Optimization Method and Optimization System for Rapidly Inspecting Inkjet State of Nozzles through a Camera

Abstract

The present invention discloses an optimization method and an optimization system for quickly inspecting the inkjet state of nozzles through a camera. The optimization method for quickly inspecting the inkjet state of nozzles through a camera includes: performing ink droplet observation and positioning on the inkjet print head, selecting a suitable inkjet print head as a qualified print head, and recording the ink droplet parameters of the qualified print head currently sprayed on the test film; obtaining the ink droplet image on the test film, filtering out the interfering ink dots in the ink droplet image through the ink droplet distribution state and morphology screening method, obtaining the filtered screened ink droplet image, and identifying the hole positions of the abnormal nozzles corresponding to the abnormal ink droplets through the ink droplet distribution state and morphology screening method; performing ink droplet observation and measurement on the screened abnormal nozzles. The present invention discloses an optimization method and an optimization system for quickly inspecting the inkjet state of nozzles through a camera, which minimizes interference, increases the robustness of the algorithm, and saves the time and efficiency of nozzle state inspection.

Description

Technical Field

[0001] The present invention relates to the technical field of inkjet printing, and particularly to an optimization method and an optimization system for quickly inspecting the inkjet state of nozzles through a camera. Background Art

[0002] In the industrial printing industry, since there are a large number of nozzles in the print head, in order to ensure the consistency of the printing state and prevent abnormal printing quality caused by nozzle blockage, it is necessary to visually manage the inkjet state of the nozzles. The existing general method is to accurately collect the speed / volume and angle of the ink droplets of the nozzles through an ink droplet observation instrument. In order to obtain more accurate measurement results, a high-power lens needs to be used, which makes the field of view of the ink droplet observation instrument small. At most 5-8 nozzles can be inspected each time. There are thousands of nozzles in a print head. If inspected by moving, it takes about 30 minutes to 1 hour for a single print head, which greatly affects the operation efficiency. If multiple print heads are used in combination in a device for printing, the detection time will increase exponentially.

[0003] Therefore, it is necessary to develop a system for quickly inspecting the inkjet state. First, the ink droplets are printed on a transparent film, then the transparent film is sampled for visual inspection and the nozzle numbers are located, and then the abnormal nozzle numbers are observed for ink droplets. Since the sampling situation is relatively complex, it is necessary to strengthen the robustness of the algorithm. Summary of the Invention

[0004] The present invention overcomes the deficiencies of the prior art and provides an optimization method and an optimization system for quickly inspecting the inkjet state of nozzles through a camera. The ejected ink droplets are used to locate whether the nozzles are abnormal, and the abnormal nozzle positions are recorded; through image preprocessing, the interference is minimized, and further reduced through automatic region segmentation. The phenomenon of uneven ink droplet spacing caused by uneven tension of the transparent film itself is optimized, the robustness of the algorithm is increased, and the time and efficiency for inspecting the state of the print head are saved.

[0005] To achieve the above object, the technical solution adopted by the present invention is: an optimization method for quickly inspecting the inkjet state of nozzles through a camera, including the following steps:

[0006] Step 1, perform ink droplet observation and positioning on the print head of the inkjet printing, select a suitable print head of the inkjet printing as a qualified print head, and record the ink droplet parameters of the qualified print head currently sprayed on the test film.

[0007] Step 2, obtain the ink droplet image on the test film in Step 1, filter out the interfering ink dots in the ink droplet image through the ink droplet distribution state and morphology screening method, obtain the filtered screened ink droplet image, and locate the positions of the abnormal nozzles corresponding to the abnormal ink droplets through the ink droplet distribution state and morphology screening method.

[0008] Step 3, perform ink droplet observation and measurement on the selected abnormal nozzles.

[0009] In a preferred embodiment of the present invention, the ink droplet observation and positioning includes the following steps:

[0010] Step 1.1, determine whether the ink droplets ejected from the nozzles of the print head meet the ink droplet requirements set in inkjet printing;

[0011] Step 1.1.1, when the ink droplets ejected from the nozzles of the print head meet the ink droplet requirements set in inkjet printing, jump to Step 2;

[0012] Step 1.1.2, when the ink droplets ejected from the nozzles of the print head do not meet the ink droplet requirements set in inkjet printing, collect the next nozzle and repeat Step 1.1;

[0013] Until a print head that can eject ink droplets meeting the ink droplet requirements set in inkjet printing is selected.

