Silicone dispensing quality evaluation method and apparatus

Abstract

Provided in the embodiments of the present disclosure are a silicone dispensing quality evaluation method and apparatus. The method comprises: acquiring a collected image of a silicone dispensing area; on the basis of the collected image, acquiring point cloud data; on the basis of the point cloud data, acquiring silicone feature data; on the basis of the silicone feature data, calculating the thickness of silicone after the silicone is flattened; and on the basis of the silicone feature data and the thickness of the silicone, evaluating the silicone dispensing quality.

Description

Glue dispensing quality assessment method and device

[0001] Cross-references to related publications

[0002] The present disclosure is based on Chinese Patent Publication No. 2023113232171 filed on October 12, 2023, entitled “Dispensing Quality Assessment Method and Device”, and claims the priority of the patent disclosure, and all the contents disclosed therein are incorporated into the present disclosure by reference. Technical Field

[0003] The embodiments of the present disclosure relate to the technical field of industrial intelligent detection, and in particular, to a method and device for evaluating dispensing quality. Background Art

[0004] During the production process of heat dissipation devices, dispensing robots are introduced into production lines and production workshops to replace manual work. Due to the chemical and physical properties of heat dissipation silicone, the corrosiveness to rubber hoses can cause abnormal valve closure and release, thereby affecting the amount of heat dissipation silicone dispensed.

[0005] In the quality inspection process, the mainstream detection technology is to use a 2D camera detection solution, using ROI (region of interest) and image processing technology to measure the colloid area. According to the size and proportion of the area, the threshold offset is set for judgment.

[0006] The above method can only vaguely determine whether there is glue in the area of ​​interest, and the result of determining the amount of glue dispensed is not accurate.

[0007] Summary of the Invention

[0008] The embodiments of the present disclosure provide a method and apparatus for evaluating glue dispensing quality, so as to at least solve the problem in the related art of improving the accuracy of determining the glue dispensing amount.

[0009] According to one embodiment of the present disclosure, a method for evaluating dispensing quality is provided, comprising:

[0010] Acquire an image of the dispensing area;

[0011] Acquire point cloud data based on the collected image;

[0012] Acquiring colloid characteristic data based on the point cloud data;

[0013] Calculating the thickness of the colloid after the colloid is flattened based on the colloid characteristic data;

[0014] The dispensing quality is evaluated based on the colloid characteristic data and the colloid thickness.

[0015] According to another embodiment of the present disclosure, a dispensing quality assessment device is provided, comprising:

[0016] An acquisition module is configured to acquire an image of the dispensing area through a binocular structured light camera;

[0017] A processing module is configured to obtain point cloud data based on the collected image; and

[0018] Acquire colloid feature data based on the point cloud data; and,

[0019] Calculating the thickness of the colloid after the colloid is flattened based on the colloid characteristic data; and

[0020] The dispensing quality is evaluated based on the colloid characteristic data and the colloid thickness.

[0021] According to another embodiment of the present disclosure, a computer-readable storage medium is provided, in which a computer program is stored. The computer program is configured to execute the steps of any one of the above method embodiments when running.

[0022] According to another embodiment of the present disclosure, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any one of the above method embodiments.

[0023] Through the present disclosure, by obtaining the real shape and volume information and calculating the colloid thickness, the dispensing quality can be evaluated more accurately, which solves the problem of inaccurate dispensing amount results in the prior art and thus improves the accuracy of judging the dispensing amount. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG1 is a flowchart of a method for evaluating dispensing quality according to an embodiment of the present disclosure;

[0025] FIG2 is a second flowchart of a method for evaluating dispensing quality according to an embodiment of the present disclosure;

[0026] FIG3 is a flowchart of a method for evaluating dispensing quality according to an embodiment of the present disclosure;

[0027] FIG4 is a fourth flow chart of a method for evaluating dispensing quality according to an embodiment of the present disclosure;

[0028] FIG5 is a schematic structural diagram of a dispensing quality assessment device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0029] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings and in conjunction with embodiments.

[0030] In this embodiment, a method for evaluating dispensing quality is provided. FIG1 is a flow chart of a method for evaluating dispensing quality according to an embodiment of the present disclosure. As shown in FIG1 , the flow chart includes the following steps:

[0031] Step S101, acquiring an image of the dispensing area;

[0032] In an exemplary embodiment, an image of the workstation to be inspected is acquired using a three-dimensional structured light camera, for example, using a DMD (digital micromirror chip) module to project coded structured light and two high-resolution industrial cameras to capture image data.

[0033] Step S102, acquiring point cloud data based on the collected image;

[0034] In an exemplary embodiment, a multi-step phase method and parallax calculation are used to obtain point cloud data from each camera perspective.

[0035] Multi-step phase shift is a method for measuring optical wavefronts. Based on the principle of interference, it measures the shape and phase information of the optical wavefront by introducing multiple phase steps in the optical path.

