Polishing allowance rapid detection method based on point cloud reconstruction

Abstract

This invention relates to the field of 3D reconstruction technology and discloses a rapid detection method for polishing allowance based on point cloud reconstruction. The method includes: filtering noise from the point cloud on the surface of the workpiece to be polished to obtain a complete point cloud set; matching the point cloud with features of a theoretical geometric model to obtain a spatial mapping relationship; generating a 3D deviation distribution map based on the spatial mapping relationship; segmenting the point cloud region based on the 3D deviation distribution map to obtain an initial abnormal region set; reconstructing the initial abnormal region set to obtain connected abnormal regions; optimizing the boundary consistency of the connected abnormal regions to obtain the precise boundary of the polishing allowance; and finally generating a detection report. This technology improves accuracy and efficiency. Noise filtering and optimization of data, mapping relationships, and distribution maps ensure accurate deviation and avoid errors. Furthermore, the abnormal region processing improves practicality, accurately locates key areas, and the report guides subsequent polishing, adapting to the needs of modern production lines.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional reconstruction technology, and in particular to a rapid method for detecting polishing allowance based on point cloud reconstruction. Background Technology

[0002] Existing polishing allowance detection methods rely on traditional contact measurement or simple point cloud analysis. The former requires manual point-by-point operation, which is extremely inefficient when dealing with complex curved workpieces and is prone to damaging the workpiece surface due to physical contact. The latter does not perform systematic noise filtering and optimization on the original point cloud data, resulting in discrete noise points in the point cloud data interfering with subsequent deviation calculations. This makes it impossible to accurately reflect the actual surface condition of the workpiece, ultimately leading to large errors in the polishing allowance detection results, which is difficult to meet the requirements of high-precision manufacturing.

[0003] Traditional methods lack standardized feature matching and region segmentation mechanisms in the inspection process, making it difficult to establish a precise spatial mapping relationship between point cloud data and theoretical geometric models. Furthermore, the identification of abnormal regions is limited to single-point marking, failing to reconstruct connected regions and optimize boundaries. This makes it impossible to quickly locate critical polishing areas during inspection and to generate inspection reports containing precise boundaries and depth distributions. This severely restricts the efficiency and quality of subsequent polishing processes and fails to meet the demands of modern production lines for rapid and accurate inspection. Summary of the Invention

[0004] This invention provides a rapid method for detecting polishing allowance based on point cloud reconstruction to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a rapid polishing allowance detection method based on point cloud reconstruction, comprising: S1. Perform noise filtering on the surface point cloud data of the workpiece to be polished to obtain the complete point cloud set of the workpiece to be polished. S2. Based on the complete point cloud, perform feature matching on the theoretical geometric model of the workpiece to be polished to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model; S3. Based on the spatial mapping relationship, the theoretical geometric model is compared for deviation to obtain a three-dimensional deviation distribution map of the workpiece to be polished; S4. Based on the three-dimensional deviation distribution map, the judgment criteria of the workpiece to be polished are divided into regions to obtain the initial abnormal region set of the workpiece to be polished. S5. Reconstruct the initial abnormal region set to obtain the connected abnormal region of the workpiece to be polished; S6. Perform boundary consistency optimization on the connected abnormal region to obtain the precise boundary of the polishing allowance of the workpiece to be polished; S7. Based on the precise boundary of the polishing allowance, generate a polishing allowance detection report for the workpiece to be polished.

[0006] In a preferred embodiment, the step of noise filtering the surface point cloud data of the workpiece to be polished to obtain the complete point cloud set of the workpiece includes: Scan the surface of the workpiece to be polished to obtain the surface point cloud data of the workpiece; Discrete noise points are filtered out from the surface point cloud data to obtain a preliminary filtered point cloud set of the workpiece to be polished; The initial filtered point cloud is optimized to obtain the complete point cloud of the workpiece to be polished.

