Digital jet printing machine nozzle automatic detection method and device based on machine vision

Abstract

The invention discloses a digital jet printing machine nozzle automatic detection method based on machine vision, and the method comprises the steps: collecting a spray needle image of a nozzle, obtaining a minimum enclosing rectangle containing the contour information of a spray needle, recording the position of a needle tip, and dividing a liquid leakage detection region and an ink jet detection region; a plurality of jet needle ink jet state diagrams are collected, the number of pixels with specified pixel values in the liquid leakage detection area and the ink jet detection area is calculated, and whether ink jet abnormity exists or not is analyzed; performing superposition processing on the plurality of jet needle ink-jet state diagrams, and performing pixel-by-pixel subtraction on the processed image and the jet needle image to obtain a jet needle ink-jet effect diagram; according to the ink jet effect picture of the jet needle, the ink jet angle is measured, and whether the ink jet angle is abnormal or not is judged. The invention further discloses a digital jet printing machine nozzle automatic detection device based on machine vision. According to the invention, by adopting a related method of machine vision, the automatic ink-jet quality detection and calibration of the digital jet printing machine nozzle are realized, and the labor cost is greatly reduced while the detection accuracy is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of digital printing, and in particular to a device and method for automatically detecting a printhead of a digital inkjet printer based on machine vision. Background Art

[0002] Digital inkjet printer is a new type of printing and dyeing equipment. Compared with the traditional dyeing process based on bleaching and dyeing, digital inkjet printer has the advantages of high precision and high energy utilization, and has been widely used in the textile, printing and other industries. The inkjet scheme of the digital inkjet printer nozzle can be divided into solid inkjet and liquid inkjet according to the working principle. In the textile industry, common digital inkjet printers use liquid inkjet schemes. Most traditional liquid inkjet schemes require a heating device to be set on the nozzle. By heating the nozzle, the ink generates bubbles and is sprayed onto the substrate. This inkjet scheme has the problems of uncontrollable droplet size and liquid waste. At the same time, the existence of the heating link also limits the types of inks.

[0003] The patent with application number 202311677664.7 discloses a printing device based on high-frequency mechanical vibration to atomize ink. The device is equipped with a brake device and an airflow guide unit at the position of the nozzle. The ink flows into the nozzle under the positive pressure of the ink supply system. The brake device drives the nozzle to vibrate at high frequency, and the ink is atomized and sprayed out from the needle tip. The airflow guide unit guides the atomized ink to the substrate by controlling the size and direction of the airflow. This new type of inkjet device has a simple structure and is easy to install. At the same time, since it no longer requires heating, this type of device can adapt to more types of inks. However, the atomization method based on mechanical vibration is greatly affected by the physical properties and installation accuracy of the nozzle, and the ink output is uncontrollable, making it difficult to effectively combine into a printing array.

[0004] Patent application number 202410018975.0 discloses an ink supply system for arraying the above-mentioned mechanical vibration-based printing device. The device takes into account that the liquid output of each nozzle unit in the nozzle array is affected by its separate liquid supply pipeline in addition to the vibration. The device uses an independent pump device and a pipeline design from the liquid source to the nozzle to ensure that the liquid output of each inkjet unit is consistent, but in actual tests, the ink output uniformity of each nozzle is still unsatisfactory. Summary of the invention

[0005] In order to solve the problem of uneven ink output from the nozzle needle of the nozzle, the present invention provides an automatic detection method for the nozzle of a digital inkjet printer based on machine vision. A camera is used to obtain the inkjet working image of the nozzle needle, and whether there is inkjet abnormality is automatically detected, which is beneficial to improving the ink output uniformity of the nozzle needle.

[0006] In order to achieve the above-mentioned invention object, the specific technical scheme adopted by the present invention is as follows:

[0007] A method for automatically detecting a printhead of a digital inkjet printer based on machine vision, comprising:

[0008] Collect the nozzle needle image, obtain the minimum circumscribed rectangle containing the nozzle outline information, record the needle tip position and divide the leakage detection area and inkjet detection area;

[0009] Collect multiple needle inkjet status images, calculate the number of pixels with specified pixel values ​​in the leakage detection area and the inkjet detection area, and analyze whether there is inkjet abnormality;

[0010] Multiple inkjet state images of the nozzle are superimposed and processed, and the processed image is subtracted from the nozzle image pixel by pixel to obtain the nozzle inkjet effect image;

[0011] Measure the inkjet angle according to the inkjet effect diagram of the nozzle, and determine whether there is an abnormal inkjet angle of the nozzle.