[0036] In the multi-step phase method, the optical wavefront is first split into two optical paths using a beam splitter. Then, in each optical path, different phase steps are introduced, such as with a plane wave plate, liquid crystal, or electrodynamic modulator, causing the phase of the light in the two paths to change differently. The light from the two optical paths is then combined again, generating an interference pattern through the interference effect. By analyzing the brightness distribution and phase differences in the interference pattern, the shape and phase information of the optical wavefront can be inferred.

[0037] The multi-step phase method provides highly accurate optical wavefront measurements and is widely used in optical system design, optical component manufacturing, and optical surface topography measurement. Its advantages include high measurement accuracy, a wide measurement range, and compatibility with light sources of different wavelengths.

[0038] In an exemplary embodiment, the following method is used to calculate the point cloud data from two camera perspectives using parallax:

[0039] Capture images from two camera perspectives: Use two cameras to capture the same scene simultaneously, maintaining a horizontal distance between the cameras to ensure parallax accuracy.

[0040] Extract feature points: For each image, use a feature point detection algorithm (such as SIFT, SURF, or ORB) to extract feature points in the image. These feature points should have properties that can be matched in both views.

[0041] Feature point matching: By matching feature points from two perspectives, we can find pairs of points with the same features in the two images. This can be achieved using feature descriptors (such as SIFT descriptors) and feature matching algorithms (such as the nearest neighbor algorithm).

[0042] Calculate disparity: For each pair of matched feature points, calculate the difference in their horizontal positions on the image, i.e., disparity. Disparity can be obtained by calculating the difference in the x-axis coordinates of the feature points.

[0043] Triangulation: Using parallax, the 3D coordinates of each matching point can be calculated through triangulation. Triangulation is the process of calculating the 3D position of a point using the camera parameters and parallax from two viewpoints. Based on the parallax and camera parameters, the 3D coordinates of each matching point can be calculated using the principles of triangulation.

[0044] Generate point cloud data: Convert the 3D coordinates of each matching point into point cloud data. Point cloud data is a collection of points, each of which contains its 3D coordinates and possible other attributes (such as color, normal, etc.).

[0045] Through the above method, parallax can be used to calculate point cloud data from the two camera perspectives. This point cloud data can be used for 3D reconstruction, depth estimation, and other related applications.

[0046] Step S103, acquiring colloid feature data based on the point cloud data;

[0047] Furthermore, the colloid characteristic data includes: total colloid volume, colloid area, colloid position and colloid height of each point, wherein the total colloid volume, colloid area and colloid position are the overall data of all single points in the dispensing area.

[0048] Step S104, calculating the thickness of the colloid after the colloid is flattened based on the colloid characteristic data;

[0049] Step S105 : evaluating the dispensing quality based on the colloid characteristic data and the colloid thickness.

[0050] By integrating point cloud data and colloid feature data and calculating colloid thickness, we address the difficulty in accurately determining dispensing volume in traditional 2D camera-based inspection solutions. Comprehensively evaluating dispensing quality improves the accuracy of dispensing volume, thereby enhancing the quality and efficiency of heat dissipation silicone dispensing.

[0051] In one embodiment, calculating the thickness of the colloid after being flattened based on the colloid characteristic data includes: calculating the thickness of the colloid after being flattened based on the total volume of the colloid and the area of ​​the dispensing region.

[0052] In an exemplary embodiment, the area of ​​the dispensing region is known before processing. The calculated total volume of the colloid is divided by the area of ​​the dispensing region to calculate the thickness of the colloid after it is flattened. Comparing the thickness of the colloid after it is flattened with a standard threshold can be used as a reference factor for evaluating dispensing quality.

[0053] In an exemplary embodiment, before preparing to flatten the colloid, a mold is generally placed in the glue dot area (for example, it can be a square frame to surround the glue dot area, and the inner wall of the square frame is used to prevent the colloid from overflowing the glue dot area during the process of flattening the colloid). The total volume of the colloid is divided by the area of ​​the glue dot area (or the area surrounded by the square frame) to obtain the thickness of the colloid after being flattened.

[0054] In an exemplary embodiment, before acquiring point cloud data based on the captured image, the method further includes: calibrating camera intrinsic parameters and system parameters with the assistance of a calibration board.

[0055] Specifically, the following methods can be used to assist in calibrating camera intrinsic parameters and system parameters:

[0056] Prepare the calibration target: Choose a calibration target with a specific geometry and known dimensions, such as a checkerboard or dot target. Ensure that the target is flat, undamaged, and clearly visible in the camera image.

[0057] Camera Setup: Place the camera in a fixed position and ensure there are no obstructions between it and the calibration plate. Adjust the camera's focal length, exposure time, and other parameters to ensure a clear image.

