A method for measuring the height and diameter of circular embossing based on 3D vision

Abstract

This invention belongs to the field of electrode embossing measurement technology, specifically relating to a method for measuring the height and diameter of circular embossing based on three-dimensional vision. This invention employs direct measurement using three-dimensional point clouds, avoiding perspective and lighting errors inherent in traditional measurements. Utilizing multi-point averaging or sorting filtering improves noise resistance. From the acquisition of the material's three-dimensional surface contour data to the final measurement, the entire process is automatically executed based on algorithms, eliminating the need for manual calibration or positioning. This significantly improves the speed and consistency of production line inspection, adapting to different materials, different mold spacings, and minor surface defects. It avoids embossing deformation caused by uneven force during rolling, improving measurement accuracy. The parameterized segmentation strategy and compensation coefficient design allow for flexible application to various specifications of circular embossing molds.

Description

Technical Field

[0001] This invention belongs to the field of electrode embossing measurement technology, specifically relating to a method for measuring the height and diameter of circular embossing based on three-dimensional vision. Background Technology

[0002] In industrial manufacturing fields such as calendering, printing, and electronic thin films, circular embossing, as an important surface feature, not only serves a decorative purpose but also has a crucial impact on product performance and assembly precision. For example, in the production of electrode materials, coated substrates, and precision molds, the height and diameter of the embossing directly determine the product's compatibility and functional properties such as electrical and thermal conductivity. In the manufacturing process of lithium-ion batteries, the embossing process on the electrode surface can significantly improve the material's physical properties (such as porosity and electrolyte wettability), thereby increasing the battery's energy density and cycle life. To ensure consistent battery performance, the height and diameter of the embossing must be monitored in real time with high precision to strictly control dimensional tolerances during the production process and ensure the consistency of the electrode embossing dimensions.

[0003] Traditional embossing size inspection mainly relies on manual point measurements using contact measuring tools such as calipers or micrometers. This method is inefficient and easily affected by the operator's experience and perspective, making it difficult to meet the online inspection needs of large-scale production lines. Furthermore, while ordinary two-dimensional vision systems are also used for embossing inspection, they can only acquire planar projection information and cannot accurately depict changes in embossing height. They are also severely affected by factors such as lighting, reflection, and perspective distortion, resulting in significant measurement errors.

[0004] In recent years, 3D vision measurement technology has gradually emerged in the field of industrial inspection. Common methods include structured light, laser scanning, and stereo vision, which can acquire relatively complete 3D surface information in a single scan. Existing technologies often use local plane fitting or height map analysis to evaluate micro-features. However, for a typical regular texture like circular embossing, these methods cannot effectively achieve accurate and rapid center positioning, region segmentation, and elevation and diameter measurement in complex backgrounds and batch inspection scenarios. They require manual selection of key points in the embossed area, which is inefficient and highly subjective, making it difficult to meet the real-time requirements of production lines with inspection speeds of tens of meters per minute. Furthermore, existing methods directly calculate height based on the global coordinate system Z-axis, ignoring the systemic errors caused by irregular undulations on the electrode surface due to uneven force during the rolling process, resulting in local deformation. Summary of the Invention

[0005] The purpose of this invention is to provide a method for measuring the height and diameter of circular embossing based on three-dimensional vision. This method can simultaneously achieve high-precision, fully automated, and online detection of embossing height and diameter, thereby meeting the requirements of modern industrial production for high efficiency, stability, and scalability in measuring embossing geometric dimensions.

[0006] The specific technical solution adopted by this invention is as follows:

[0007] A method for measuring the height and diameter of circular embossing based on three-dimensional vision includes:

[0008] Obtain the three-dimensional point cloud data of the electrode embossing, obtain the two-dimensional depth image based on the three-dimensional point cloud data, and obtain the center position of the embossing.

[0009] The embossed flower's neighboring regions are obtained based on the center position of the embossed flower, and a segmentation strategy is obtained. The neighboring regions of the embossed flower are then segmented based on the segmentation strategy to obtain the segmented regions.

[0010] Obtain the reference plane based on segmented regions;

[0011] Obtain all three-dimensional points within the segmented area; obtain a set of height three-dimensional points that meet the first preset condition based on all three-dimensional points and the reference plane; obtain the embossing height based on the set of height three-dimensional points and the reference plane.

[0012] A local coordinate system is established based on the reference plane. A set of three-dimensional points for the diameter that meets the second preset condition is obtained based on the local coordinate system. The embossing diameter is obtained based on the set of three-dimensional points for the diameter and the local coordinate system.

[0013] In a preferred embodiment, the steps of acquiring three-dimensional point cloud data of the electrode embossing, acquiring a two-dimensional depth image based on the three-dimensional point cloud data, and obtaining the center position of the embossing include:

[0014] Obtain the 3D point cloud data of electrode embossing;

[0015] The 3D point cloud data is preprocessed to obtain preprocessed 3D point cloud data, wherein the preprocessing includes filtering;

[0016] The sampling parameters of the three-dimensional point cloud data of the collected electrode embossing are obtained. Based on the sampling parameters, the preprocessed three-dimensional point cloud data is projected into a two-dimensional depth image, where the pixel gray value corresponds to the Z value of the three-dimensional point.

