Gluing quality detection method and system

Abstract

The invention relates to the technical field of battery manufacturing, and particularly discloses a gluing quality detection method and system. The gluing quality detection method comprises the following steps: driving a camera to return along the original path of a gluing track by using a robot which finishes gluing to carry out segmented shooting on the gluing track of a workpiece, so as to obtain a plurality of glue line images; performing parallel processing on the plurality of glue line images to obtain a plurality of colloid regions and obtain colloid boundary contours in the colloid regions; calculating a glue width mean value, a glue width standard deviation or a glue spacing mean value of the gluing track according to the glue boundary contours in the plurality of glue areas; and according to the glue width mean value, the glue width standard deviation or the glue spacing mean value, the gluing track quality is judged. The gluing robot is used for driving the camera to return along the gluing track along the original path for shooting, the camera moves along the whole gluing path and collects images through the track synchronous flying shooting technology, and the problem of detection blind areas caused by view limitation of large-size workpieces is fundamentally solved.

Description

Technical Field

[0001] This invention relates to the field of battery manufacturing technology, and in particular to a method and system for detecting the quality of adhesive coating. Background Technology

[0002] In the manufacturing process of lithium battery packs, adhesive is typically applied to the casing, cover plate, and end plates to ensure the overall structural strength of the battery pack, thereby fixing the cell modules and buffering friction between the cell modules and the casing. This adhesive application process places extremely high demands on the continuity, uniformity of width, and accuracy of the adhesive lines. Defects such as broken adhesive lines, missed areas, uneven width, or misalignment can directly lead to battery pack sealing failure, posing potential safety risks.

[0003] Currently, the commonly used adhesive coating quality inspection methods in the industry mainly suffer from the following pain points: First, many production lines still rely on manual visual inspection. This method is highly subjective, inefficient, and prone to missed detections due to visual fatigue, making it difficult to meet the cycle time and consistency requirements of automated production lines. Second, some inspection solutions using machine vision often employ fixed camera installations to photograph stationary workpieces. For large battery boxes, the field of view of a single camera is limited, making it impossible to capture the entire adhesive coating trajectory at once. This necessitates the deployment of multiple cameras, resulting in high system costs and complex calibration. If a single camera is used for segmented photography, the robotic arm needs to frequently move to multiple predetermined photography points. The movement trajectory is unrelated to the adhesive coating trajectory, which is not only inefficient but also prone to image stitching difficulties or partial obstruction of adhesive lines due to inconsistent photographic poses, making it impossible to achieve 100% inspection of the entire path.

[0004] Therefore, there is an urgent need for a low-cost, high-efficiency online inspection solution that can be closely integrated with the adhesive application process to solve the problem of quality inspection of the entire adhesive application path for large-size battery boxes. Summary of the Invention

[0005] Based on this, the purpose of this invention is to provide a method and system for detecting adhesive coating quality. The system utilizes an adhesive coating robot to drive a camera to return along the original adhesive coating trajectory to take pictures. By using trajectory synchronous flying shooting technology, the system breaks through the field of view limitations of traditional static photography and realizes full-process two-dimensional detection of adhesive coating quality for large-sized workpieces.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: This invention first provides a method for detecting the quality of adhesive coating, characterized in that it is used to detect the quality of adhesive coating on a workpiece after coating, and includes the following steps: S1. Using a robot that has completed the glue application, drive a camera to return along the original path of the glue application and take segmented pictures of the glue application trajectory of the workpiece to obtain multiple glue line images; S2. Perform parallel processing on multiple adhesive line images to obtain multiple adhesive regions and acquire the adhesive boundary contours within each adhesive region; S3. Calculate the mean glue width, standard deviation of glue width, or mean glue spacing of the coating trajectory based on the glue boundary contours within multiple colloidal regions; S4. Determine the quality of the adhesive application trajectory based on the mean adhesive width, standard deviation of adhesive width, or mean adhesive spacing.