[0014] In a preferred embodiment of the present invention, in Step 1.1, an ink droplet observation instrument is used to sequentially check the nozzles in each row of the print head, and it is determined whether the ink droplets ejected from the nozzles meet the preset ink droplet volume requirements according to the volume of the ejected ink droplets. Those that meet the requirements are determined to be qualified. When the current ink droplet volume is qualified, the operation of Step 1.1 is stopped, and the offset of the first ink droplet in the current row is recorded, and then marked in the ink droplet image of the current row according to the recorded position.

[0015] In a preferred embodiment of the present invention, in Step 2, obtaining the ink droplet image on the test film in Step 1 includes the following steps:

[0016] Collect the ink droplet image printed on the test film by the nozzles selected in the current Step 1 through a grayscale line scan camera.

[0017] In a preferred embodiment of the present invention, the ink droplet distribution state and morphology screening method includes the following steps:

[0018] Step 2.1, perform binarization operation on the abnormal interference on the ink droplet image through a set threshold, select the corresponding kernel, and perform preliminary removal of interfering ink dots through morphological filtering opening operation of the image to obtain a preliminary filtered ink droplet image after removing interfering ink dots;

[0019] Step 2.2, take the center points of all the ink droplets in the preliminary filtered ink droplet image, sort the Y values of each ink droplet image, and then automatically segment them into several groups of values according to the sorting results by using a segmentation algorithm. At this time, each group of ink droplets still contains interfering points;

[0020] Step 2.3, calculate the distance between all the points in the group and the center value of their respective classification clusters. If the distance is greater than the set threshold, it is considered interference, and then this point is deleted in the group. At this time, the ink droplet spacing basically removes the interference of nozzle missing and scattered ink droplets, and a deeply filtered ink droplet image is obtained;

[0021] Step 2.4, search for single missing points in the deep-filtered ink droplet image through the single missing point search algorithm, and search for single missing points in the deep-filtered ink droplet image through the continuous missing ink point search algorithm; until the corresponding point coordinates of all the nozzle holes in this row are found, and mark the serial numbers.

[0022] In a preferred embodiment of the present invention, in step 2.2, the segmentation algorithm includes the following steps:

[0023] Step 2.2.1, mark the pixel value of the sample point in the Y direction in the image as Y i , where i represents the number of samples, and for Y i with the same value, merge and record as Y n , where n represents the number of merged Y samples, and record the quantity;

[0024] Step 2.2.2, dynamically segment the clusters for the merged Y n , when Y n - Y n-1 = 1, it is considered that two coordinates are adjacent and classified into the same cluster. When Yn - Yn-1 > 1, it is considered that two coordinates are not adjacent and classified into different clusters. For each cluster, record as C m , M represents the number of clusters, and calculate the mean value A m of all points in the cluster and the quantity N m ;

[0025] Step 2.2.3, screen out the 3 clusters with the largest number of points in the cluster;

[0026] Step 2.2.4, take the mean value of the adjacent cluster centers to calculate the segmentation line value S m = (A m + A m+1 ) / 2;

[0027] Step 2.2.5, segment the original samples according to the segmentation line value S m .

[0028] In a preferred embodiment of the present invention, the single missing point search algorithm includes the following steps:

[0029] Calculate the distance D i between two adjacent points in the X direction for each group of points, and sort the distance D i to generate D i ’, take the median value D i ’(I / 2) of D i ’, and obtain the median value of the distance as the actual ink droplet distance D = D i ’(I / 2), where I is the number of all samples;

[0030] Draw a circle with a radius of a set value with the center at a standard distance D increased to the right from the current point. Search for ink dots among all points in each group from above within the circular area. If there is an ink dot, it is determined as a normal ink dot, and then search for ink dots to the right successively based on the currently found ink dots.

[0031] In a preferred embodiment of the present invention, the algorithm for finding continuously missing ink dots includes the following steps: When it is detected that several adjacent ink drops are all missing points, the hole positions are corrected again by referring to the results of the adjacent front row and back row; and so on, find the corresponding point coordinates of all nozzle holes in this row in the image and mark the serial numbers.

[0032] Among them, correcting the hole positions is based on the characteristics of the nozzle holes of the print head to know that the distances after projection of all holes in the ideal state are equal.

[0033] In a preferred embodiment of the present invention, abnormal interferences include ink drop scatter points, ink drop scratches, or ink drop adhesions; the nucleus is smaller than a normal ink drop and larger than the median value of ink drop scatter points; normal ink drops cannot be removed when initially removing interfering ink dots.

[0034] In a preferred embodiment of the present invention, an optimized system for quickly checking the ink jet state of nozzle holes by a camera is implemented by using an optimized method for quickly checking the ink jet state of nozzle holes by a camera, and includes:

[0035] A data acquisition module, which is used to observe and locate ink drops of the print head for ink jet printing, select a suitable print head for ink jet printing as a qualified print head, and record the ink drop parameters of the qualified print head currently sprayed on the test film.