[0017] The center position of the embossed flower is obtained based on a two-dimensional depth image.

[0018] In a preferred embodiment, the step of obtaining the embossing center position based on a two-dimensional depth image includes:

[0019] Obtain the roller die parameters for electrode embossing, and obtain the spacing between multiple adjacent embossing dies based on the roller die parameters;

[0020] Sort the spacing between multiple adjacent embossing molds in ascending order, and take the spacing between the first adjacent embossing mold as the filter mask size;

[0021] The two-dimensional depth image is filtered based on the filter mask size to obtain the filtered two-dimensional depth image.

[0022] Obtain the difference in grayscale value of each pixel before and after filtering;

[0023] Obtain the grayscale difference threshold, filter the pixels whose grayscale difference exceeds the grayscale difference threshold, and form a candidate feature pixel set;

[0024] Connected components are extracted from candidate feature pixel sets, and embossed connected components are obtained from the connected components.

[0025] Obtain the center pixel coordinates of the connected domain of the embossed flower, and map the center pixel coordinates to three-dimensional space to obtain the center position of the embossed flower.

[0026] In a preferred embodiment, the steps of obtaining the embossed neighboring region based on the embossing center position, obtaining a segmentation strategy, and segmenting the embossed neighboring region based on the segmentation strategy to obtain the segmented region include:

[0027] Get the preset adjacent edge dimensions;

[0028] Using the center position of the embossing as the reference point, extract the three-dimensional point cloud data within the preset adjacent edge size range as the embossing neighboring area;

[0029] Obtain the roller die parameters for electrode embossing, and obtain the segmentation strategy based on the roller die parameters;

[0030] The segmented regions are obtained by segmenting the adjacent areas of the embossed area based on the segmentation strategy. The segmented regions include planar areas and embossed areas.

[0031] In a preferred embodiment, the steps of obtaining the roller die parameters for electrode embossing and obtaining a segmentation strategy based on the roller die parameters include:

[0032] Obtain the roller mold parameters for electrode embossing, and based on the roller mold parameters, obtain the embossing mold diameter and the spacing between multiple embossing molds;

[0033] Sort the spacing between multiple adjacent embossing molds in ascending order, and take the spacing between the first adjacent embossing mold as the target embossing mold spacing.

[0034] Obtain the compensation coefficient;

[0035] The segmentation value is obtained based on the embossing die diameter, the target embossing die spacing, and the compensation coefficient;

[0036] Obtain the segmentation table, which includes multiple segmentation value ranges and the segmentation strategy corresponding to each segmentation value range;

[0037] The corresponding segmentation strategy is obtained from the segmentation table based on the segmentation value range corresponding to the segmentation value.

[0038] In a preferred embodiment, the step of obtaining the compensation coefficient includes:

[0039] The spacing fluctuation value is obtained based on the spacing of multiple embossing molds;

[0040] Obtain the fluctuation table, which includes multiple interval fluctuation value ranges and the corresponding compensation coefficient for each interval fluctuation value range;

[0041] The corresponding compensation coefficient is obtained from the fluctuation table based on the interval of the interval fluctuation value.

[0042] In a preferred embodiment, the step of obtaining the reference plane based on the segmented region includes:

[0043] Obtaining planar regions based on segmented regions;

[0044] Obtain all 3D points in the planar region and aggregate them into a planar 3D point set;

[0045] Equation of reference plane based on planar 3D point set;

[0046] The equation parameters are obtained based on the equation of the reference plane, and the plane characterized by the equation parameters is used as the reference plane.

[0047] In a preferred embodiment, the steps of acquiring all three-dimensional points within the segmented region, acquiring a set of height-based three-dimensional points that meet a first preset condition based on all three-dimensional points and a reference plane, and acquiring the embossing height based on the set of height-based three-dimensional points and the reference plane include:

[0048] Obtain all 3D points within the segmented region;

[0049] Obtain the distance from each 3D point to the reference plane and mark it as the first distance;

[0050] Sort the multiple first distances in descending order to obtain the first sorting table;

[0051] Obtain the preset distance quantity, select three-dimensional points that meet the preset distance quantity from the first sorting table in descending order, and summarize them into a height three-dimensional point set;

[0052] Obtain the average of the sums of all first distances in the height 3D point set, and use it as the embossing height.

[0053] In a preferred embodiment, the steps of establishing a local coordinate system based on a reference plane, obtaining a set of three-dimensional points representing the diameter that meet a second preset condition based on the local coordinate system, and obtaining the embossing diameter based on the set of three-dimensional points representing the diameter and the local coordinate system include:

[0054] Establish a local coordinate system based on the reference plane;

[0055] Transform the coordinates of the 3D point cloud data within the segmented region to the coordinates corresponding to the local coordinate system;

[0056] Extract the Z-axis coordinates of each 3D point in the local coordinate system;

[0057] Obtain a preset height threshold, select 3D points whose Z-axis coordinates are greater than the preset height threshold, and summarize them into a diameter 3D point set;

[0058] Obtain the bounding box of the 3D point set of diameter, and obtain the span difference of the bounding box in the Y-axis direction, and use it as the embossing diameter.