[0007] As a further improvement of the above-mentioned solution of the present invention, the glue application is performed by a robot driving a glue gun to apply glue according to a preset glue application path; in step S1, the method of segmenting the glue application trajectory on the workpiece is as follows: a camera is installed at the end of the robot, and after the glue gun finishes applying glue, the robot drives the camera to return along the original path, and the camera segments the glue application trajectory on the workpiece.

[0008] As a further improvement to the above-mentioned solution of the present invention, the segmented shooting method is as follows: the camera is triggered to take a picture once at every predetermined displacement.

[0009] As a further improvement to the above-mentioned solution of the present invention, in step S2, the processing includes: removing noise by median filtering, enhancing the contrast between the colloid and the workpiece substrate by histogram equalization, then highlighting the colloid region by background subtraction or dynamic threshold segmentation, eliminating noise and burrs by morphological opening operation, and then using connected component analysis to screen out the effective colloid region.

[0010] As a further improvement to the above-mentioned solution of the present invention, in step S2, when the adhesive application trajectory is a single adhesive line, the average adhesive width of the adhesive application trajectory is calculated based on the adhesive boundary contour of each adhesive region; when the adhesive application trajectory is a double adhesive line, the average adhesive spacing of the adhesive application trajectory is calculated based on the two adhesive boundary contours of each adhesive region.

[0011] As a further improvement to the above-mentioned solution of the present invention, in step S2, the colloidal boundary contour of the colloidal region is obtained using the Sobel edge detection algorithm or the Canny edge detection algorithm.

[0012] As a further improvement to the above-mentioned solution of the present invention, step S3, calculating the mean glue width and standard deviation of the glue application trajectory includes: First, calculate the glue line width of each glue region: along the normal direction of the glue line centerline, calculate the pixel distance between corresponding points on both sides of the glue boundary contour, and then convert it into the actual physical size through calibration coefficients to obtain the glue line width; The average width of the adhesive line in multiple adhesive regions is calculated, which is the average adhesive width of the coating trajectory. Calculate the standard deviation of the glue line width for multiple colloidal regions.

[0013] As a further improvement to the above-described solution of the present invention, step S3, calculating the average glue spacing of the glue application trajectory includes: First, calculate the spacing between the colloids in each colloid region: calculate the distance between the center lines of two adjacent colloid lines to obtain the colloid spacing in the colloid region; The average value of the adhesive spacing between multiple adhesive regions is the average adhesive spacing of the coating trajectory.

[0014] As a further improvement to the above-mentioned solution of the present invention, in step S4, the method for determining the quality of the adhesive application trajectory is as follows: The average glue width is compared with the preset glue width standard, and the standard deviation of the glue width is compared with a predetermined value. If the average glue width is within the glue width standard range and the standard deviation of the glue width is less than the predetermined value, the glue coating quality is judged to be qualified; otherwise, the glue coating quality is judged to be unqualified. Alternatively, the average glue spacing value can be compared with a preset glue spacing standard. If the average glue spacing value is within the glue spacing standard range, the glue application quality is judged to be qualified; otherwise, the glue application quality is judged to be unqualified.

[0015] As a further improvement to the above-described solution of the present invention, step S4 further includes: Determine if there are adhesive lines with a width of 0 or adhesive lines exceeding the standard adhesive width. If there is a glue line width of 0, the defect type of the glue application trajectory is recorded as glue breakage; If the width of the glue line exceeds the glue width standard, the defect type of the glue application trajectory will be recorded as glue overflow. If the standard deviation of the adhesive width is greater than the preset value, the defect type of the adhesive application trajectory will be recorded as uneven adhesive width.