[0012] Preferably, the needle image is binarized using the Otsu binarization method to remove noise from the binarized image, and the needle contour information is obtained through edge detection.

[0013] Preferably, all the needle contour information is traversed and the minimum circumscribed rectangle is calculated, and screening is performed according to the area S and aspect ratio K of each minimum circumscribed rectangle. The position of the needle tip in the minimum circumscribed rectangle after brushing is recorded, and the area above the needle tip is used as the leakage detection area, and the area below the needle tip is used as the inkjet detection area.

[0014] Preferably, the collected inkjet needle inkjet state diagram is binarized using the Otsu binarization method, and the number of pixels with a pixel value of 255 in the leakage detection area and the inkjet detection area is calculated and recorded as N1 and N2;

[0015] When N1 is greater than 200, it means that there is ink leakage from the nozzle;

[0016] When N2 is less than 200, it means that the inkjet amount of the nozzle is too low.

[0017] Preferably, a plurality of inkjet needle inkjet state images are superimposed, and each pixel value after superposition is an average value of the pixel value at the corresponding position of the initial image.

[0018] Preferably, according to the inkjet effect diagram, Gaussian filtering and image corrosion operations are used to remove scattered points, and the inkjet effect diagram obtained can be regarded as a mask of the inkjet effect.

[0019] Preferably, the specific process of measuring the inkjet angle according to the inkjet effect diagram of the inkjet needle is as follows:

[0020] The needle tip is taken as vertex C, and the coordinates are (x p ,y p), the row height of the entire image is H, and the width is W; set two active points A and B on the bottom edge of the image, where point A moves from (0, H-1) to (x p ,H-1), point B moves from (W-1,H-1) to (x p ,H-1) movement, connect the two straight lines AC and BC, and record their lengths as L respectively ac and L bc ; Count the number of pixels with a pixel value of 255 on lines AC and BC as N ac and N bc ; When points A and B move to the point where the conditions are met When , the detection ends; taking point C as the vertex, calculate the inkjet angle ∠ACB;

[0021] When ∠ACB is within the range of 35°~40°, the inkjet needle sprays ink normally, otherwise it is recorded that the inkjet needle sprays ink abnormally.

[0022] Preferably, the method further includes calibrating the ink output of the inkjet needle:

[0023] The ink output of each nozzle per unit time is measured in turn, and the nozzles to be adjusted are selected according to the set error tolerance value;

[0024] The vibration frequency of the corresponding nozzle needle is adjusted, and the ink output is recalculated until the ink output of the nozzle needle is within the set error tolerance value compared with the ink output of the reference needle.

[0025] According to the above method, the present invention also provides a digital inkjet printer nozzle automatic detection device based on machine vision, comprising:

[0026] Base;

[0027] A bracket having three axes of freedom on the base;

[0028] The light source and camera are mounted on a bracket.

[0029] And a processor connected to the camera is used for image processing and implementing the above-mentioned automatic detection method for the printhead of a digital inkjet printer.

[0030] Preferably, an ink cartridge and an electronic analytical balance are installed on the bracket for calibrating and detecting the ink output of the inkjet needle.

[0031] The present invention realizes automatic inkjet quality detection and calibration of the digital inkjet printer nozzle by adopting relevant methods of machine vision, effectively improving the detection accuracy while greatly reducing labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A schematic diagram of a detection device provided by the present invention;

[0033] Figure 2 A schematic diagram of a nozzle tested by the present invention;

[0034] Figure 3 It is the overall framework flow chart of the present invention;

[0035] Figure 4 It is a flow chart of the inkjet abnormality detection stage of the inkjet needle of the present invention;

[0036] Figure 5 Schematic diagram of the inkjet detection effect of the present invention; wherein Figure (a) is an image in a non-inkjet state, Figure (b) is an image in an inkjet state, Figure (c) is a binarization result of the image in the inkjet state, Figure (d) is a binarization difference between an image in the inkjet state and an image in the non-inkjet state, and Figure (e) is an inkjet angle detection result;

[0037] Figure 6 It is a schematic diagram of step (2-5) of the present invention. DETAILED DESCRIPTION

[0038] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0039] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0040] The detection object of the present invention is a digital inkjet printer nozzle. The processing platform used in the present invention is a combination of Intel i7-12700H and 16G RAM. The maximum frame rate of the camera used is 90FPS, and the resolution of the electronic analytical balance used is 0.001 grams.