[0059] And, a terminal for measuring the height and diameter of circular embossing based on three-dimensional vision, comprising:

[0060] One or more processors;

[0061] A storage device on which one or more programs are stored;

[0062] When one or more programs are executed by one or more processors, the one or more processors implement a method for measuring the height and diameter of circular embossing based on three-dimensional vision.

[0063] The technical effects achieved by this invention are as follows:

[0064] This invention employs direct measurement using 3D point clouds, avoiding perspective and lighting errors inherent in traditional measurements. Utilizing multi-point averaging or sorting for better noise resistance, the entire process, from acquiring 3D surface contour data to final measurement, is automatically executed based on an algorithm, eliminating the need for manual calibration or positioning. This significantly improves production line inspection speed and consistency, adapting to different materials, mold spacings, and minor surface defects. It avoids embossing deformation caused by uneven stress during rolling, enhancing measurement accuracy. The parameterized segmentation strategy and compensation coefficient design allow for flexible application to various specifications of circular embossing molds, enabling rapid switching with only a few parameter adjustments. Combined with line lasers or high-speed cameras, the algorithm's processing speed meets production line cycle requirements, supporting online real-time monitoring and timely removal of defective products. Attached Figure Description

[0065] Figure 1 This is a flowchart of the method provided by the present invention. Detailed Implementation

[0066] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0067] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0068] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in a preferred embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that mutually excludes other embodiments.

[0069] Furthermore, the present invention will be described in detail with reference to the schematic diagrams. When describing the embodiments of the present invention in detail, the schematic diagrams are merely examples for ease of explanation and should not limit the scope of protection of the present invention.

[0070] Please see the appendix Figure 1 As shown, a method for measuring the height and diameter of circular embossing based on three-dimensional vision is provided, including:

[0071] S1. Obtain the three-dimensional point cloud data of the electrode embossing, obtain the two-dimensional depth image based on the three-dimensional point cloud data, and obtain the center position of the embossing.

[0072] S2. Obtain the embossed neighboring region based on the embossed center position, and obtain the segmentation strategy. Segment the embossed neighboring region based on the segmentation strategy to obtain the segmented region.

[0073] S3. Obtain the reference plane based on the segmented region;

[0074] S4. Obtain all three-dimensional points within the segmented area, obtain a set of height three-dimensional points that meet the first preset condition based on all three-dimensional points and the reference plane, and obtain the embossing height based on the set of height three-dimensional points and the reference plane.

[0075] S5. Establish a local coordinate system based on the reference plane, obtain a set of three-dimensional points of diameter that meet the second preset conditions based on the local coordinate system, and obtain the embossing diameter based on the set of three-dimensional points of diameter and the local coordinate system.

[0076] As described in steps S1 to S5 above, a camera or line laser sensor is used to collect three-dimensional point cloud data of the embossed area on the electrode surface through structured light, laser, or multi-view reconstruction. Based on the sampling parameters during sensor acquisition (such as resolution, pixel mapping relationship, and scanning step distance), the three-dimensional point cloud is projected onto a two-dimensional plane to generate a two-dimensional depth image. Through a preset projection filtering and connected component analysis algorithm, the two-dimensional pixel coordinates (or point cloud centroid) of the embossed pattern center are extracted and mapped back to three-dimensional space to obtain the center position of the embossing in the three-dimensional coordinate system. Using the embossed center as a reference, a preset range is selected. The point cloud serves as the neighboring region for embossing. Based on embossing mold parameters (such as mold diameter and adjacent spacing) and empirical compensation coefficients, a segmentation strategy (such as threshold, filter mask size, and segmentation entries) is dynamically generated to segment the neighboring region's point cloud, resulting in two point sets: a planar region and the embossing region. All points are extracted from the segmented planar region, and a best-fit plane equation is calculated using least squares or minimum median variance fitting methods. This best-fit plane serves as the measurement reference plane. The point cloud within the entire segmented region is then acquired, and the distance from each point to the reference plane is calculated. Points are sorted from largest to smallest distance, and several points meeting the first preset criteria are selected. Points with a distance greater than the first N or exceeding a certain threshold are used to form a height point set. The average distance from these points to the reference plane is used as the height value of the embossing. A local coordinate system is established on the reference plane (the Z-axis is aligned with the plane normal, and the X-axis is aligned with the original coordinate system). The point cloud of the segmented region is transformed to this local coordinate system. Points with a height higher than the second preset height threshold are selected to form a diameter point set. A directed bounding box (OBB) is calculated for this point set, and the maximum span of the bounding box in the Y direction (Ymax–Ymin) is used as the diameter value of the embossing. Direct measurement using 3D point cloud avoids the perspective and lighting issues of traditional measurement. To mitigate errors, multi-point averaging or sorting screening is used to improve measurement noise resistance. From the acquisition of three-dimensional surface contour data of the material to the final measurement, the entire process is automatically executed based on algorithms, eliminating the need for manual calibration or positioning. This significantly improves the inspection speed and consistency of the production line. It is adaptable to different materials, different mold spacings, and minor surface defects. The parameterized segmentation strategy and compensation coefficient design can be flexibly applied to various specifications of circular embossing molds. Only a few sets of parameters need to be adjusted for quick switching. Combined with line lasers or high-speed cameras, the algorithm processing speed can meet the cycle time requirements of the production line, supporting online real-time monitoring and timely rejection of unqualified products.