[0036] A data processing module, which is used to obtain the ink drop image on the test film, filter out the interfering ink dots in the ink drop image through the ink drop distribution state and morphology screening method, obtain the filtered screened ink drop image, and find out the hole positions of the abnormal nozzles corresponding to the abnormal ink drops through the ink drop distribution state and morphology screening method.

[0037] A data compounding module, which is used to perform ink drop observation and measurement on the screened abnormal nozzle holes.

[0038] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0039] An optimized method and an optimized system for quickly checking the ink jet state of nozzle holes by a camera disclosed by the present invention.

[0040] 1. The present invention uses a line scan camera image acquisition system, uses a transparent film as an intermediate carrier, locates whether the nozzle holes are abnormal with the ejected ink drops, and records the abnormal hole positions.

[0041] 2. In this system, through image preprocessing, the interference of scatter points / scratches / adhesions is minimized, and further reduced through automatic region segmentation.

[0042] 3. In this system, the algorithm enhances the interference of scatter points / scratches / adhesions in the image. At the same time, through the dynamic spacing segmentation method, the phenomenon of uneven ink droplet spacing caused by uneven tension of the transparent film itself is optimized, increasing the robustness of the algorithm.

[0043] 4. After positioning, a targeted ink droplet observer is used for precise remeasurement, greatly saving the time and improving the efficiency of nozzle state inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The present invention will be further described below in conjunction with the drawings and embodiments.

[0045] Figure 1 is a schematic flow chart of an optimization method and an optimization system for quickly inspecting the ink jet state of nozzle holes through a camera according to the present invention;

[0046] Figure 2 is a schematic diagram of ink droplet observation and positioning in an optimization method and an optimization system for quickly inspecting the ink jet state of nozzle holes through a camera according to the present invention;

[0047] Figure 3 is a schematic diagram of the nozzle spraying ink droplets onto a test film in an optimization method and an optimization system for quickly inspecting the ink jet state of nozzle holes through a camera according to the present invention;

[0048] Figure 4 is a schematic diagram of a gray-scale line-scan camera collecting an ink droplet image of ink droplets on a test film in an optimization method and an optimization system for quickly inspecting the ink jet state of nozzle holes through a camera according to the present invention;

[0049] Figure 5 is a schematic diagram of the hole positions of the abnormal nozzles corresponding to the abnormal ink droplets screened out by the ink droplet distribution state and morphology screening method in an optimization method and an optimization system for quickly inspecting the ink jet state of nozzle holes through a camera according to the present invention;

[0050] Figure 6 is a schematic diagram of sequentially inspecting nozzle holes in an optimization method and an optimization system for quickly inspecting the ink jet state of nozzle holes through a camera according to the present invention;

[0051] Figure 7 is an image of ink droplet sampling on a test film using a single gray-scale line-scan camera and a backlight method in an optimization method and an optimization system for quickly inspecting the ink jet state of nozzle holes through a camera according to the present invention;

[0052] Figure 8It is the image after binarization processing of the sampled image in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0053] Figure 9 It is the image after iconography processing of the image after binarization processing in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0054] Figure 10 It is Chart 1 for implementing a segmentation algorithm in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0055] Figure 11 It is Chart 2 for implementing a segmentation algorithm in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0056] Figure 12 It is the original image corresponding to the actual segmentation of the original sample according to the segmentation line value Sm in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0057] Figure 13 It is the data table after deleting the interference points whose distances from all points in the group and the central values of their respective classification clusters are greater than the set threshold in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0058] Figure 14 It is the image after the interference of nozzle hole missing and scattered ink droplets is preliminarily removed from the ink droplet spacing in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0059] Figure 15 It is the schematic diagram for implementing the algorithm for finding a single missing point in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0060] Figure 16 It is the schematic diagram for implementing the algorithm for finding continuous missing ink droplets in an optimization method and an optimization system for quickly inspecting the inkjet state of nozzle holes through a camera according to the present invention;

[0061] Wherein, 1 - nozzle, 2 - ink droplet observer, 3 - backlight, 4 - grayscale line scan camera, 5 - test film, 6 - coaxial light source. Detailed implementation manner

[0062] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. Without conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0063] The term "and / or" is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " generally indicates that the associated objects before and after are in an "or" relationship. Embodiment 1

[0064] As Figures 1 - 16 shown, the present invention discloses an optimized method for quickly inspecting the inkjet state of nozzle holes through a camera, including the following steps:

[0065] Step 1: Perform ink droplet observation and positioning on the inkjet print head 1, select a suitable inkjet print head 1 as the qualified print head 1, and record the ink droplet parameters of the current inkjet on the test film 5 by the qualified print head 1.