[0041] A typical embodiment of the present invention is described in detail below with reference to the accompanying drawings:

[0042] The digital inkjet printer nozzle automatic detection device based on machine vision in this embodiment includes: a base; a bracket with three-axis freedom on the base; a light source and a camera installed on the bracket, and a processor connected to the camera.

[0043] Specific as Figure 1 The detection device shown in the figure mainly includes three parts:

[0044] Three-axis robotic arm: As the base of other modules, it can realize movement in the three axes of X, Y, and Z. Figure 1The motor is connected to the host computer through a communication line, and the host computer controls the real-time position and movement speed of the motor through the control interface provided by the motor.

[0045] Weighing module: It consists of a high-precision electronic analytical balance 40 and an ink container 41. The ink container is a hollow rectangular parallelepiped with a gap on one side for the needle to move left and right. The container is fixed on the electronic analytical balance, and the electronic analytical balance is connected to the host computer (processor) through a communication line.

[0046] Vision module: It is mounted on the bracket 50 above the three-axis robot arm in parallel with the weighing module. The two slide bars 60 and 61 are connected by a connector 63. The vision module is fixed on the slide bar 61 to achieve continuous adjustment of the position in the horizontal and vertical directions. The vision module consists of a high-speed area array camera 70 and a linear array light source 71. The light source 71 is installed directly above the camera and has an adjustable pitch angle. The camera is connected to the host computer via a communication cable.

[0047] In another preferred embodiment, a method for detecting a digital inkjet printer nozzle based on machine vision is also provided, comprising:

[0048] Collect the nozzle needle image, obtain the minimum circumscribed rectangle containing the nozzle outline information, record the needle tip position and divide the leakage detection area and inkjet detection area;

[0049] Collect multiple needle inkjet status images, calculate the number of pixels with specified pixel values ​​in the leakage detection area and the inkjet detection area, and analyze whether there is inkjet abnormality;

[0050] Multiple inkjet state images of the nozzle are superimposed and processed, and the processed image is subtracted from the nozzle image pixel by pixel to obtain the nozzle inkjet effect image;

[0051] Measure the inkjet angle according to the inkjet effect diagram of the nozzle, and determine whether there is an abnormal inkjet angle.

[0052] The digital inkjet printer nozzle detection method of this embodiment can be implemented according to the above-mentioned detection device embodiment, and can also be implemented in conjunction with other visual detection platforms.

[0053] Combination Figure 1 In the device embodiment, the digital inkjet printer nozzle detection method based on machine vision is as follows Figure 3 Proceed as shown in the steps:

[0054] (1) Nozzle detection settings;

[0055] (2) Abnormal detection of inkjet needle;

[0056] (3) Calibration of ink output from the nozzle.

[0057] Step (1) specifically includes:

[0058] Set up the nozzle to be tested in front of the nozzle detection device and operate the three-axis robot arm. Figure 2 As shown, the first needle 81 of the nozzle 80 appears in the center of the camera screen; set the camera exposure time to 5000us, the image acquisition format to Mono8, and the image resolution to 2048x2048; adjust the light source angle so that the light source can illuminate the position below the needle; set the number of needles of the nozzle, the number of control boards, and confirm the initial address of the board communication in the software.

[0059] The flowchart of step (2) is as follows Figure 4 and Figure 5 As shown, specifically including:

[0060] (2-1) When the inkjet needle is not spraying ink, a frame of image is collected, which is recorded as M1. The segmentation threshold T is calculated using the Otsu method, and the image M1 is binarized using the threshold.