[0077] In a preferred embodiment, the steps of acquiring three-dimensional point cloud data of the electrode embossing, acquiring a two-dimensional depth image based on the three-dimensional point cloud data, and obtaining the embossing center position include:

[0078] S101. Obtain the three-dimensional point cloud data of the electrode embossing;

[0079] S102. Preprocess the 3D point cloud data to obtain preprocessed 3D point cloud data, wherein the preprocessing includes filtering;

[0080] S103. Obtain the sampling parameters of the three-dimensional point cloud data of the collected electrode embossing, and project the preprocessed three-dimensional point cloud data into a two-dimensional depth image based on the sampling parameters, wherein the pixel gray value corresponds to the Z value of the three-dimensional point.

[0081] S104. Obtain the center position of the embossing based on the two-dimensional depth image.

[0082] As described in steps S101 to S104 above, a high-precision 3D imaging sensor (such as a line laser or structured light) is used to scan the electrode surface. (The general scanning process is as follows: a single scan reconstructs a line contour of the target in 3D. The target is scanned step by step from position A to B. The 3D line contours scanned during this period are stitched together to form the overall 3D surface contour, i.e., 3D point cloud data. From the 3D surface contour, a 2D depth image is constructed. Here, the number of 3D reconstruction points per scan by the line laser camera is the number of columns in the image, the number of scans from A to B is the number of rows in the image, and the depth Z-value information of each point is the pixel value of each point in the image.) The sensor synchronously records the depth information corresponding to each pixel. Through camera calibration parameters and projection models, complete 3D point cloud data is calculated. The original point cloud is input into algorithms such as statistical filtering, radius filtering, or voxel mesh filtering to remove noise points and outliers, and smooth the point cloud distribution. This step removes false points caused by light spots, reflections, or environmental interference. The sensor samples are read. Parameters (such as scan line spacing, frame rate, and camera focal length) are used to project the preprocessed 3D point cloud onto a regular 2D pixel grid. The grayscale value of each pixel is directly mapped to the Z coordinate (depth) of the corresponding 3D point, forming a depth grayscale map. The depth grayscale map is then filtered by mean or median (the mask size can be determined by the minimum spacing of the embossing) to highlight the grayscale abrupt changes caused by concave and convex deformation. The grayscale difference before and after filtering is calculated, and areas with obvious elevation changes are selected by thresholding. Connected components are extracted and interference is filtered by area or shape. The center pixel coordinates of the embossing connected components are taken, and the projection relationship is combined to calculate back to the 3D space, which is the embossing center position. Image and depth information are obtained simultaneously through a single scan, eliminating the need for multiple focusing or mechanical repositioning, resulting in high measurement efficiency. The method of combining filtering and connected component analysis can quickly eliminate point cloud noise and image artifacts, and the positioning error can be controlled to the sub-millimeter level. The projection parameters and filter mask size can be dynamically adjusted according to different mold sizes and spacings, making it easy to switch between embossed workpieces of different specifications.

[0083] In a preferred embodiment, the step of obtaining the embossing center position based on a two-dimensional depth image includes:

[0084] S1041. Obtain the roller die parameters for electrode embossing, and obtain the spacing between multiple adjacent embossing dies based on the roller die parameters;

[0085] S1042. Sort the spacing between multiple adjacent embossing molds in ascending order, and obtain the spacing between the first adjacent embossing mold as the filter mask size.

[0086] S1043. Filter the two-dimensional depth image based on the filter mask size to obtain the filtered two-dimensional depth image.

[0087] S1044. Obtain the difference in grayscale value of each pixel before and after filtering;

[0088] S1045. Obtain the gray value difference threshold, filter the pixels whose gray value difference exceeds the gray value difference threshold, and form a candidate feature pixel set.

[0089] S1046. Extract connected components based on candidate feature pixel set, and obtain embossed connected components based on connected components;

[0090] S1047. Obtain the center pixel coordinates of the embossed flower connected domain and map the center pixel coordinates to three-dimensional space to obtain the center position of the embossed flower.

[0091] As described in steps S1041 to S1047 above, the diameter of the roller mold corresponding to the embossed workpiece and the arrangement parameters of adjacent embossing molds are read. The actual spacing between all adjacent embossing molds is calculated. Based on the known mold geometry, this step can obtain accurate spatial interval values ​​without additional measurement. Multiple spacing values ​​are sorted from smallest to largest, and the minimum value is selected as the mask size used for filtering. The minimum spacing usually corresponds to the width of the densest embossing pattern. Using it as the filter kernel size can just smooth the background between adjacent patterns without excessively blurring the embossing edges. By applying mean or median filtering to the two-dimensional depth grayscale image with a determined mask size, the grayscale abrupt changes at the peaks and valleys of the patterns are locally smoothed, removing minor noise and material imperfections while preserving the large-scale features. The embossing outline is determined by pixel-by-pixel subtraction between the original grayscale image and the filtered grayscale image. This difference map highlights areas with dramatic elevation (depth) changes, i.e., the locations of embossed protrusions or depressions. After setting a grayscale difference threshold, pixels with a difference greater than this threshold are selected to form a candidate feature pixel set. This process eliminates background noise and low-contrast areas, retaining only the high-contrast pixels of the embossed outline. Connectivity analysis is then performed on the candidate feature pixel set to identify individual region blocks. Based on prior information such as area and shape (e.g., circular or approximately circular), the true embossed region is selected from these connected regions. The centroid (center pixel coordinates) of the selected connected region is calculated, and the arithmetic mean of all its pixel coordinates is used as the center pixel coordinate. The calculation formula is as follows: In the formula, Let represent the coordinates of the center pixel, and ... The coordinates of the i-th pixel within the embossing connected region are represented by the projection parameters. These pixel coordinates are then used to calculate the 3D point cloud space, yielding accurate 3D coordinates of the embossing center. The filter mask size is automatically determined using the mold spacing, which smooths background noise while preserving the embossing boundary. This avoids the problem of low adaptability of fixed sizes to different embossing specifications. Based on dual screening using grayscale difference and connected region analysis, false features caused by material reflection, dust, or minor scratches are effectively eliminated, ensuring that the final positioning is based on the real embossing pattern.