[0058] Capture images: Use the camera to capture a series of images containing the calibration plate. Capture multiple images at different angles, distances, and poses, ensuring that the calibration plate occupies different positions in the images.

[0059] Extract corner points: Use image processing software or computer vision libraries to extract corner points from each image. Corner points are the intersection points of specific geometric shapes on the calibration plate and can be extracted by detecting lines and edges in the image.

[0060] Calibrate camera intrinsic parameters: Using the extracted corner coordinates, apply a camera calibration algorithm to calculate the camera's intrinsic parameters, including focal length, principal point coordinates, and distortion parameters. Commonly used camera calibration algorithms include the Zhang Zhengyou calibration method and the Tsai algorithm.

[0061] Calibrate system parameters: Based on the known dimensions of the calibration plate and the camera intrinsic parameters, use the calibration plate's geometric relationship and the camera model to derive system parameters, such as the distance from the camera to the calibration plate and the camera pose.

[0062] Verification and optimization: Use the calibration results to perform distortion correction and pose estimation on new images to verify the accuracy of the calibration. If the calibration results do not meet the requirements, you can optimize them by increasing the number of images taken, changing the position and pose of the calibration plate, etc.

[0063] Therefore, through the above steps, the calibration board can be used to assist in the calibration of camera intrinsic parameters and system parameters, thereby improving the measurement and analysis accuracy of camera images.

[0064] In one embodiment, obtaining colloid feature data based on the point cloud data, as shown in FIG2 , specifically includes:

[0065] Step S201, converting point cloud data into a depth map to obtain depth information;

[0066] Point cloud data is three-dimensional data consisting of a series of point coordinates, each representing a specific location on the colloid surface. A depth map converts point cloud data into a two-dimensional image, where each pixel represents the depth information of the corresponding point. By converting point cloud data into a depth map, it is easier to obtain the depth information of each point.

[0067] Step S202 : obtaining the colloid height of each point based on the depth information.

[0068] Using the depth information in the depth map, the height of the colloid at each point can be calculated. Depth information represents the distance of each point in the point cloud data relative to the camera. By calculating the depth value of each point, the height of the colloid at that point can be determined. The height of the colloid refers to the distance from the surface to the bottom of the colloid. By obtaining the height of the colloid at each point, the three-dimensional shape and thickness of the colloid can be determined.

[0069] Through the above steps, the point cloud data can be converted into a depth map, and the height of the colloid at each point can be obtained based on the depth information. This can provide more accurate information on the colloid morphology and thickness, providing a stronger basis for subsequent dispensing quality assessment.

[0070] In one embodiment, after converting the point cloud data into a depth map and obtaining depth information, as shown in FIG3 , the following steps are further included:

[0071] Step S301, converting the depth map into a grayscale map;

[0072] A depth map is an image that contains depth information for each pixel, while a grayscale image is an image that converts the depth information in the depth map into grayscale values. After conversion to grayscale, the grayscale value of each pixel represents the depth information at that point. Converting a depth map to grayscale makes it easier to process and analyze depth information.

[0073] Step S302 : Separate the colloid region from the standard plane based on the grayscale image to determine the colloid region.

[0074] By applying threshold segmentation or other image processing algorithms to the grayscale image, the colloid region can be separated from the background. The colloid region represents the location and morphology of the colloid in the depth map, while the standard plane is a pre-determined reference plane used to determine the height and thickness of the colloid. By separating the colloid region from the standard plane, an accurate colloid region can be obtained, providing accurate data for subsequent colloid quality assessment.

[0075] Through the above steps, the depth map can be converted into a grayscale image, and the colloid area can be separated from the standard plane based on the grayscale image to determine the colloid area. In this way, an accurate colloid area can be obtained, providing a more powerful basis for subsequent dispensing quality evaluation.

[0076] In one embodiment, after separating the colloid region from the standard plane based on the grayscale image to determine the colloid region, the method further includes:

[0077] Use multiple standard squares of preset sizes to fill the colloid area to obtain multiple segmented regions.

[0078] A standard square is a square of known dimensions, such as 1 square millimeter, that can be used as a reference. By filling the colloid region with multiple standard squares, it is possible to cover part or all of the colloid. This allows the colloid image to be segmented into multiple segments, allowing the data corresponding to the entire colloid image to be divided, revealing more detail for subsequent calculations and analysis, thereby improving accuracy.

[0079] In one embodiment, after the colloid region is filled with a plurality of standard squares of a preset size to obtain a plurality of segmented regions, as shown in FIG4 , the method further includes:

[0080] Step S401, calculating the average height of the colloid in the segmented area based on the colloid height of each point in the segmented area;

[0081] The segmented area may include multiple adjacent point cloud data. The height average is obtained by averaging the colloid heights corresponding to the multiple point cloud data, and the height average is used as the height of the colloid in the segmented area.

