2) is the rotation matrix of the current nozzle, is the displacement deflection, S represents the expansion and contraction change of the image size caused by the error between the paper and the motor. Let the X direction be the paper feeding direction and the Y direction be the nozzle arrangement direction: ; Represents a regularized variable to prevent excessive scaling changes from affecting other parameters. , is the smoothing weight parameter, is the random error threshold, which represents how much noise the model can tolerate. The last entropy term allows the allocation matrix to be continuously improved during the optimization process. is the control parameter, as The continuous increase of iterations; The specific process of S3.2 is as follows: S3.2.1: Normalize the K rows and columns of the iterated allocation matrix by the Sinkhorn algorithm; S3.2.2: Use least squares to update the transformation relationship S, R, t. The goal is to: ; is the objective function after K matrix iteration; S3.2.3: Repeat steps S3.2.1 and S3.2.2 until the objective function converges; S4: classifying the abnormal points existing in the point set correspondence model, extracting the ink points with blockage and oblique spraying defects, and determining the numbers of all defective nozzles in the current printhead in combination with the ink point correspondence conditions marked in S2 and outputting them; the abnormal points are ink points that have no corresponding relationship in the model results; S5: After matching and normalizing the transformation parameters of all adjacent nozzles one by one, the stitching error between the nozzles is quantified and obtained to evaluate the current printer performance.

2. A OnePass inkjet printing defect detection method according to claim 1, characterized in that: The specific process of S2 is as follows: S2.1: grayscale the scanned sub-image to obtain a grayscale image; S2.2: Binarize the grayscale image and perform contour recognition, then screen all contours and mark the contours with relevant defect features; S2.3: Calculate the centroid position of each contour and extract the position information of all ink points.

3. A OnePass inkjet printing defect detection method according to claim 2, characterized in that: The specific process of S4 is as follows: S4.1: Extract all outliers in point set A and compare their positions with those in point set B; S4.2: Output the nozzle number with specific defects according to different abnormal corresponding situations and special marking points.