[0092] In a preferred embodiment, the step of obtaining the embossed neighboring region based on the embossing center position, obtaining a segmentation strategy, and segmenting the embossed neighboring region based on the segmentation strategy to obtain the segmented region includes:

[0093] S201. Obtain the preset adjacent side dimensions;

[0094] S202. Using the center position of the embossing as the reference point, extract the three-dimensional point cloud data within the preset adjacent side size range as the embossing neighboring area;

[0095] S203. Obtain the roller mold parameters for electrode embossing, and obtain the segmentation strategy based on the roller mold parameters;

[0096] S204. Based on the segmentation strategy, the adjacent areas of the embossed area are segmented to obtain the segmented area, wherein the segmented area includes the planar area and the embossed area.

[0097] As described in steps S201 to S204 above, based on the embossing mold diameter and production process requirements, a pre-defined adjacent edge size (i.e., the radius or side length of the envelope) is set. This size should be a certain multiple greater than the maximum diameter of the embossing to ensure complete coverage of the embossing and its surrounding planar area. Using the center point of the 3D embossing as a reference, the adjacent edge size range is expanded on the X and Y planes, and all 3D point clouds within this range are extracted to form the embossing adjacent area, including the embossing and part of the surrounding background plane. The roller mold parameters (such as mold diameter, adjacent spacing, compensation coefficient, etc.) are read, and based on the predefined segmentation table or algorithm logic, the appropriate segmentation threshold, filter kernel size, minimum connected component area, and other strategy parameters are calculated. Combined with the segmentation strategy, the adjacent area of ​​the embossing is segmented... Point clouds or projected images are classified into planar regions (points near the reference plane with no significant elevation changes) and embossed regions (points above the plane with corresponding pattern protrusions or depressions). Segmentation methods can employ simple binarization based on height thresholds, or combine multimodal fusion strategies such as plane fitting residuals, region growing, and clustering. By presetting a large neighboring edge size, it is ensured that regardless of slight offsets in the embossing position or workpiece vibration, the neighboring region can completely encompass the entire embossing and surrounding plane. The segmentation strategy relies entirely on mold parameters and a pre-configured segmentation table, requiring no manual fine-tuning. This enables one-click switching and batch detection of different embossing specifications. Combined with geometric thresholds and statistical analysis, it can clearly distinguish between planar and embossed regions, suppressing missegmentation caused by surface textures, dust, or minor scratches.

[0098] In a preferred embodiment, the step of obtaining the roller die parameters for electrode embossing and obtaining a segmentation strategy based on the roller die parameters includes:

[0099] S2031. Obtain the roller mold parameters for electrode embossing, and obtain the embossing mold diameter and the spacing between multiple embossing molds based on the roller mold parameters;

[0100] S2032. Sort multiple adjacent embossing mold spacings in ascending order, and obtain the first adjacent embossing mold spacing as the target embossing mold spacing.

[0101] S2033, Obtain the compensation coefficient;

[0102] S2034. Obtain the segmentation value based on the embossing die diameter, the target embossing die spacing, and the compensation coefficient;

[0103] S2035. Obtain the segmentation table, wherein the segmentation table includes multiple segmentation value intervals and the segmentation strategy corresponding to each segmentation value interval;

[0104] S2036. Obtain the corresponding segmentation strategy from the segmentation table based on the segmentation value range corresponding to the segmentation value.