[0066] Specifically, the ink droplet observation and positioning includes the following steps:

[0067] Step 1.1: Determine whether the ink droplets ejected from the nozzle holes of the print head 1 meet the ink droplet requirements set in inkjet printing.

[0068] Step 1.1.1: When the ink droplets ejected from the nozzle holes of the print head 1 meet the ink droplet requirements set in inkjet printing, jump to Step 2.

[0069] Step 1.1.2: When the ink droplets ejected from the nozzle holes of the print head 1 do not meet the ink droplet requirements set in inkjet printing, collect the next nozzle hole and repeat Step 1.1.

[0070] Until a print head 1 that can eject ink droplets meeting the ink droplet requirements set in inkjet printing is selected.

[0071] More specifically, in Step 1.1, an ink droplet observation instrument 2 is used to sequentially check the nozzle holes in each row of the print head 1, determine whether the volume of the ink droplets ejected from the nozzle holes meets the preset ink droplet volume requirements, and judge those that meet as qualified. When the current ink droplet volume is qualified, stop the operation of Step 1.1, record the offset of the first ink droplet in the current row, and then mark it in the ink droplet diagram of the current row according to the recorded position.

[0072] Step 2: Obtain the ink droplet image on the test film 5 in Step 1, filter out the interfering ink dots in the ink droplet image through the ink droplet distribution state and morphology screening method, obtain the filtered screened ink droplet image, and identify the hole positions of the abnormal nozzles corresponding to the abnormal ink droplets through the ink droplet distribution state and morphology screening method. Abnormal interference includes ink droplet scatter points, ink droplet scratches, or ink droplet adhesion; the nucleus is smaller than a normal ink droplet and larger than the median value of the ink droplet scatter points; normal ink droplets cannot be removed during the preliminary removal of interfering ink dots.

[0073] Specifically, in Step 2, obtaining the ink droplet image on the test film 5 in Step 1 includes the following steps:

[0074] Collect the ink droplet image of the ink droplets printed by the nozzles selected in the current Step 1 on the test film 5 through the grayscale line scan camera 4.

[0075] Furthermore, the ink droplet distribution state and morphology screening method includes the following steps:

[0076] Step 2.1: For the abnormal interference on the ink droplet image, perform a binarization operation through a set threshold, select the corresponding nucleus, and perform a preliminary removal of interfering ink dots through the opening operation of image morphological filtering to obtain the initially filtered ink droplet image after the preliminary removal of interfering ink dots;

[0077] Step 2.2: Take the center points of all the ink droplets in the initially filtered ink droplet image, sort the Y values of each ink droplet image, and then automatically segment them into several groups of values according to the sorting results. At this time, each group of ink droplets still contains interfering points.

[0078] Furthermore, in Step 2.2, the segmentation algorithm includes the following steps:

[0079] Step 2.2.1: Mark the pixel value of the sample point in the Y direction in the image as Y i , where i represents all the sample quantities, and for Y i with the same values, merge and record them as Y n , where n represents the number of merged Y samples and record the quantity;

[0080] Step 2.2.2: Dynamically segment the clusters (heaps) for the merged Y n . When Y n - Y n-1 = 1, it is considered that the two coordinates are adjacent and classified into the same cluster. When Y n - Y n-1 > 1, it is considered that the two coordinates are not adjacent and classified into different clusters. Denote each cluster as C m , M represents the number of clusters, calculate the mean value A m of all the points in the cluster and the quantity N m ;

[0081] Step 2.2.3: Select the 3 clusters with the largest number of in-cluster points;

[0082] Step 2.2.4: Calculate the dividing line value S between clusters by taking the mean of adjacent cluster centers m =(A m +A m+1 ) / 2;

[0083] Step 2.2.5: Divide the original samples according to the dividing line value S m and actually correspond to the original image record.

[0084] Step 2.3: Calculate the distance between all points in this group and the center value of their respective classification clusters. If the distance is greater than the set threshold, it is considered interference, and then this point is deleted from this group. At this time, the ink droplet spacing basically removes the interference of nozzle missing and scattered ink droplets, and a deep-filtered ink droplet image is obtained;

[0085] Step 2.4: Search for single missing points in the deep-filtered ink droplet image through the single missing point search algorithm, and search for single missing points in the deep-filtered ink droplet image through the continuous missing ink point search algorithm; until the image corresponding point coordinates of all nozzle numbers in this row are found and numbered.