[0061] (2-2) For the binarized image, a 5x5 convolution kernel is first used to perform an erosion operation on the image to remove small noise points, and then further edge detection is performed to obtain contour information; all contour information is traversed and the minimum enclosing rectangle is calculated; the area S and aspect ratio K of all minimum enclosing rectangles are calculated, and screening is performed according to the following rules:

[0062]

[0063] The minimum bounding rectangle obtained after screening contains the location information of the target needle, so as to determine and record the location of the needle tip (x p ,y p ), and the area above the needle tip is used as the leakage detection area, and the area below the needle tip is used as the inkjet detection area.

[0064] (2-3) Start the inkjet, and collect three inkjet images M2, M3, and M4 at intervals of 1000 ms. First, the three images are binarized in the same way, and the number of pixels with a pixel value of 255 in the leakage detection area and the inkjet detection area is calculated and recorded as N1 and N2. When N1 is greater than 200, it means that the nozzle has a certain degree of ink leakage. When N2 is less than 200, it means that the nozzle has too little ink. If either of the two does not meet the requirements, it is judged that the nozzle is abnormal.

[0065] (2-4) If the abnormality detection fails, the needle with that number is marked in the report and the abnormality is recorded. The program controls the three-axis robot arm so that the camera angle faces the next needle and repeats step (2).

[0066] The images M2, M3 and M4 are superimposed, and the superposition method is shown in Formula 3:

[0067]

[0068] Pix in the formula n (x, y) represents the grayscale value of the pixel at the coordinate (x, y) on the corresponding image.

[0069] Image M5 is obtained. Similarly, image M5 is subtracted from image M1 pixel by pixel to obtain inkjet image M6. Image M6 is binarized using Otsu's method, and then scattered points are removed using Gaussian filtering and image erosion operations to obtain image M7, which can be regarded as a mask for the inkjet effect of the needle.

[0070] (2-5) After obtaining the mask of the inkjet effect of the nozzle, start measuring the inkjet angle.

[0071] like Figure 6 As shown, the needle tip is taken as vertex C, and its coordinates are (x p ,y p ), let the row height of the whole image be H and the width be W; set two active points A and B on the bottom edge of the image, where point A moves from (0, H-1) to (x p ,H-1), point B moves from (W-1,H-1) to (x p ,H-1) movement, connect the two straight lines AC and BC, and record their lengths as L respectively ac and L bc ; Count the number of pixels with a pixel value of 255 on lines AC and BC as N ac and N bc ; When points A and B move to the point where the following conditions are met, the detection ends:

[0072]

[0073] N in the formula ac and L ac When the ratio is greater than 0.67, it indicates that edge AC is already at the edge of the inkjet area.

[0074] Take point C as the vertex and calculate ∠ACB:

[0075]

[0076] When ∠ACB is within the range [35,40], in degrees, it is recorded that the inkjet needle is ejecting ink normally; otherwise, it is recorded that the inkjet needle is ejecting ink abnormally.

[0077] (2-5) Repeat the above steps until all the nozzles on the nozzle are tested.

[0078] In this embodiment, step (3) specifically includes:

[0079] (3-1) By controlling the three-axis robotic arm, the spray needle is extended into Figure 1 In the opening of the ink cartridge 41 shown, the ink ejection amount is detected sequentially starting from the first needle.

[0080] (3-2) Record the initial reading W1 of the electronic balance, then start the inkjet system of the corresponding nozzle, set the timer to turn off the inkjet after 40 seconds, wait for 5 seconds for the reading to stabilize, and then record the reading W2. Record the difference W between the two. Also record the vibration frequency and nozzle pressure of the nozzle simultaneously.

[0081] (3-3) Loop step (3-2) to complete the detection of all nozzles. After completing the inkjet volume detection of the entire nozzle, select the inkjet volume of a needle as the benchmark according to the pre-settings, and screen out the nozzles that meet the requirements and those that do not meet the requirements with an error tolerance value of ±5%, and put the nozzles that do not meet the requirements into the queue to be adjusted.

[0082] (3-4) Debug the needles in the adjustment queue separately, and modify the vibration frequency of the corresponding needles with 20% of the current vibration frequency of the needles as the basic step. After the modification is completed, repeat step (3-2). When the step is too large, adjust the step value to half of the original step. Adjust multiple times until the ink output of the needle is within ±5% of the ink output of the reference needle.

[0083] After completing the individual calibration of all the nozzles to be adjusted, the ink jetting uniformity of the nozzles on the print head is guaranteed.