[0105] As described in steps S2031 to S2036 above, the roller mold parameters corresponding to the current electrode embossing are read, including the embossing mold diameter, i.e., the forming diameter of a single embossing pattern, and the set of adjacent embossing mold spacings, representing the actual interval between two adjacent embossing centers on the roller surface. The spacing sets are sorted from smallest to largest, with the minimum value corresponding to the densest adjacent embossing pairs. The minimum value is taken as the target embossing mold spacing. Based on the actual working conditions of the production line, such as temperature, material elasticity, and roller wear, a compensation coefficient corresponding to the current minimum value is obtained from a pre-calibrated fluctuation table or compensation library. This coefficient is used to correct measurement deviations caused by dynamic working conditions (irregular deformation of the electrode material surface due to uneven force during roller pressing, and deformation of the embossing). The embossing mold diameter, target spacing, and compensation coefficient are combined to calculate a segmentation value. This value reflects the height threshold or projection distance threshold of the plane and embossing transition area. The formula for calculating the segmentation value is as follows: In the formula, S represents the segmentation value, k represents the compensation coefficient, D represents the diameter of the embossing mold, and L represents the target embossing mold spacing. A segmentation table is predefined, containing several segmentation value intervals and corresponding segmentation strategies for each interval (such as binarization threshold, region growth parameters, filter kernel size, etc.). The currently calculated segmentation value is compared to determine which interval in the segmentation table it falls into. Then, the segmentation parameter set corresponding to that interval is extracted as the specific segmentation strategy for the adjacent area of ​​this embossing. The segmentation strategy is explicitly associated with the mold parameters and working condition compensation. There is no need to manually set the threshold. It is suitable for embossing molds with different diameters and spacings. Only the mold parameters and compensation library need to be updated, and the compensation coefficient is introduced to automatically correct the deformation caused by material expansion, temperature changes, or equipment wear, ensuring that the segmentation value accurately matches the real embossing boundary. Through the accurately calculated segmentation value and targeted strategy, the plane and embossing contour can be distinguished more accurately, reducing the measurement deviation caused by missegmentation.

[0106] In a preferred embodiment, the step of obtaining the compensation coefficient includes:

[0107] S20331. Obtain the spacing fluctuation value based on the spacing of multiple embossing molds;

[0108] S20332. Obtain the fluctuation table, wherein the fluctuation table includes multiple interval fluctuation value ranges and the compensation coefficient corresponding to each interval fluctuation value range;

[0109] S20333: Obtain the corresponding compensation coefficient from the fluctuation table based on the interval of the interval fluctuation value.

[0110] As described in steps S20331 to S20333 above, the spacing fluctuation value is calculated from the spacing between multiple sets of adjacent embossing molds. The formula for calculating the spacing fluctuation value is as follows: In the formula, k represents the spacing fluctuation value, and t represents the number of the spacing between multiple embossing dies, t=1,2,3…m. Let t be the spacing of the embossing die. A pre-established mapping table of spacing fluctuation and compensation coefficient (fluctuation table) is created. The table contains multiple fluctuation value ranges and corresponding compensation coefficients. This table is based on equipment calibration and experimental data statistics and can reflect the influence of working condition changes (irregular deformation of the electrode material surface caused by uneven force during rolling and deformation of embossing) on ​​the measurement. The actual calculated fluctuation value is compared with each interval in the fluctuation table. After determining the interval, the compensation coefficient corresponding to that interval is directly read to ensure that the segmentation threshold matches the actual situation. By quantifying the current die spacing fluctuation, the compensation coefficient is dynamically adjusted to avoid embossing deformation caused by uneven force during rolling, improve measurement accuracy, and effectively offset measurement errors. The fluctuation table is based on a large amount of experimental and calibration data. The compensation coefficient can finely correct the segmentation threshold, making the segmentation of the plane and embossing area more accurate, thereby improving the accuracy of height and diameter measurement.

[0111] In a preferred embodiment, the step of obtaining the reference plane based on the segmented region includes:

[0112] S301. Obtain a planar region based on segmented regions;

[0113] S302. Obtain all three-dimensional points in the planar region and summarize them into a planar three-dimensional point set;

[0114] S303. Obtain the equation of the reference plane based on a three-dimensional point set in a plane;

[0115] S304. Obtain equation parameters based on the datum plane equation, and use the plane characterized by the equation parameters as the datum plane.

[0116] As described in steps S301 to S304 above, all point clouds belonging to the planar portion are selected from the segmented region, and all three-dimensional coordinate points within the planar region are collected to form a complete planar three-dimensional point set. This point set covers the substrate surface around the embossing, providing sufficient data support for planar fitting. The planar three-dimensional point set is fitted using the least squares method or the minimum median variance method to solve the general form of the planar equation, A*x+B*y+C*z+E=0, where the normal vector n=(A,B,C) and the intercept E are output parameters. The fitted equation parameters (A,B,C,E) are used as the mathematical description of the reference plane. Statistical fitting methods (least squares or minimum median variance) can automatically eliminate the influence of a small number of outliers, ensuring that the planar parameters accurately reflect the actual tilt and unevenness of the material. The point cloud of the planar region is specifically fitted to avoid interference from the embossing protrusions in the calculation of the reference plane, ensuring that the zero reference plane for height measurement is true and reliable.

[0117] In a preferred embodiment, the steps of acquiring all three-dimensional points within the segmented region, acquiring a set of height three-dimensional points that meet a first preset condition based on all three-dimensional points and a reference plane, and acquiring the embossing height based on the set of height three-dimensional points and the reference plane include:

[0118] S401. Obtain all 3D points within the segmented region;

[0119] S402. Obtain the distance from each 3D point to the reference plane and mark it as the first distance;

[0120] S403. Sort the multiple first distances in descending order to obtain the first sorting table;

[0121] S404. Obtain the preset distance quantity, select three-dimensional points that meet the preset distance quantity from the first sorting table in descending order, and summarize them into a height three-dimensional point set;

[0122] S405. Obtain the average of the sum of all first distances in the height 3D point set, and use it as the embossing height.