[0007] In a preferred embodiment, the step of performing feature matching on the theoretical geometric model of the workpiece to be polished based on the complete point cloud to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model includes: Feature extraction is performed on the complete point cloud to obtain the feature points of the workpiece to be polished; Spatial registration is performed between the feature points and the feature points of the theoretical geometric model in the workpiece to be polished to obtain the spatial correspondence between the feature points and the feature points of the theoretical geometric model. The spatial correspondence is iteratively registered to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model.

[0008] In a preferred embodiment, the step of comparing the deviation of the theoretical geometric model based on the spatial mapping relationship to obtain a three-dimensional deviation distribution map of the workpiece to be polished includes: A spatial mapping relationship transformation is performed on the complete point cloud to obtain the standard model coordinate system of the workpiece to be polished; The normal distance from the complete point cloud to the theoretical geometric model is calculated based on the standard model coordinate system. The formula for calculating the normal distance is as follows: in, Indicates the first Normal distance of each point The first point in the point cloud set The coordinates of the points This represents the coordinates of the nearest point on the surface of the theoretical model. The theoretical model surface represents the first... Normal vectors at each point; A three-dimensional deviation distribution map of the workpiece to be polished is generated based on the normal distance.

[0009] In a preferred embodiment, generating a three-dimensional deviation distribution map of the workpiece to be polished based on the normal distance includes: According to the principle of visual distinction, the normal distance is color-coded to obtain the color information corresponding to the normal distance; The data points in the complete point cloud are colored based on the color information to obtain the colored point cloud of the workpiece to be polished. The colored point cloud is superimposed on the theoretical geometric model to obtain the three-dimensional deviation distribution map.

[0010] In a preferred embodiment, the step of segmenting the workpiece to be polished based on the three-dimensional deviation distribution map to obtain an initial set of abnormal regions of the workpiece to be polished includes: Based on the theoretical geometric model, a deviation analysis is performed on the workpiece to be polished to obtain the judgment criteria for the workpiece to be polished. Data points that exceed the judgment benchmark in the three-dimensional deviation distribution map are extracted and marked to obtain the abnormal points of the workpiece to be polished; Spatial clustering is performed on the abnormal points to obtain the initial abnormal region set of the workpiece to be polished.

[0011] In a preferred embodiment, the step of reconstructing the initial abnormal region set to obtain the connected abnormal regions of the workpiece to be polished includes: The initial set of abnormal regions is integrated to obtain a preliminary combination of the abnormal points. A spatial layout analysis is performed on the preliminary assembly to obtain the spatial location and extent of the assembly. Based on the judgment criteria, the spatial position and range of the assembly are analyzed and screened to obtain the abnormal connectivity area of ​​the workpiece to be polished.

[0012] In a preferred embodiment, the step of optimizing the boundary consistency of the connectivity anomaly region to obtain the precise boundary of the polishing allowance of the workpiece to be polished includes: Spatially sort the data points on the periphery of the connectivity anomaly region to obtain the contour point sequence of the connectivity anomaly region; Boundary reconstruction is performed on the contour point sequence to obtain the contour curve of the connected anomaly region; The contour curve is refined to obtain the precise boundary of the polishing allowance of the workpiece to be polished.

[0013] In a preferred embodiment, generating a polishing allowance inspection report for the workpiece to be polished based on the precise boundary of the polishing allowance includes: Based on the precise boundary of the polishing allowance, a three-dimensional depth analysis is performed on the polishing allowance area of ​​the workpiece to be polished to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model. Based on the theoretical geometric model, the polishing allowance area is geometrically mapped to obtain the position and contour shape of the polishing allowance area; A polishing allowance inspection report for the workpiece to be polished is generated based on the maximum indentation depth value, the location, and the contour shape.