[0082] Step S402, obtaining the volume of the colloid within the segmented area based on the area and height of the standard square;

[0083] By multiplying the area of ​​the standard square and the average height in the corresponding segmented area, the colloid volume in the corresponding colloid area can be obtained.

[0084] Step S403: obtaining the total volume of the colloid based on the volume of the colloid.

[0085] The total volume of the colloids can be obtained by adding up the multiple colloid volumes obtained in step S402.

[0086] The colloid volume in each segmented area is first calculated, and then the total colloid volume is obtained by accumulating multiple colloids, which improves the accuracy of the measurement results.

[0087] In one embodiment, after calculating the colloid area based on the standard area, the method may further include:

[0088] Calculate the height mean based on the colloid height at each point;

[0089] For each point within the colloid region, the corresponding colloid height is obtained based on its depth or grayscale value. The mean height is obtained by summing the colloid heights of all points and dividing by the total number of points. The mean height represents the average height level of the colloid and can be used to describe the thickness or height distribution of the colloid.

[0090] The total volume of the colloid was obtained based on the area and height of the colloid.

[0091] The colloid area was calculated in the previous step, and the mean height was calculated in the same way. The total volume of the colloid can be calculated by multiplying the colloid area by the mean height. This is because the total volume of the colloid can be considered as the vertical extension of the colloid area, that is, the colloid area multiplied by the mean height of the colloid.

[0092] The above method can be used to calculate the average height of the colloid at each point, and the total volume of the colloid can be obtained based on the colloid area and the average height. This allows for convenient and rapid colloid volume information, providing accurate data for subsequent colloid quality assessment and improving the time efficiency of testing.

[0093] In an exemplary embodiment, the main factors affecting the thickness of the colloid after being flattened include the magnitude of the applied pressure, the properties of the colloid, the structure of the colloid, and other factors. However, during one or more processing steps, the magnitude of the applied pressure is fixed, the properties and structure of the colloid used are also fixed, and the location or area of ​​the dispensing is fixed. Therefore, through the method disclosed herein, under the premise that the magnitude of the applied pressure, the properties of the colloid, the structure of the colloid, the location or area of ​​the dispensing are fixed, after measuring the area and volume of the colloid before the colloid is flattened, the thickness of the colloid after the colloid is flattened can be calculated.

[0094] Furthermore, during the actual processing process, the location or area of ​​the glue dispensed may deviate. However, before applying pressure to the colloid, factors such as the magnitude of the applied pressure, the properties of the colloid, and the colloid structure are determined. Through the method disclosed herein, after first measuring the correct position of the colloid, the thickness of the colloid after the colloid is flattened can be calculated by combining the colloid volume and colloid area. Specifically, when the colloid volume and colloid area are known, through multiple experiments, by varying the magnitude of the pressure, the properties of the colloid, the colloid structure, and other factors, the corresponding colloid thickness after the colloid is flattened can be obtained. The specific experimental method will not be repeated here.

[0095] In one embodiment, after converting the depth map into a grayscale map, the method further includes:

[0096] The brightness difference and details of the colloidal area are enhanced by grayscale expansion method.

[0097] Through the above method, the following effects can be achieved:

[0098] 1. Enhance the light and dark contrast of the colloid area: Grayscale expansion can make the bright part of the colloid area brighter and the dark part darker, thereby enhancing the light and dark contrast and making the details of the colloid area more prominent.

[0099] 2. Enhance the details and texture of the colloid area: Grayscale expansion can make the details of the colloid area clearer and richer, and make its texture more obvious, thereby improving the detail expression ability of the colloid area.

[0100] 3. Improve the visual effect of the colloid area: By enhancing the brightness difference and details of the colloid area, grayscale expansion can make the colloid area fuller, three-dimensional and layered, thereby improving the visual effect and appreciation of the colloid area.

[0101] In summary, increasing the brightness difference and details of the colloidal area through the grayscale expansion method can make the colloidal area more prominent, clearer and richer, thereby improving the quality and visual effect of the image.

[0102] In one embodiment, after improving the brightness difference and details of the colloidal region by the grayscale dilation method, the method further includes:

[0103] The multiple discrete colloidal regions generated by the grayscale expansion method are merged into a continuous colloidal region through the region growing method.

[0104] Through the above method, the following effects can be achieved:

[0105] 1. Continuity: The merged colloidal region will no longer be discrete, but a continuous region. This can make the colloidal region more natural and realistic, and more in line with the actual situation.

[0106] 2. Smoothness: The region growing method can smooth multiple discrete colloidal regions through certain rules and algorithms. This can reduce the jaggedness or discontinuity of the colloidal region boundaries and make the merged region smoother.

[0107] 3. Accuracy: The region growing method can merge similar colloidal regions based on the characteristics and properties of the image. This can improve the precision and accuracy of the merged colloidal regions, making the results more reliable.