[0086] Further, the single missing point search algorithm includes the following steps:

[0087] Calculate the distance D between two adjacent points in the X direction for each group of points i , and sort the distance D i to generate D i ’, take the median D i ’(I / 2) of D i ’, and obtain the median value of the distance as the actual ink droplet spacing D = D i ’(I / 2), where I is the number of all samples; draw a circle with a radius of a set value with the current point plus a standard distance D to the right as the center, and search for ink points in the inner area of the circle from each group of all points above. If there is one, it is determined as a normal ink point, and then search for ink points to the right based on the currently found ink point in turn.

[0088] Further, the continuous missing ink point search algorithm includes the following steps: When it is detected that several adjacent ink droplets are all missing points, the hole positions are re-corrected with reference to the results of the adjacent front row and back row; and so on, until the image corresponding point coordinates of all nozzle numbers in this row are found and numbered. Among them, the hole position correction is based on the characteristics of the nozzles of nozzle 1 to know that the projected distances of all holes in the ideal state are equal.

[0089] Step 3: Conduct ink droplet observation and measurement on the selected abnormal nozzles. Example 2

[0090] As Figures 1 - 16 shown, an optimized method for quickly inspecting the inkjet state of nozzles through a camera includes the following steps:

[0091] Step 1: Perform ink droplet observation and positioning on the inkjet printhead 1, select a suitable inkjet printhead 1 as the qualified printhead 1, and record the ink droplet parameters of the qualified printhead 1 currently sprayed on the test film 5;

[0092] Specifically, the ink droplet observation and positioning includes the following steps:

[0093] Step 1.1: Determine whether the ink droplets ejected from the nozzles of the printhead 1 meet the ink droplet requirements set in inkjet printing;

[0094] Step 1.1.1: When the ink droplets ejected from the nozzles of the printhead 1 meet the ink droplet requirements set in inkjet printing, jump to Step 2;

[0095] Step 1.1.2: When the ink droplets ejected from the nozzles of the printhead 1 do not meet the ink droplet requirements set in inkjet printing, collect the next nozzle and repeat Step 1.1;

[0096] Until a printhead 1 that can eject ink droplets meeting the ink droplet requirements set in inkjet printing is selected.

[0097] More specifically, in Step 1.1, an ink droplet observation instrument 2 is used to sequentially check each row of nozzles of the printhead 1, and it is determined whether the ink droplets ejected from the nozzles meet the preset ink droplet volume requirements according to the volume of the ejected ink droplets. Those that meet the requirements are determined as qualified. When the current ink droplet volume is qualified, the operation of Step 1.1 is stopped, and the offset of the first ink droplet in the current row is recorded, and then it is marked in the ink droplet image of the current row according to the recorded position.

[0098] For example: Through the mutual cooperation of the ink droplet observation instrument 2 and the coaxial light source 6 located on both sides of the printhead 1, each row of nozzles is sequentially checked, and it is determined whether it is qualified according to the volume. If the current ink droplet volume is qualified, the first step operation is stopped, and the offset of the first ink droplet in the current row is recorded, and then it is marked in the figure (assuming that the printhead 1 has 3 rows of nozzles, the following results are Gap: 2 in the first row, Gap: 1 in the second row, Gap: 1 in the third row).

[0099] Step 2: Obtain the ink droplet image on the test film 5 in Step 1, filter out the interfering ink dots in the ink droplet image through the ink droplet distribution state and morphology screening method, obtain the filtered screened ink droplet image, and identify the hole positions of the abnormal nozzles corresponding to the abnormal ink droplets through the ink droplet distribution state and morphology screening method. Abnormal interference includes ink droplet scatter points, ink droplet scratches, or ink droplet adhesion; the nucleus is smaller than a normal ink droplet and larger than the median value of the ink droplet scatter points; normal ink droplets cannot be removed when initially removing the interfering ink dots.

[0100] Specifically, in step 2, obtaining the ink droplet image on the test film 5 in step 1 includes the following steps:

[0101] Collect the ink droplet image of the ink droplets printed on the test film 5 by the spray holes selected in the current step 1 through the grayscale line scan camera 4. And a backlight 3 is arranged on the lower side of the test film 5, and the test film 5 uses a transparent test film. It is convenient to cooperate with the backlight 3 to improve the accuracy and stability of ink dot sampling.

[0102] Furthermore, the ink droplet distribution state and morphology screening method includes the following steps:

[0103] Step 2.1, for the abnormal interference on the ink droplet image, perform binary operation through a set threshold, select the corresponding kernel, and perform preliminary removal of interfering ink dots through morphological filtering opening operation of the image to obtain the initially filtered ink droplet image after preliminary removal of interfering ink dots;

[0104] Step 2.2, take the center points of all the ink droplets in the initially filtered ink droplet image, sort the Y values of each ink droplet image, and then automatically segment them into several groups of values according to the sorting results. At this time, each group of ink droplets still contains interfering points.