[0014] To address the aforementioned problems, this invention also provides a rapid detection of polishing allowance based on point cloud reconstruction, comprising: The step of performing three-dimensional depth analysis on the polishing allowance area of ​​the workpiece to be polished based on the precise boundary of the polishing allowance to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model includes: Based on the precise boundary of the polishing allowance, spatial deviation analysis is performed on the point cloud data within the polishing allowance area to obtain the positional difference between the point cloud data and the precise boundary of the polishing allowance. The indentation depth of the polishing allowance area is calculated based on the positional difference to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model. The formula for calculating the indentation depth of the polishing allowance area is as follows: in, Indicates the maximum indentation depth value. The reference height value representing the theoretical reference surface. Indicates the first The height measurement value of a point cloud data point. This represents the total number of point cloud data within the polishing allowance area; The formula for calculating the weighted maximum depth of the polishing allowance area is as follows: This represents the maximum depth value of the weighted region. Indicates the number of sub-regions. Indicates the first Sub-regions Indicates the first The area value of each sub-region Indicates the first The height measurement value of the point cloud data.

[0015] Record the distribution location of the maximum indentation depth value in the polishing allowance area.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. This technology improves the accuracy and efficiency of polishing allowance detection through a standardized seven-step inspection process. First, noise filtering is performed on the point cloud data of the workpiece surface to remove discrete noise points and optimize the point cloud, obtaining a complete and high-quality point cloud set, laying an accurate data foundation for subsequent inspection. Then, feature matching is used to establish a spatial mapping relationship between the point cloud set and the theoretical geometric model. Combined with normal distance calculation, a three-dimensional deviation distribution map is generated, which can intuitively and accurately reflect the deviation between the workpiece surface and the theoretical model, avoiding the impact of data errors on the inspection results.

[0017] 2. This technology further enhances the practicality and reliability of detection by processing abnormal areas. It obtains connected abnormal areas through region segmentation and reconstruction, and then obtains precise boundaries for polishing allowance through boundary consistency optimization, which can accurately locate key polishing areas. Finally, it combines three-dimensional depth analysis to obtain information such as the maximum depression depth value, and generates a detection report containing the location, contour and depth of the allowance area. This can provide precise guidance for subsequent polishing processes, effectively ensure polishing quality, and meet the needs of modern production lines for efficient and accurate detection. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a method for rapid detection of polishing allowance based on point cloud reconstruction according to an embodiment of the present invention. The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0020] This application provides a method for rapid detection of polishing allowance based on point cloud reconstruction. The execution entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for rapid detection of polishing allowance based on point cloud reconstruction can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0021] Reference Figure 1The diagram shown is a flowchart illustrating a rapid polishing allowance detection method based on point cloud reconstruction according to an embodiment of the present invention. In this embodiment, the rapid polishing allowance detection method based on point cloud reconstruction includes: S1. Perform noise filtering on the surface point cloud data of the workpiece to be polished to obtain the complete point cloud set of the workpiece to be polished. In this embodiment of the invention, the step of performing noise filtering on the surface point cloud data of the workpiece to be polished to obtain the complete point cloud set of the workpiece includes: Scan the surface of the workpiece to be polished to obtain the surface point cloud data of the workpiece; Discrete noise points are filtered out from the surface point cloud data to obtain a preliminary filtered point cloud set of the workpiece to be polished; The initial filtered point cloud is optimized to obtain the complete point cloud of the workpiece to be polished.

[0022] Specifically, a comprehensive scan of the surface of the workpiece to be polished is performed. During the scanning process, the distance between the scanner and the workpiece surface is kept stable to ensure that the scan covers all the outer surfaces of the workpiece, including curved surfaces, corners and other details. Through the laser emission and reception device of the scanner, the spatial position information of the workpiece surface is converted into discrete point coordinate data. The set of these discrete point coordinate data is the surface point cloud data of the workpiece to be polished.

[0023] Furthermore, a combination of manual screening and regional comparison is used to filter out discrete noise points in the surface point cloud data. First, the overall distribution pattern of the surface point cloud data is observed to identify isolated points that deviate significantly from the normal point cloud distribution range. These isolated points are discrete noise points. Then, the positions of suspected noise points are compared with those of their neighboring points. If the distance between a point and most of its surrounding points is much greater than the point spacing of the normal point cloud, and the point cannot form a continuous surface pattern with its surrounding points, then the point is determined to be a discrete noise point and is deleted from the surface point cloud data. After filtering out all discrete noise points, the remaining point cloud data set is the preliminary filtered point cloud set of the workpiece to be polished.