[0108] 4. Efficiency: The region growing method can merge multiple discrete colloidal regions into a continuous region, thereby reducing the number of objects to be processed. This can improve the efficiency and speed of the algorithm and speed up the processing process.

[0109] In general, merging multiple discrete colloidal regions generated by the grayscale expansion method into a continuous colloidal region through the region growing method can achieve a more natural, smooth, precise and efficient effect.

[0110] In one embodiment, after merging multiple discrete colloidal regions generated by the grayscale expansion method into a continuous colloidal region using a region growing method, the method further includes:

[0111] The edge outline of the continuous colloidal area was confirmed by the window sliding method.

[0112] Through the above method, the following effects can be achieved:

[0113] 1. Determine the edge contours of the colloidal region: The window sliding method slides a fixed-size window across the image and calculates the average value or other characteristic value of the pixels within the window to determine the edge contours of the colloidal region. This can be used to segment colloidal particles in the image and obtain their shape and position information.

[0114] 2. Analyze the size and shape of colloidal particles: After determining the edge contours of a continuous colloidal region using the window sliding method, the size and shape of colloidal particles can be calculated and analyzed. For example, parameters such as particle diameter, area, and perimeter can be calculated, allowing for the study of particle properties and behavior.

[0115] 3. Image processing and computer vision applications: The window sliding method can be used as a technical means of image processing and computer vision to segment and extract target areas of interest.

[0116] In summary, confirming the edge contours of continuous colloidal regions through the window sliding method can realize the segmentation and shape analysis of colloidal particles, as well as its applications in image processing and computer vision.

[0117] In one embodiment, after confirming the edge contour of the continuous colloid region by the window sliding method, the method further comprises:

[0118] When it is confirmed that the range of the colloid area exceeds the preset range of the dispensing area, it is confirmed that glue connection will occur.

[0119] Through the above method, the following effects can be achieved:

[0120] 1. Enhance the bonding strength between colloidal particles: Since there are connections between colloidal particles, their bonding strength will be enhanced.

[0121] 2. Improve the mechanical properties of materials: The connected colloidal particles form a continuous structure, which can increase the strength and stiffness of the material.

[0122] 3. Formation of specific structures and morphologies: The connected colloidal particles can form specific structures and morphologies, such as network, spindle or chain structures.

[0123] 4. Influence on fluid properties and flow behavior: The structure and morphology formed between the colloidal particles can affect the properties and flow behavior of the fluid.

[0124] In short, the bonding strength between colloidal particles can be enhanced, the mechanical properties of the material can be improved, a specific structure and morphology can be formed, and the fluid properties and flow behavior can be affected.

[0125] In one embodiment, the dispensing quality is evaluated based on the colloid characteristic data and the colloid thickness, including:

[0126] The dispensing quality is evaluated by comparing the fluid characteristic data and fluid thickness with the preset thresholds in the workpiece drawing.

[0127] Through the above method, the following effects can be achieved:

[0128] 1. Dispensing Quality Assessment: Dispensing quality can be assessed by comparing the colloid characteristic data with preset thresholds. If the colloid characteristic data and colloid thickness are within the preset threshold range, the dispensing quality is considered good. Conversely, if the colloid characteristic data or colloid thickness exceeds the preset threshold, it may indicate a dispensing problem, requiring correction or re-dispensing.

[0129] 2. Determine whether the dispensing result meets the requirements: Comparing the colloid characteristic data with the preset threshold value can determine whether the dispensing result meets the requirements. Based on the preset threshold value in the workpiece drawing, it can be determined whether the size, shape, or position of the colloid is within a reasonable range. If it does, the dispensing result meets the design requirements. If it does not meet the requirements, the dispensing process may need to be adjusted or the dispensing process may need to be repeated.

[0130] 3. Automated quality control: Automated quality control is achieved by comparing colloid characteristic data and colloid thickness with preset thresholds. By comparing these data with the thresholds, the dispensing quality can be automatically determined, reducing the subjectivity and errors of manual judgment. This can improve production efficiency and consistency and reduce the risk of quality issues.

[0131] In summary, comparing colloid feature data and colloid thickness with preset thresholds in the workpiece drawing allows for dispensing quality assessment, determining whether dispensing results meet requirements, and automated quality control. These effects can improve the reliability and consistency of dispensing quality and reduce the generation of defective products.

[0132] In one embodiment, after acquiring an image of the dispensing area by a binocular structured light camera, the method further includes:

[0133] The point cloud data is preprocessed to remove noisy and redundant data.

[0134] Through the above method, the following effects can be achieved:

[0135] 1. Enhance point cloud data quality: Point cloud data quality can be improved by eliminating noisy data. Noisy data can be caused by factors such as sensor noise, environmental interference, or data acquisition errors. Preprocessing can reduce or completely eliminate this noisy data through filtering, smoothing, or outlier removal, thereby improving the accuracy and reliability of point cloud data.