[0084] The above description is only an example of a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatic detection of digital inkjet printer nozzles based on machine vision, characterized in that: include: Collect the nozzle needle image, obtain the minimum circumscribed rectangle containing the nozzle outline information, record the needle tip position and divide the leakage detection area and inkjet detection area; Collect multiple needle inkjet status images, calculate the number of pixels with specified pixel values ​​in the leakage detection area and the inkjet detection area, and analyze whether there is inkjet abnormality; Multiple inkjet state images of the nozzle are superimposed and processed, and the processed image is subtracted from the nozzle image pixel by pixel to obtain the nozzle inkjet effect image; Measure the inkjet angle according to the inkjet effect diagram of the nozzle, and determine whether there is an abnormal inkjet angle.

2. The method for automatic detection of a digital inkjet printer nozzle according to claim 1, characterized in that: The needle image is binarized using the Otsu binarization method to remove noise from the binarized image, and the needle contour information is obtained through edge detection.

3. The method for automatic detection of a digital inkjet printer nozzle according to claim 2, characterized in that: Traverse all the needle contour information and calculate the minimum circumscribed rectangle, screen according to the area S and aspect ratio K of each minimum circumscribed rectangle, record the position of the needle tip in the minimum circumscribed rectangle after brushing, use the area above the needle tip as the leakage detection area, and use the area below the needle tip as the inkjet detection area.

4. The method for automatic detection of a digital inkjet printer nozzle according to claim 1, characterized in that: The collected inkjet state image of the nozzle is binarized using the Otsu binarization method, and the number of pixels with a pixel value of 255 in the leakage detection area and the inkjet detection area is calculated and recorded as N1 and N2; When N1 is greater than 200, it means that there is ink leakage from the nozzle; When N2 is less than 200, it means that the inkjet amount of the nozzle is too low.

5. The method for automatic detection of a digital inkjet printer nozzle according to claim 1, characterized in that: Multiple inkjet state images of the nozzle are superimposed, and the pixel value after superposition is the average value of the pixel value at the corresponding position of the initial image.

6. The method for automatic detection of a digital inkjet printer nozzle according to claim 1, characterized in that: According to the inkjet effect diagram, scattered points are removed by using Gaussian filtering and image corrosion operations, and the inkjet effect diagram obtained can be regarded as a mask of the inkjet effect.

7. The method for automatic detection of a digital inkjet printer nozzle according to claim 6, characterized in that: The specific process of measuring the inkjet angle according to the inkjet effect diagram of the nozzle is as follows: The needle tip is taken as vertex C, and the coordinates are (x p ,y p ), the row height of the entire image is H, and the width is W; set two active points A and B on the bottom edge of the image, where point A moves from (0, H-1) to (x p ,H-1), point B moves from (W-1,H-1) to (x p ,H-1) movement, connect the two straight lines AC and BC, and record their lengths as L respectively ac and L bc ; Count the number of pixels with a pixel value of 255 on lines AC and BC as N ac and N bc ; When points A and B move to the point where the conditions are met When , the detection ends; Take point C as the vertex and calculate the inkjet angle ∠ACB; When ∠ACB is within the range of 35°~40°, the inkjet needle sprays ink normally, otherwise it is recorded that the inkjet needle sprays ink abnormally.

8. The method for automatic detection of a digital inkjet printer nozzle according to claim 1, characterized in that: Also includes needle ink output calibration: The ink output of each nozzle per unit time is measured in turn, and the nozzles to be adjusted are selected according to the set error tolerance value; The vibration frequency of the corresponding nozzle needle is adjusted, and the ink output is recalculated until the ink output of the nozzle needle is within the set error tolerance value compared with the ink output of the reference needle.

9. A digital inkjet printer nozzle automatic detection device based on machine vision, characterized in that: include: Base; A bracket having three axes of freedom on the base; The light source and camera are mounted on a bracket. And a processor connected to the camera, used for image processing and implementing the automatic detection method for the digital inkjet printer nozzle according to any one of claims 1 to 8.

10. The automatic detection device for the printhead of a digital inkjet printer according to claim 9, characterized in that: The bracket is equipped with an ink cartridge and an electronic analytical balance for calibrating and detecting the ink output of the spray needle.

Images