[0024] Furthermore, when optimizing the preliminary filtered point cloud, the system first checks for areas with uneven point cloud density. For areas with low point cloud density, it analyzes the distribution pattern of points around the area and generates new point coordinates in the blank areas based on the positional relationship of adjacent points, ensuring that the point cloud density of the area is consistent with other areas. For areas with high point cloud density, it removes some duplicate or overly close points, retaining key points that accurately reflect the surface morphology of the workpiece. At the same time, it calibrates the coordinates of all points in the preliminary filtered point cloud to ensure that the spatial position of each point perfectly matches the actual surface position of the workpiece. After density adjustment and coordinate calibration, the resulting point cloud data set is the complete point cloud set of the workpiece to be polished.

[0025] S2. Based on the complete point cloud, perform feature matching on the theoretical geometric model of the workpiece to be polished to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model; In this embodiment of the invention, the step of performing feature matching on the theoretical geometric model of the workpiece to be polished based on the complete point cloud to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model includes: Feature extraction is performed on the complete point cloud to obtain the feature points of the workpiece to be polished; Spatial registration is performed between the feature points and the feature points of the theoretical geometric model in the workpiece to be polished to obtain the spatial correspondence between the feature points and the feature points of the theoretical geometric model. The spatial correspondence is iteratively registered to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model.

[0026] Specifically, by observing the surface morphology of the workpiece to be polished as presented by the complete point cloud, structures with obvious geometric features on the workpiece surface, such as edges, corners, protrusions, depressions, and process holes, are identified. These structures will show different distribution density and arrangement direction of points in the point cloud data compared to the surrounding area. For these identified special structures, the spatial positional relationship of points within the structure is analyzed one by one, such as the distance between adjacent points and the distribution trend of points. Key points that can represent the geometric attributes of the structure are selected. The set of these key points is the feature point of the workpiece to be polished.

[0027] Furthermore, in the theoretical geometric model of the workpiece to be polished, the geometric structures corresponding to the workpiece feature points extracted in the first step are found, and the theoretical feature points and their coordinates on the theoretical geometric model are determined. With the spatial position of the workpiece feature points as a reference, the spatial orientation of the theoretical geometric model is adjusted, including operations such as rotation and translation, so that the theoretical feature points on the theoretical geometric model coincide with the workpiece feature points in spatial position as much as possible. Then, the positional offset of each workpiece feature point and the corresponding theoretical feature point in the X, Y, and Z coordinate axes is recorded. These offsets and the corresponding feature point pairs are associated records, which together constitute the spatial correspondence between the feature points and the feature points of the theoretical geometric model.

[0028] Furthermore, based on the spatial correspondence obtained in the second step, the overall positional deviation between all workpiece feature points and their corresponding theoretical feature points is calculated. It is then determined whether the deviation meets the preset matching accuracy requirements. If the deviation does not meet the requirements, the spatial orientation of the theoretical geometric model is readjusted based on the current overall deviation. After adjustment, the spatial correspondence between the workpiece feature points and the theoretical feature points is recalculated, and the overall positional deviation is evaluated again. The process of "calculating deviation - adjusting model - updating spatial correspondence - evaluating deviation" is repeated until the overall positional deviation between the workpiece feature points and the theoretical feature points reaches the preset accuracy. At this point, a stable spatial positional association rule is formed between the complete point set and the theoretical geometric model. This rule is the spatial mapping relationship between the complete point set and the theoretical geometric model.