[0136] 2. Improving point cloud data stability: Preprocessing can improve point cloud data stability by removing redundant data. Excessive data can be caused by sensor oversampling or scanning at different angles. By filtering, sampling, or resampling, redundant information in point cloud data can be reduced, thereby improving data consistency and stability.

[0137] 3. Reduce computational and storage overhead: Eliminating noisy and redundant data can reduce computational and storage overhead for point cloud data. Noise and redundant data increase the volume and complexity of point cloud data, increasing computational and storage requirements. Preprocessing can reduce data volume and simplify the data structure, thereby reducing computational and storage overhead and improving processing efficiency.

[0138] 4. Improve the effectiveness of subsequent data processing and analysis: Preprocessing can provide a better foundation for subsequent data processing and analysis. Eliminating noise and redundant data can reduce interference with subsequent algorithms and methods, providing more accurate and reliable data, thereby improving the effectiveness of operations such as point cloud data segmentation, registration, and recognition.

[0139] In summary, preprocessing point cloud data to eliminate noisy and redundant data can enhance data quality, improve data stability, reduce computing and storage overhead, and improve the results of subsequent data processing and analysis.

[0140] The following examples illustrate the above content:

[0141] Grayscale dilation compares the grayscale value of each pixel with the structural element and sets the grayscale value of the pixel to the maximum grayscale value in its neighborhood. This operation can make the darker areas "expand" and gradually brighten, thereby enhancing the brightness differences and details in the image.

[0142] The steps of the grayscale expansion method are as follows:

[0143] (1) Define a structural element that has a specific shape and size.

[0144] (2) Overlap the structural element with each pixel of the image.

[0145] (3) Compare the grayscale value of the structural element with that of each pixel and find the maximum grayscale value in its neighborhood.

[0146] (4) Set the grayscale value of the current pixel to the maximum grayscale value in the neighborhood.

[0147] (5) Repeat steps (2) to (4) until the entire image is traversed.

[0148] The basic idea of ​​the region growing method is to start from a seed point and gradually add adjacent pixels to an initial region until a stopping condition is met. These adjacent pixels are selected based on some predefined pixel similarity criteria. For example, pixel grayscale value, color, texture, or other characteristics can be used to determine similarity.

[0149] The steps of the region growing method are as follows:

[0150] (1) Select a seed point (starting point), which is usually selected by the user through interaction.

[0151] (2) Initialize an initial region and add seed points to the region.

[0152] (3) According to the similarity criterion, a neighborhood pixel is selected to be compared with the pixels in the current area.

[0153] (4) If the neighboring pixel is similar to the pixel in the current region, the pixel is added to the current region and marked as visited.

[0154] (5) Repeat steps (3) to (4) until no more similar pixels can be added or the stopping condition is reached.

[0155] In steps (3), (4), and (5), the basic idea of ​​volume calculation is to segment the image using the smallest standard unit. The height value is fitted to the region with the mean value as the unit height. All unit volumes are accumulated and added together to finally obtain the volume of each region. The window sliding method uses a fixed-size window to slide on the image and intercept the image blocks within the window. You can choose to slide the window with a fixed step size or overlap to cover the entire image. In this way, the image can be segmented into multiple fixed-size image blocks. The steps are as follows:

[0156] (1) Determine the size of the window: Set the unit to the minimum window unit of 1 square millimeter.

[0157] (2) Initialize the window position: Set the initial position of the window to the upper left corner of the colloid projection plane image.

[0158] (3) Capture the image block within the window: Based on the position of the window, capture the image block at the corresponding position from the original image.

[0159] (4) Move the window position: Move the window position according to the set step size or overlap.

[0160] (5) Repeat steps (3) to (5): Repeat steps (3) to (5) until the window slides to all parts of the image.

[0161] Therefore, by using the above-mentioned window sliding method, not only the edge area of ​​the colloid can be obtained, but also the colloid area can be obtained according to the number of window sliding times and the surface area of ​​the window.

[0162] The colloid volume calculated in the above way and the area of ​​the required silicone coating obtained from the workpiece drawing can be used to calculate the theoretical thickness of the area, thereby completing the evaluation of the colloid thickness in a quantitative way to avoid overflow / shortage of glue (i.e. too much / too little colloid volume) in subsequent production links.

[0163] In addition, for the glue ligation problem generated during the glue dispensing process described above, the same method can also be used, that is, grayscale expansion of the depth map and other methods to extract the depth part higher than the standard plane. If the obtained range exceeds the threshold of the framed ROI area, it can be determined that there is glue ligation in this part of the glue dispensing.

[0164] In actual production, some workpiece surfaces have inclined surfaces and other areas that require silicone dispensing. For such areas, 2D cameras, lacking angle and depth information, can lead to inaccurate area estimation. However, a dispensing measurement solution using a 3D structured light camera, after 3D reconstruction, can effectively extract the colloid from the inclined surface. Similarly, by projecting or placing the inclined surface on a standard plane, the colloid volume calculation and measurement method can be used to accurately calculate the colloid volume of the inclined surface.