[0123] As described in steps S401 to S405 above, from the segmented embossed area and planar area, focus on the entire segmented area, collect the coordinates of all three-dimensional points within it, forming a complete point cloud. For each three-dimensional point within the segmented area, calculate its vertical distance to the plane according to the reference plane equation (the distance formula from point to plane can be used for calculation). Record the distance of each point as the first distance. Sort all the first distances from largest to smallest to generate a first sorting table. Based on the preset distance quantity parameter (e.g., select the top 5 or top 10 maximum distance points), take the points ranked at the preset distance quantity from the first sorting table and summarize them into a height three-dimensional point set. Calculate the average value of all the first distances in the height three-dimensional point set. This average value is the final embossed height measurement value, which can reflect the typical characteristic height of embossed protrusions or depressions. Selecting the average value of several maximum distance points can suppress single-point errors caused by isolated noise or extreme outliers, and improve the stability of height measurement. The preset distance quantity can be adjusted according to the complexity of the embossed texture and the measurement accuracy requirements.

[0124] In a preferred embodiment, the steps of establishing a local coordinate system based on a reference plane, obtaining a set of three-dimensional diameter points that meet a second preset condition based on the local coordinate system, and obtaining the embossing diameter based on the set of three-dimensional diameter points and the local coordinate system include:

[0125] S501. Establish a local coordinate system based on the reference plane;

[0126] S502. Transform the coordinates of the 3D point cloud data within the segmented region to the coordinates corresponding to the local coordinate system;

[0127] S503. Extract the Z-axis coordinates of each 3D point in the local coordinate system;

[0128] S504. Obtain a preset height threshold, select three-dimensional points whose Z-axis coordinates are greater than the preset height threshold, and summarize them into a diameter three-dimensional point set.

[0129] S505. Obtain the bounding box of the three-dimensional point set of the diameter, and obtain the span difference of the bounding box in the Y-axis direction, and use it as the embossing diameter.

[0130] As described in steps S501 to S505 above, a new three-dimensional coordinate system (i.e., a local coordinate system) is defined with the reference plane as a reference. The Z′ axis is parallel to the normal vector of the reference plane and points outward from the plane. The X′ axis remains in the same direction as the original global coordinate system X-axis. The Y′ axis is determined by the right-hand rule, making (X′, Y′, Z′) a right-hand coordinate system. All three-dimensional points (x, y, z) within the segmented area are transformed into the local coordinate system according to the rotation matrix and translation vector defined above, resulting in new coordinates (x′, y′, z′). For each transformed point, its coordinate value z′ along the local Z′ axis is read. This value represents the height of the point relative to the reference plane. Based on the second preset condition (such as a height threshold, usually slightly lower than the highest point of the embossing), all points less than the height threshold are selected to form a straight line. A 3D point set is used to establish an oriented bounding box (OBB) for the diameter point set. This bounding box is oriented along the three axes of the local coordinate system. The minimum and maximum values ​​of the bounding box in the local Y′ axis direction are read, and their difference is the diameter of the embossing on that section. The Z′ axis of the local coordinate system is completely parallel to the reference plane, so that all subsequent height and lateral dimension calculations can be performed in an ideal plane and normal system, eliminating the influence of workpiece tilt that may exist in the global coordinate. After coordinate transformation, the extraction of embossing height and diameter depends only on a single coordinate component of each point, avoiding complex 3D geometric calculations. The algorithm is more intuitive and easier to implement. By setting a height threshold, only points close to the top edge of the embossing are retained, which can effectively exclude the substrate plane and local noise points, ensuring that the bounding box calculation only covers the true embossing outline.

[0131] And, a terminal for measuring the height and diameter of circular embossing based on three-dimensional vision, comprising:

[0132] One or more processors;

[0133] A storage device on which one or more programs are stored;

[0134] When one or more programs are executed by one or more processors, the one or more processors implement a method for measuring the height and diameter of circular embossing based on three-dimensional vision.

[0135] The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.