[0016] The present invention also provides a coating quality inspection system, which employs the coating quality inspection method described above, and includes: A camera, driven by a robot that has completed the glue application, returns along the original path to capture segmented images of the glue application trajectory of the workpiece to obtain multiple glue line images. An image processing module is used to process multiple adhesive coating images in parallel to obtain multiple adhesive regions and to acquire the adhesive boundary contours within each adhesive region. The calculation module is used to calculate the mean glue width, standard deviation of glue width, or mean glue spacing of the coating trajectory based on the glue boundary contours within multiple colloidal regions; and The quality assessment module is used to determine the quality of the adhesive application trajectory based on the mean adhesive width, the standard deviation of adhesive width, or the mean adhesive spacing.

[0017] Compared with the prior art, the present invention has the following beneficial effects: Achieving 100% full-path detection: This invention utilizes a glue-applying robot to drive a camera to return along the glue-applying trajectory for image capture. Through trajectory-synchronous flying shooting technology, the camera moves along the entire glue-applying path and collects images, breaking through the field-of-view limitations of traditional static photography. This enables full-process two-dimensional detection of the glue-applying quality of large-sized workpieces, fundamentally solving the problem of blind spots in detection caused by limited field of view for large-sized workpieces.

[0018] Significantly reduces system costs: The camera used in the glue application process can be directly used to capture photos of the glue application trajectory, enabling the simultaneous completion of two major functions—visual guidance positioning and glue application quality inspection—with just one camera. This eliminates the need to deploy multiple cameras for the inspection process, significantly reducing hardware costs and system complexity.

[0019] Ultimately improves inspection efficiency: The process of taking photos after applying glue is combined with the robot's return to the standby position. The movement path is smooth and continuous, avoiding unnecessary pauses and point jumps, greatly shortening the inspection cycle and increasing production efficiency by more than 3 times. It is particularly suitable for automated glue application inspection scenarios of planar box-shaped workpieces, providing a reliable quality control solution for intelligent manufacturing.

[0020] Ensuring high precision and consistency: The fully automated testing process eliminates human interference and uses precise two-dimensional image algorithms for quantitative evaluation. The test results are objective, reliable, and traceable, greatly improving the consistency and stability of product quality.

[0021] High system integration and easy deployment: The glue application path guidance and detection functions are deeply integrated into the same software platform. Based on two-dimensional image processing technology, it not only ensures detection accuracy, but also avoids the high cost of complex three-dimensional vision systems, simplifies the system architecture, and facilitates the transformation and integration on existing automated production lines. Attached Figure Description

[0022] Figure 1 This is a flowchart of a method for detecting adhesive coating quality provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating the process from applying adhesive to inspecting the quality of the adhesive application in an embodiment of the present invention. Detailed Implementation

[0023] To facilitate understanding of the present invention, a more comprehensive description will be given below with reference to specific embodiments. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of the present invention.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0025] Reference Figure 1 This embodiment proposes a method for detecting the quality of adhesive coating, which includes the following four steps S1-S4.

[0026] S1. Using a robot that has completed the glue application, drive a camera to return along the original path of the glue application and take segmented pictures of the glue application trajectory of the workpiece to obtain multiple glue line images.

[0027] In this embodiment, the glue application is performed by a robot driving a glue gun according to a preset glue application path. Combined with... Figure 2 Taking the box as an example, the following is a general introduction to the glue application process.

[0028] System Construction and Calibration: A CCD camera (or a CMOS industrial camera) and a glue gun are synchronously fixed to the end flange of the industrial robot via a rigid mounting bracket. The CCD camera axis maintains a fixed offset relationship with the center line of the glue gun nozzle. The CCD camera is equipped with a telecentric lens and a ring LED light source to ensure that the imaging plane is parallel to the glue application surface. A two-dimensional nine-point calibration method is used with a planar checkerboard calibration board to calculate the homography transformation matrix between the robot's base coordinate system and the CCD camera's two-dimensional coordinate system, completing the hand-eye calibration and laying the foundation for subsequent coordinate transformation.

[0029] Acquiring and preprocessing images of the box: The robot, carrying a CCD camera, moves to the predetermined shooting position above the box, triggering the CCD camera to acquire a two-dimensional image of the workpiece, and preprocessing the box image (such as Gaussian filtering and contrast enhancement).