[0165] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by adding the necessary general hardware platform with the help of software, of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD-ROM), including a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present disclosure.

[0166] In this embodiment, a dispensing quality assessment device is also provided, which is configured to implement the above-mentioned embodiments and preferred embodiments, and will not be repeated hereafter. As used below, the term "module" may be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.

[0167] FIG5 is a structural block diagram of a dispensing quality evaluation device according to an embodiment of the present disclosure. As shown in FIG5 , the device includes:

[0168] An acquisition module 51 is configured to acquire an image of the dispensing area through a binocular structured light camera;

[0169] A processing module 52 is configured to obtain point cloud data based on the collected image; and

[0170] Acquiring colloid feature data based on point cloud data; and,

[0171] Calculating the thickness of the colloid after the colloid is flattened based on the colloid characteristic data; and,

[0172] Evaluate dispensing quality based on colloid characteristic data and colloid thickness.

[0173] Furthermore, the device also includes an output module 53 configured to output the evaluation result.

[0174] Furthermore, the colloid characteristic data includes: total volume of the colloid, area of ​​the colloid, position of the colloid and height of the colloid at each point, wherein the colloid volume, area of ​​the colloid and position of the colloid are the overall data of all individual points in the dispensing area.

[0175] Furthermore, the device is further configured to calculate the thickness of the colloid after being flattened based on the colloid characteristic data, including calculating the thickness of the colloid after being flattened based on the total volume of the colloid and the area of ​​the dispensing region.

[0176] In one exemplary embodiment, the area of ​​the dispensing region is known before processing. The device then divides the calculated total volume of the colloid by the area of ​​the dispensing region to calculate the thickness of the colloid after it is flattened. The thickness of the colloid after it is flattened is compared with a standard threshold value, which can be used as a reference factor for evaluating dispensing quality.

[0177] In an exemplary embodiment, before preparing to flatten the colloid, a mold is generally placed in the dotted area (for example, a square frame is used to surround the dotted area, and the inner wall of the square frame is used to prevent the colloid from overflowing the dotted area during the flattening process). The device uses the total volume of the colloid divided by the area of ​​the dotted area (or the area surrounded by the square frame) to obtain the thickness of the colloid after the colloid is flattened.

[0178] Furthermore, the device is further configured to: convert the point cloud data into a depth map to obtain depth information;

[0179] Get the colloid height at each point based on the depth information.

[0180] Furthermore, the device is further configured to: convert the point cloud data into a depth map, and after obtaining the depth information, further include:

[0181] Convert the depth map into a grayscale image;

[0182] The colloid region was separated from the standard plane based on the grayscale image to determine the colloid area.

[0183] Furthermore, the device is further configured to: separate the colloid area from the standard plane based on the grayscale image to determine the colloid area, and further include: filling the colloid area with multiple standard squares of preset sizes to obtain multiple segmented areas.

[0184] Furthermore, the device is further configured to: fill the colloid region with a plurality of standard squares of a preset size to obtain a plurality of segmented regions, and further comprises:

[0185] Calculate the average height of the colloid in the segmented area based on the colloid height of each point in the segmented area;

[0186] The colloid volume within the segmented region is obtained based on the area and height of the standard square;

[0187] The total volume of the colloid is obtained based on the volume of the colloid. Further, the device can be configured to: after calculating the colloid area based on the standard area, it can also include:

[0188] Calculate the height mean based on the colloid height at each point;

[0189] The total volume of the colloid was obtained based on the area and height of the colloid.

[0190] Furthermore, the device is further configured to: after converting the depth map into a grayscale map, further include:

[0191] The brightness difference and details of the colloidal area are enhanced by grayscale expansion method.

[0192] Furthermore, the device is further configured to: after improving the brightness difference and details of the colloidal area by the grayscale expansion method, further include:

[0193] The multiple discrete colloidal regions generated by the grayscale expansion method are merged into a continuous colloidal region through the region growing method.

[0194] Furthermore, the device is further configured to: after merging multiple discrete colloidal regions generated by the grayscale expansion method into a continuous colloidal region by a region growing method, further comprising:

[0195] The edge outline of the continuous colloidal area was confirmed by the window sliding method.

[0196] Furthermore, the device is further configured to: after confirming the edge contour of the continuous colloid area by the window sliding method, further include:

[0197] When it is confirmed that the range of the colloid area exceeds the preset range of the dispensing area, it is confirmed that glue connection will occur.

[0198] Furthermore, the device is further configured to: evaluate the dispensing quality based on the colloid characteristic data and the colloid thickness, including:

[0199] The dispensing quality is evaluated by comparing the fluid characteristic data and fluid thickness with the preset thresholds in the workpiece drawing.