[0105] Furthermore, in step 2.2, the segmentation algorithm includes the following steps:

[0106] Step 2.2.1, mark the pixel value of the sample point in the Y direction in the image as Y i , where i represents all the sample quantities, for Y i with the same value, merge and record it as Y n , where n represents the number of merged Y samples and records the quantity;

[0107] Step 2.2.2, dynamically segment the merged Y n into clusters (heaps). When Y n - Y n-1 = 1, it is considered that the two coordinates are adjacent and classified into the same cluster. When Yn - Yn - 1 > 1, it is considered that the two coordinates are not adjacent and classified into different clusters. Denote each cluster as C m , M represents the number of clusters, calculate the mean value A m of all the points in the cluster and the quantity N m ;

[0108] Step 2.2.3, screen out the 3 clusters with the largest number of points in the cluster;

[0109] Step 2.2.4, take the mean value of the adjacent cluster centers to calculate the segmentation line value S m = (A m + A m+1 ) / 2;

[0110] Step 2.2.5, segment the original samples according to the segmentation line value S m for actual corresponding original image records.

[0111] Step 2.3, calculate the distances between all points in this group and the central values of their respective classification clusters. If the distance is greater than the set threshold, it is considered interference, and this point is deleted from this group. At this time, the ink droplet spacing basically removes the interference of nozzle missing and scattered ink droplets, and a deep-filtered ink droplet image is obtained.

[0112] For example: calculate all points Y in this group n and the central value A m (62, 162, 269) of their respective classification clusters, and the distance D n = Y n - A m . If the distance is greater than the set threshold (assumed to be 10), it is considered interference, and this point is deleted from this group.

[0113] Step 2.4, check for single missing points in the deep-filtered ink droplet image through the single missing point search algorithm, and check for single missing points in the deep-filtered ink droplet image through the continuous missing ink point search algorithm; until the image corresponding point coordinates of all nozzle numbers in this row are found and numbered.

[0114] Furthermore, the single missing point search algorithm includes the following steps:

[0115] Calculate the spacing D between two adjacent points in the X direction for each group of points i , and sort the spacing D i to generate D i ’. Take the median value D i ’(I / 2) of D i ’ to obtain the median value of the spacing as the actual ink droplet spacing D = D i ’(I / 2), where I is the number of all samples; draw a circle with a radius of the set value with the center at the current point plus one standard spacing D to the right. Search for ink points in the inner area of the circle from each group of all points above. If there is one, it is determined as a normal ink point, and then search for ink points to the right in turn according to the currently found ink point.

[0116] For example, when calculating the spacing D between two adjacent points in the X direction for each group of points, D i = X i - X i-1 (I represents the number of all samples), and sort the spacing D i to generate D i ’, take the median value D i ’ of D i ’(I / 2) to obtain the median value of the spacing as the actual ink droplet spacing D = D i’(I / 2). Draw a circle with a radius of a set value with the center at a standard spacing D increased to the right from the current point. Search for ink dots among all points in each group from the top within the circular area. If any are found, they are determined to be normal ink dots, and then search for ink dots to the right in sequence based on the currently found ink dots.

[0117] The following is the search process:

[0118] Assume that for the second row of the droplet observation result, Gap = 1, for classification C of the second row 2 All points are sorted in the X direction. Since Gap = 1, it is considered that there is no missing head start point, so the point with the smallest X value is marked as X 21 (No. 1 in the second row);

[0119] Within the range Traverse all points in the second row. If a point meets the range, mark this point as X 22 ;

[0120] Within the range Traverse all points in the second row. If no point meets the requirement, it is considered that the missing point X 23 = X 22 + D;

[0121] Within the range Traverse all points in the second row. If a point meets the range, mark this point as X 24 .

[0122] Furthermore, the algorithm for finding consecutive missing ink dots includes the following steps: When it is detected that several adjacent ink droplets are all missing points, the hole positions are re-corrected with reference to the results of the adjacent front row and back row; and so on, find the image corresponding point coordinates of all nozzle holes in this row and mark the serial numbers. Among them, correcting the hole positions is based on the characteristics of the nozzle holes of nozzle 1, and it is known that the projected spacing of all holes in the ideal state is equal.