[0030] Template matching and edge detection: The template matching algorithm is used to quickly and coarsely locate the corner points or feature regions of the box, and then the Canny edge detection operator is combined to accurately extract the sub-pixel contour of the edge of the box to be coated with glue.

[0031] Coordinate transformation and glue application path generation: Based on the transformation matrix obtained from calibration, the contour pixel coordinates are converted into world coordinates in the robot coordinate system, and then a glue application path that is parallel to the edge of the box and can be executed by the robot is generated through the equidistant offset algorithm.

[0032] The robot drives the glue gun to move along the glue application path at a constant speed (e.g., 200 mm / s) according to the generated glue application path. At the same time, the controller sends a signal to the glue application valve to open and close the glue application, thus completing the application of sealant.

[0033] The moment the adhesive application is completed, the robot immediately begins its return along the original application path at the same constant speed without pausing. During this process, the robot controller sends a trigger pulse to the CCD camera every fixed displacement (e.g., corresponding to 80% of the camera's field of view, i.e., 120mm) via the EtherCAT bus. The CCD camera then performs continuous, segmented aerial photography. Upon receiving the pulse, the CCD camera immediately performs exposure acquisition (exposure time set to within 1ms to avoid motion blur) and transmits the image to the industrial control computer in real time. For a 1200mm long side, approximately 10 images are needed for complete coverage, with an overlap of 30mm between adjacent images.

[0034] The trigger frequency is matched with the robot's movement speed and the CCD camera's field of view to ensure a certain overlap between adjacent acquired images, thereby achieving seamless, full-coverage image acquisition of the entire long film line. By precisely controlling the synchronization between the camera's exposure time and the robot's movement speed, motion blur is effectively suppressed.

[0035] S2. Perform parallel processing on multiple adhesive line images to obtain multiple adhesive regions and acquire the adhesive boundary contours within each adhesive region.

[0036] The industrial control computer performs parallel processing on multiple received adhesive line images, including: removing noise through median filtering, enhancing the contrast between the adhesive and the metal background of the box through CLAHE (histogram equalization), then highlighting the adhesive area using background subtraction or dynamic threshold segmentation, eliminating small edge protrusions and possible isolated noise points in the image through morphological opening operations, and finally using connected component analysis to filter out effective adhesive areas.

[0037] During the initial debugging of the system, a standard background image without adhesive is pre-collected. At this time, the grayscale value of each aerial photograph is subtracted from that of the standard background image. Since the adhesive region is newly added, the grayscale value of this region changes significantly after the difference. Then, the adhesive can be easily binarized and segmented by applying a fixed threshold.

[0038] For the processed binary image, an edge tracking algorithm (Sobel edge detection algorithm or Canny edge detection algorithm) is used to obtain the closed contour of the colloid.

[0039] S3. Calculate the mean glue width, standard deviation of glue width, or mean glue spacing of the coating trajectory based on the glue boundary contours of multiple colloidal regions.

[0040] Along the direction of the colloid, calculate the colloid width (unit: pixel) in the normal direction at regular intervals (e.g., 5 pixels). Based on the calibrated pixel equivalent (e.g., 0.05 mm / pixel), convert all pixel widths into actual physical widths (unit: mm).

[0041] When the glue application trajectory is a single glue line, calculate the average glue width of the glue application trajectory according to the glue boundary contours of each glue body area. Specifically: (1) First, calculate the glue line width of each glue body area: Along the normal direction of the glue line center line, calculate the pixel distance between the corresponding points on both sides of the glue boundary contour, and then convert it to the actual physical size through the calibration coefficient to obtain the glue line width; (2) Calculate the average value of the glue line widths of multiple glue body areas, which is the average glue width of the glue application trajectory; (3) Calculate the standard deviation of the glue width of the glue line widths of multiple glue body areas.