[0200] Furthermore, the device is further configured to: after acquiring the collected image of the dispensing area through the binocular structured light camera, further include:

[0201] The point cloud data is preprocessed to remove noisy and redundant data.

[0202] It should be noted that the above modules can be implemented through software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0203] An embodiment of the present disclosure further provides a computer-readable storage medium, in which a computer program is stored. The computer program is configured to execute the steps of any one of the above method embodiments when run.

[0204] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0205] An embodiment of the present disclosure further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0206] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0207] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail here.

[0208] Obviously, those skilled in the art should understand that the modules or steps of the present disclosure described above can be implemented using a general-purpose computing device, they can be concentrated on a single computing device, or distributed across a network composed of multiple computing devices, they can be implemented using program code executable by the computing device, and thus, they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be performed in a different order than herein, or they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. Thus, the present disclosure is not limited to any particular combination of hardware and software.

[0209] The foregoing description is merely a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. Those skilled in the art will readily appreciate that various modifications and variations of the present disclosure are possible. Any modifications, equivalent substitutions, or improvements made within the principles of the present disclosure shall be included within the scope of protection of the present disclosure.

Claims

1. A method for evaluating dispensing quality, comprising: Acquire a collection image of the dispensing area; Acquire point cloud data based on the collected image; Acquire colloid characteristic data based on the point cloud data; Calculating the thickness of the colloid after the colloid is flattened based on the colloid characteristic data; The dispensing quality is evaluated based on the colloid characteristic data and the colloid thickness.

2. The method according to claim 1, wherein: Acquiring colloid characteristic data based on the point cloud data includes: Convert the point cloud data into a depth map to obtain depth information; The height of the colloid at each point is obtained based on the depth information.

3. The method according to claim 2, wherein: After converting the point cloud data into a depth map and obtaining the depth information, the method further includes: Converting the depth map into a grayscale image; The colloid region is separated from a standard plane based on the grayscale image to determine the colloid region.

4. The method according to claim 3, wherein: After separating the colloid region from the standard plane based on the grayscale image to determine the colloid region, the method further includes: The colloid region is filled with a plurality of standard squares of a preset size to obtain a plurality of segmented regions.

5. The method according to claim 4, wherein: After filling the colloid region with a plurality of standard squares of a preset size to obtain a plurality of segmented regions, the method further includes: Calculating the average height of the colloid in the segmented area based on the colloid height of each point in the segmented area; Obtaining the volume of the colloid in the segmented region based on the area of ​​the standard square and the height; The total volume of the colloid was obtained based on the colloid volume.

6. The method according to claim 3, wherein: After converting the depth map into a grayscale map, the method further includes: The brightness difference and details of the colloidal area are enhanced by grayscale dilation method.

7. The method according to claim 6, wherein: After improving the brightness difference and details of the colloidal area by the grayscale expansion method, it also includes: The plurality of discrete colloidal regions generated by the grayscale expansion method are merged into a continuous colloidal region by a region growing method.

8. The method according to claim 7, wherein: After merging the plurality of discrete colloid regions generated by the grayscale expansion method into a continuous colloid region by a region growing method, the method further includes: The edge contour of the continuous colloid region was confirmed by the window sliding method.

9. The method according to claim 8, wherein: After confirming the edge contour of the continuous colloid region by the window sliding method, the method further includes: When it is confirmed that the range of the colloid area exceeds the preset range of the glue dispensing area, it is confirmed that glue connection will occur.

10. The method according to claim 5, wherein: Calculating the thickness of the colloid after the colloid is flattened based on the colloid characteristic data, including: The thickness of the colloid after the colloid is flattened is calculated based on the total volume of the colloid and the area of ​​the glue-dispensing region.

11. The method according to claim 1, wherein: Based on the colloid characteristic data and the colloid thickness, the dispensing quality is evaluated, including: The colloid characteristic data, the colloid thickness and the preset threshold value in the workpiece drawing are compared to evaluate the dispensing quality.

12. The method according to claim 1, wherein: After acquiring the acquisition image of the dispensing area, it also includes: The point cloud data is preprocessed to eliminate noise data and redundant data.

13. A dispensing quality assessment device, comprising: An acquisition module, configured to acquire a collection image of a dispensing area; A processing module, configured to obtain point cloud data based on the collected image; as well as, Acquire colloid characteristic data based on the point cloud data; as well as, Calculating the thickness of the colloid after the colloid is flattened based on the colloid characteristic data; and, The dispensing quality is evaluated based on the colloid characteristic data and the colloid thickness.

14. A computer-readable storage medium having a computer program stored therein, wherein: When the computer program is executed by a processor, the steps of the method described in any one of claims 1 to 12 are implemented.

15. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in any one of claims 1 to 12 when executing the computer program.

Images