[0123] For example, when it is detected that X 23 at is a missing point, and X 24 is also a missing point, then the hole positions need to be re-corrected with reference to the results of the first row and the third row. X 24 =(X 14 + X 34 ) / 2. Because of the characteristics of the nozzle holes of nozzle 1, the projected spacing of all holes is theoretically equal. When it is detected that X 24 at is a missing point, and X 25 is also a missing point, then X 25 =(X 15 + X 35 ) / 2.

[0124] And so on, find the image corresponding point coordinates of all nozzle holes in this row and mark the serial numbers.

[0125] Step 3: Conduct ink droplet observation and measurement on the selected abnormal nozzle holes. Embodiment III

[0126] Based on Embodiment I or Embodiment II, an optimized system for quickly inspecting the inkjet state of nozzle holes by a camera is implemented using an optimized method for quickly inspecting the inkjet state of nozzle holes by a camera, including:

[0127] A data acquisition module, which is used to perform ink droplet observation and positioning on the inkjet print head 1, select a suitable inkjet print head 1 as a qualified print head, and record the ink droplet parameters of the qualified print head 1 currently sprayed on the test film 5.

[0128] A data processing module, which is used to obtain the ink droplet image on the test film 5, filter out the interfering ink dots in the ink droplet image through the ink droplet distribution state and morphology screening method, obtain the filtered screened ink droplet image, and identify the hole positions of the abnormal nozzle holes corresponding to the abnormal ink droplets through the ink droplet distribution state and morphology screening method.

[0129] A data compounding module, which is used to conduct ink droplet observation and measurement on the selected abnormal nozzle holes.

[0130] Working principle:

[0131] For the optimized method and optimized system for quickly inspecting the inkjet state of nozzle holes of the present invention, the angle position state of the print head 1 is detected through the ink droplet observation system to establish the connection between the points on the transparent film and the actual nozzle holes for precise positioning; after binarizing the picture to extract blobs and then performing morphological opening operation, most of the scattered points / adhesions and thin film scratches in the original state are removed, and then the ink droplet rows are automatically segmented according to the position distribution state in the Y direction; the incorrect nozzle hole numbers caused by the uneven ink droplet spacing caused by the warping of the thin film are solved by dynamically updating the reference points; the precise positions of the continuous abnormal nozzle hole sequences during large-area nozzle hole blockage are calculated by referring to the correct X-direction positions of the other rows of nozzle holes.

[0132] Enlightened by the ideal embodiments of the present invention, through the above description, relevant personnel can make various changes and modifications completely within the scope of not deviating from the technical idea of this invention. The technical scope of this invention is not limited to the content in the specification, and the technical scope must be determined according to the scope of the claims.

Claims

1. An optimization method for quickly checking the inkjet status of a nozzle through a camera, characterized in that: The following steps are involved: Step 1, observe and locate the ink droplets of the inkjet printing nozzle, select a suitable inkjet printing nozzle as a qualified nozzle, and record the ink droplet parameters currently sprayed on the test film by the qualified nozzle; Step 2, obtaining an ink drop image on the test film, filtering out interfering ink dots in the ink drop image by ink drop distribution state and morphology screening method, obtaining a filtered screening ink drop image, and finding out the abnormal nozzle position corresponding to the abnormal ink drop by ink drop distribution state and morphology screening method; Ink droplet distribution state and morphology screening method, including: Step 2.1, for the abnormal interference of the ink droplet image, a binarization operation is performed using a set threshold, and a corresponding kernel is selected to preliminarily remove the interfering ink dots using an image morphological filter operation to obtain a preliminarily filtered ink droplet image after the interfering ink dots are preliminarily removed; Step 2.2, take the center point of all ink droplets in the initial filtered ink droplet image, sort the Y value of each ink droplet image, and automatically segment it into several groups of values ​​using a segmentation algorithm according to the sorting result, and each group of ink droplets still contains interference points; The segmentation algorithms include: Step 2.2.1, the pixel value of the sample point in the Y direction in the image is marked as Y i , i is the number of all samples, for Y i The same values ​​are combined and recorded as Y n , n represents the number of Y sample records after merging; Step 2.2.2, after merging Y n Dynamic partitioning clusters; when Y n -Y n-1 = 1, the two coordinates are adjacent and classified into the same cluster; n -Y n-1 >1, the two coordinates are not adjacent and are classified into different clusters. Each cluster is denoted by C m , M is the number of clusters, calculate the mean A of all points in the cluster m and the number N m ; Step 2.2.3, select the three clusters with the largest number of points; Step 2.2.4: Take the mean of the centers of adjacent clusters to calculate the dividing line value S between clusters m =(A m +A m+1 ) / 2; Step 2.2.5: The original sample is divided into two groups according to the segmentation line value S. m To split; Step 3: perform ink droplet observation and measurement on the screened abnormal nozzles.