[0042] When the glue application trajectory is a double glue line, calculate the average glue spacing of the glue application trajectory according to the two glue boundary contours of each glue body area. Specifically: (1) First, calculate the spacing between the glues of each glue body area: Calculate the distance between the center lines of two adjacent glue lines to obtain the glue spacing of the glue body area; (2) Calculate the average value of the glue spacings of multiple glue body areas, which is the average glue spacing of the glue application trajectory.

[0043] S4. Judge the quality of the glue application trajectory according to the average glue width, the standard deviation of the glue width or the average glue spacing.

[0044] In this embodiment, for the glue application trajectory of a single glue line, the method for judging the quality of the glue application trajectory is as follows: Compare the average glue width with the preset glue width standard (2.0 - 3.0 mm), compare the standard deviation of the glue width with the preset value (0.15 mm), and judge whether there is a glue line width of 0 (i.e., broken glue) and whether there is a glue line width exceeding the glue width standard (i.e., overflow glue); if the average glue width is within the glue width standard range and the standard deviation of the glue width is less than 0.15 mm, and there is no glue line width of 0 and no glue line width exceeding the glue width standard, then judge that the glue application quality is qualified. Otherwise, judge that the glue application quality is unqualified. If there is a glue line width of 0, record the defect type of the glue application trajectory as broken glue; if there is a glue line width exceeding the glue width standard, record the defect type of the glue application trajectory as overflow glue; if the standard deviation of the glue width is greater than the preset value, record the defect type of the glue application trajectory as uneven glue width.

[0045] For the glue application trajectory of a double glue line, the method for judging the quality of the glue application trajectory is as follows: Compare the average glue spacing with the preset glue spacing standard; if the average glue spacing is within the glue spacing standard range, then judge that the glue application quality is qualified. Otherwise, judge that the glue application quality is unqualified.

[0046] Based on the above glue application quality detection method, this embodiment also provides a glue application quality detection system, which includes: A camera, which is driven by the robot that has completed glue application to return along the original glue application path to segmentally photograph the glue application trajectory of the workpiece to obtain multiple glue line images; the camera in this embodiment is a CCD camera, and is synchronously fixedly connected to the end flange of the robot through a rigid mounting bracket together with the glue gun; An image processing module is used to process multiple adhesive coating images in parallel to obtain multiple adhesive regions and to acquire the adhesive boundary contours within each adhesive region; the image processing module is equipped with an industrial control computer and image processing algorithms; The calculation module is used to calculate the mean glue width, standard deviation of glue width, or mean glue spacing of the coating trajectory based on the glue boundary contours within multiple colloidal regions; and The quality assessment module is used to determine the quality of the adhesive application trajectory based on the mean adhesive width, standard deviation of adhesive width, or mean adhesive spacing.

[0047] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0048] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for detecting the quality of adhesive coating, characterized in that, The adhesive application is performed by a robot driving a glue gun according to a preset glue application path; the adhesive application quality inspection method includes the following steps: S1. Using a robot that has completed the glue application, drive a camera to return along the original path of the glue application and take segmented pictures of the glue application trajectory of the workpiece to obtain multiple glue line images; S2. Perform parallel processing on multiple adhesive line images to obtain multiple adhesive regions and acquire the adhesive boundary contours within each adhesive region; S3. Calculate the mean glue width, standard deviation of glue width, or mean glue spacing of the coating trajectory based on the glue boundary contours within multiple colloidal regions; S4. Determine the quality of the adhesive application trajectory based on the mean adhesive width, standard deviation of adhesive width, or mean adhesive spacing.

2. The method for detecting adhesive coating quality according to claim 1, characterized in that, In step S1, the method for segmenting and photographing the adhesive application trajectory on the workpiece is as follows: a camera is installed at the end of the robot. After the adhesive application gun completes the application, the robot drives the camera to return along the original path of the adhesive application and photographs the adhesive application trajectory on the workpiece segment by segment using the camera.