Claims

1. A method for measuring the height and diameter of circular embossed patterns based on three-dimensional vision, characterized in that, include: Obtain the three-dimensional point cloud data of the electrode embossing, obtain the two-dimensional depth image based on the three-dimensional point cloud data, and obtain the center position of the embossing. The embossed flower's neighboring regions are obtained based on the center position of the embossed flower, and a segmentation strategy is obtained. The neighboring regions of the embossed flower are then segmented based on the segmentation strategy to obtain the segmented regions. Obtain the reference plane based on segmented regions; Obtain all three-dimensional points within the segmented area; obtain a set of height three-dimensional points that meet the first preset condition based on all three-dimensional points and the reference plane; obtain the embossing height based on the set of height three-dimensional points and the reference plane. A local coordinate system is established based on the reference plane. A set of three-dimensional points for the diameter that meets the second preset condition is obtained based on the local coordinate system. The embossing diameter is obtained based on the set of three-dimensional points for the diameter and the local coordinate system. The steps for acquiring 3D point cloud data of electrode embossing, obtaining a 2D depth image based on the 3D point cloud data, and determining the center position of the embossing include: Obtain the 3D point cloud data of electrode embossing; The 3D point cloud data is preprocessed to obtain preprocessed 3D point cloud data, wherein the preprocessing includes filtering; The sampling parameters of the three-dimensional point cloud data of the collected electrode embossing are obtained. Based on the sampling parameters, the preprocessed three-dimensional point cloud data is projected into a two-dimensional depth image, where the pixel gray value corresponds to the Z value of the three-dimensional point. The center position of the embossed pattern is obtained based on a two-dimensional depth image. The steps for obtaining the center position of embossed flowers based on a two-dimensional depth image include: Obtain the roller die parameters for electrode embossing, and obtain the spacing between multiple adjacent embossing dies based on the roller die parameters; Sort the spacing between multiple adjacent embossing molds in ascending order, and take the spacing between the first adjacent embossing mold as the filter mask size; The two-dimensional depth image is filtered based on the filter mask size to obtain the filtered two-dimensional depth image. Obtain the difference in grayscale value of each pixel before and after filtering; Obtain the grayscale difference threshold, filter the pixels whose grayscale difference exceeds the grayscale difference threshold, and form a candidate feature pixel set; Connected components are extracted from candidate feature pixel sets, and embossed connected components are obtained from the connected components. Obtain the center pixel coordinates of the embossed flower's connected domain, and map these center pixel coordinates to three-dimensional space to obtain the center position of the embossed flower. The steps of obtaining the embossed flower's neighboring region based on the embossed flower's center location, obtaining a segmentation strategy, and segmenting the embossed flower's neighboring region based on the segmentation strategy to obtain the segmented region include: Get the preset adjacent edge dimensions; Using the center position of the embossing as the reference point, extract the three-dimensional point cloud data within the preset adjacent edge size range as the embossing neighboring area; Obtain the roller die parameters for electrode embossing, and obtain the segmentation strategy based on the roller die parameters; The segmented region is obtained by segmenting the adjacent areas of the embossed area based on the segmentation strategy. The segmented region includes the planar area and the embossed area. The steps for obtaining the roller die parameters for electrode embossing and obtaining the segmentation strategy based on the roller die parameters include: Obtain the roller mold parameters for electrode embossing, and based on the roller mold parameters, obtain the embossing mold diameter and the spacing between multiple embossing molds; Sort the spacing between multiple adjacent embossing molds in ascending order, and take the spacing between the first adjacent embossing mold as the target embossing mold spacing. Obtain the compensation coefficient; The segmentation value is obtained based on the embossing die diameter, the target embossing die spacing, and the compensation coefficient; Obtain the segmentation table, which includes multiple segmentation value ranges and the segmentation strategy corresponding to each segmentation value range; The corresponding segmentation strategy is obtained from the segmentation table based on the segmentation value range corresponding to the segmentation value; The steps of acquiring all three-dimensional points within the segmented region, acquiring a set of height-based three-dimensional points that meet the first preset condition based on all three-dimensional points and the reference plane, and acquiring the embossing height based on the set of height-based three-dimensional points and the reference plane include: Obtain all 3D points within the segmented region; Obtain the distance from each 3D point to the reference plane and mark it as the first distance; Sort the multiple first distances in descending order to obtain the first sorting table; Obtain the preset distance quantity, select three-dimensional points that meet the preset distance quantity from the first sorting table in descending order, and summarize them into a height three-dimensional point set; Obtain the average of the sums of all first distances in the height 3D point set, and use it as the embossing height; The steps of establishing a local coordinate system based on a reference plane, obtaining a set of three-dimensional points representing the diameter that meet the second preset condition based on the local coordinate system, and obtaining the embossing diameter based on the set of three-dimensional points representing the diameter and the local coordinate system include: Establish a local coordinate system based on the reference plane; Transform the coordinates of the 3D point cloud data within the segmented region to the coordinates corresponding to the local coordinate system; Extract the Z-axis coordinates of each 3D point in the local coordinate system; Obtain a preset height threshold, select 3D points whose Z-axis coordinates are greater than the preset height threshold, and summarize them into a diameter 3D point set; Obtain the bounding box of the 3D point set of diameter, and obtain the span difference of the bounding box in the Y-axis direction, and use it as the embossing diameter.

2. The method for measuring the height and diameter of circular embossing based on three-dimensional vision according to claim 1, characterized in that, The steps to obtain the compensation coefficient include: The spacing fluctuation value is obtained based on the spacing of multiple embossing molds; Obtain the fluctuation table, which includes multiple interval fluctuation value ranges and the corresponding compensation coefficient for each interval fluctuation value range; The corresponding compensation coefficient is obtained from the fluctuation table based on the interval of the interval fluctuation value.

3. The method for measuring the height and diameter of circular embossing based on three-dimensional vision according to claim 1, characterized in that, The steps for obtaining the reference plane based on the segmented region include: Obtaining planar regions based on segmented regions; Obtain all 3D points in the planar region and aggregate them into a planar 3D point set; Equation of reference plane based on planar 3D point set; The equation parameters are obtained based on the equation of the reference plane, and the plane characterized by the equation parameters is used as the reference plane.

4. A terminal for measuring the height and diameter of circular embossed patterns based on three-dimensional vision, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When one or more programs are executed by one or more processors, the one or more processors implement the method for measuring the height and diameter of circular embossing based on three-dimensional vision as described in any one of claims 1 to 3.

Images