2. The optimization method for quickly checking the inkjet status of a nozzle through a camera according to claim 1, characterized in that: The ink droplet observation and positioning comprises the following steps: Step 1.1, determining whether the ink droplets ejected from the nozzle orifice of the nozzle meet the ink droplet requirements set in inkjet printing; Step 1.1.1, when the ink droplets ejected from the nozzle orifice of the nozzle meet the ink droplet requirements set in inkjet printing, jump to step 2; Step 1.1.2, when the ink droplets ejected from the nozzle orifice of the nozzle do not meet the ink droplet requirements set in inkjet printing, collect the next nozzle orifice and repeat step 1.1; Until a nozzle is selected that can spray ink droplets that meet the requirements set in inkjet printing.

3. The optimization method for quickly checking the inkjet status of a nozzle through a camera according to claim 2, characterized in that: In step 1.1, an ink droplet observer is used to check the nozzles of each row of the nozzles in turn, and the volume of the ink droplets ejected from the nozzles is used to determine whether it meets the preset ink droplet volume requirements. If it meets the requirements, it is determined to be qualified. If the current ink droplet volume is qualified, the operation of step 1.1 is stopped, and the offset of the first ink droplet in the current row is recorded, and then marked in the ink droplet map of the current row according to the recorded point.

4. The optimization method for quickly checking the inkjet status of a nozzle through a camera according to claim 3, characterized in that: In step 2, obtaining the ink droplet image on the test film in step 1 includes the following steps: The ink droplet image printed on the test film by the nozzle selected in the current step 1 is collected by a grayscale line scan camera.

5. The optimization method for quickly checking the inkjet status of a nozzle through a camera according to claim 4, characterized in that: The ink droplet distribution state and morphology screening method also includes: Step 2.3, calculate the distance between all points in the group and the center value of each classification cluster. If the distance is greater than the set threshold, it is considered interference, and the point is deleted from the group. At this time, the ink drop spacing basically eliminates the interference of nozzle missing and scattered ink droplets, and obtains a deep filter ink drop image; Step 2.4, use the algorithm for finding single missing points to check for single missing points in the deep filter ink droplet image, and use the algorithm for finding continuous missing points to check for single missing points in the deep filter ink droplet image; until the image corresponding point coordinates of all nozzle numbers in this row are found and the serial numbers are marked.

6. The optimization method for quickly checking the inkjet status of a nozzle through a camera according to claim 5, characterized in that: The algorithm for finding a single missing point includes the following steps: Calculate the distance D between two adjacent points in the X direction of each group of points i , and the spacing D i Sort and generate D i ', for D i 'Take the median value D i '(I / 2), get the median value of the spacing as the actual ink drop spacing D = D i '(I / 2), I is the number of all samples; Draw a circle with a set radius as the center by adding a standard spacing D to the right of the current point. Search for ink dots from all the points in each group above in the area inside the circle. If they exist, they are determined to be normal ink dots. Search for ink dots to the right based on the currently found ink dots.

7. The optimization method for quickly checking the inkjet status of a nozzle through a camera according to claim 6, characterized in that: The algorithm for finding consecutive missing ink dots includes the following steps: when it is detected that several adjacent ink droplets are all missing dots, the hole positions are corrected again with reference to the results of the adjacent front and back rows; and by analogy, the image corresponding point coordinates of all nozzles in this row are found and the serial numbers are marked.

8. The optimization method for quickly checking the inkjet status of a nozzle through a camera according to claim 1, characterized in that: Abnormal interference includes ink droplet scatter, ink droplet scratch, or ink droplet adhesion; The nucleus is smaller than a normal ink droplet and larger than the median value of the ink droplet scatter points; Normal ink droplets cannot be removed during the preliminary removal of interfering ink dots.

9. An optimized system for quickly checking the inkjet status of a nozzle through a camera, characterized in that: The method is implemented by using an optimization method for quickly checking the inkjet state of a nozzle through a camera as described in any one of claims 1 to 6, comprising: A data acquisition module, the data acquisition module is used to observe and locate ink droplets of inkjet printing nozzles, select a suitable inkjet printing nozzle as a qualified nozzle, and record the parameters of ink droplets currently sprayed on the test film by the qualified nozzle; A data processing module, the data processing module is used to obtain an ink drop image on the test film, filter out interfering ink dots in the ink drop image by ink drop distribution state and morphology screening method, obtain the filtered screened ink drop image, and identify the hole position of the abnormal nozzle corresponding to the abnormal ink drop by the ink drop distribution state and morphology screening method; A data compound module is used to observe and measure ink droplets on the screened abnormal nozzles.

Images