3. The method for detecting adhesive coating quality according to claim 2, characterized in that, The segmented shooting method is as follows: the camera is triggered to take a picture once at predetermined displacements.

4. The method for detecting adhesive coating quality according to claim 1, characterized in that, In step S2, the processing includes: removing noise by median filtering, enhancing the contrast between the colloid and the workpiece substrate by histogram equalization, then highlighting the colloid region by background subtraction or dynamic threshold segmentation, eliminating noise and burrs by morphological opening operation, and finally screening out the effective colloid region by connected component analysis.

5. The method for detecting adhesive coating quality according to claim 1, characterized in that, In step S2, the colloidal boundary contour of the colloidal region is obtained using the Sobel edge detection algorithm or the Canny edge detection algorithm.

6. The method for detecting adhesive coating quality according to claim 1, characterized in that, In step S3, when the adhesive application trajectory is a single adhesive line, the average adhesive width of the adhesive application trajectory is calculated based on the adhesive boundary contour of each adhesive region; when the adhesive application trajectory is a double adhesive line, the average adhesive spacing of the adhesive application trajectory is calculated based on the two adhesive boundary contours of each adhesive region.

7. The method for detecting adhesive coating quality according to claim 1, characterized in that, In step S3, calculating the mean glue width and standard deviation of the glue application trajectory includes: First, calculate the glue line width of each glue region: along the normal direction of the glue line centerline, calculate the pixel distance between corresponding points on both sides of the glue boundary contour, and then convert it into the actual physical size through calibration coefficients to obtain the glue line width; The average width of the adhesive line in multiple adhesive regions is calculated, which is the average adhesive width of the coating trajectory. Calculate the standard deviation of the adhesive line width for multiple colloidal regions; And / or, in step S3, calculating the average glue spacing of the glue application trajectory includes: First, calculate the spacing between the colloids in each colloid region: calculate the distance between the center lines of two adjacent colloid lines to obtain the colloid spacing in the colloid region; The average value of the adhesive spacing between multiple adhesive regions is the average adhesive spacing of the coating trajectory.

8. The method for detecting adhesive coating quality according to claim 1, characterized in that, In step S4, the method for determining the quality of the adhesive application trajectory is as follows: The average glue width is compared with the preset glue width standard. If the average glue width is within the glue width standard range and the glue width standard deviation is less than the predetermined value, the glue coating quality is judged to be qualified; otherwise, the glue coating quality is judged to be unqualified. Alternatively, the average glue spacing value can be compared with a preset glue spacing standard. If the average glue spacing value is within the glue spacing standard range, the glue coating quality is judged to be qualified. Otherwise, the adhesive application quality is deemed unqualified.

9. The method for detecting adhesive coating quality according to claim 8, characterized in that, Step S4 also includes: Determine if there are adhesive lines with a width of 0 or adhesive lines exceeding the standard adhesive width. If there is a glue line width of 0, the defect type of the glue application trajectory is recorded as glue breakage; If the width of the glue line exceeds the glue width standard, the defect type of the glue application trajectory will be recorded as glue overflow. If the standard deviation of the adhesive width is greater than the preset value, the defect type of the adhesive application trajectory will be recorded as uneven adhesive width.

10. A glue coating quality inspection system, characterized in that, It employs the adhesive coating quality inspection method as described in any one of claims 1-9, which includes: A camera, driven by a robot that has completed the glue application, returns along the original path to capture segmented images of the glue application trajectory of the workpiece to obtain multiple glue line images. An image processing module is used to process multiple adhesive coating images in parallel to obtain multiple adhesive regions and to acquire the adhesive boundary contours within each adhesive region. The calculation module is used to calculate the mean glue width, standard deviation of glue width, or mean glue spacing of the coating trajectory based on the glue boundary contours within multiple colloidal regions; and The quality assessment module is used to determine the quality of the adhesive application trajectory based on the mean adhesive width, the standard deviation of adhesive width, or the mean adhesive spacing.

Images