[0029] S3. Based on the spatial mapping relationship, the theoretical geometric model is compared for deviation to obtain a three-dimensional deviation distribution map of the workpiece to be polished; In this embodiment of the invention, the step of comparing the deviation of the theoretical geometric model based on the spatial mapping relationship to obtain a three-dimensional deviation distribution map of the workpiece to be polished includes: A spatial mapping relationship transformation is performed on the complete point cloud to obtain the standard model coordinate system of the workpiece to be polished; The normal distance from the complete point cloud to the theoretical geometric model is calculated based on the standard model coordinate system. The formula for calculating the normal distance is as follows: in, Indicates the first Normal distance of each point The first point in the point cloud set The coordinates of the points This represents the coordinates of the nearest point on the surface of the theoretical model. The theoretical model surface represents the first... Normal vectors at each point; A three-dimensional deviation distribution map of the workpiece to be polished is generated based on the normal distance.

[0030] The step of generating a three-dimensional deviation distribution map of the workpiece to be polished based on the normal distance includes: According to the principle of visual distinction, the normal distance is color-coded to obtain the color information corresponding to the normal distance; The data points in the complete point cloud are colored based on the color information to obtain the colored point cloud of the workpiece to be polished. Specifically, the colored point cloud is superimposed on the theoretical geometric model to obtain the three-dimensional deviation distribution map. Based on the spatial mapping relationship between the obtained complete point cloud set and the theoretical geometric model, the spatial position and orientation of all points in the complete point cloud set are adjusted. Through operations such as rotation and translation, the coordinates of each point in the complete point cloud set are transformed to the coordinate system used by the theoretical geometric model, so that the complete point cloud set and the theoretical geometric model are in the same coordinate system. After this transformation process, the coordinate system obtained is consistent with the coordinate system of the theoretical geometric model, which is the standard model coordinate system of the workpiece to be polished.

[0031] Furthermore, in the obtained standard model coordinate system, for each point in the complete point cloud, the point with the closest spatial distance to that point is found on the surface of the theoretical geometric model. At the same time, the normal vector of the theoretical geometric model surface at this closest point is determined. Then, the vector between the point in the complete point cloud and the corresponding closest point is calculated. This vector is multiplied by the above normal vector, and the absolute value of the result is taken. In this way, the distance from each point in the complete point cloud to the theoretical geometric model is calculated one by one. These distances are the normal distances from the complete point cloud to the theoretical geometric model.

[0032] Furthermore, following the principle of visually distinguishable distances, corresponding colors are assigned to normal distances of different numerical ranges. For example, smaller normal distances correspond to lighter colors, and larger normal distances correspond to darker colors. This establishes a correspondence between normal distances and colors, obtaining color information for each normal distance. Based on this color information, each point in the complete point cloud is assigned a corresponding color, completing the coloring process of the complete point cloud and obtaining the colored point cloud of the workpiece to be polished. The colored point cloud and the theoretical geometric model of the workpiece to be polished are superimposed on the same display interface, ensuring that their spatial positions correspond completely. The resulting visualization, which includes the colored point cloud and the theoretical geometric model, is the three-dimensional deviation distribution map of the workpiece to be polished.

[0033] Specifically, the numerical range of all normal distances is statistically analyzed and evenly divided into multiple continuous sub-intervals. Each sub-interval corresponds to a specific degree of deviation, ensuring that the numerical boundaries of adjacent sub-intervals are clear and non-overlapping. Based on the principle of visual differentiation, each sub-interval is assigned a unique color. For example, the sub-interval with the smallest deviation is assigned light green, the sub-interval with a slightly larger deviation is assigned yellow, the sub-interval with a relatively large deviation is assigned orange, and the sub-interval with the largest deviation is assigned red. This ensures that the colors of different sub-intervals will not cause confusion during visual observation. Subsequently, association rules between normal distance values ​​and corresponding colors are established, clarifying the sub-interval to which each normal distance value belongs and the color corresponding to that sub-interval. The set of these association rules constitutes the color information corresponding to the normal distance.

[0034] Furthermore, each data point is extracted one by one from the complete point cloud set, and the normal distance value corresponding to each data point is obtained. Based on the color information corresponding to the normal distance obtained in the first step, the sub-interval to which the normal distance value of the current data point belongs is found, and the color corresponding to the sub-interval is determined. The determined color is directly assigned to the currently extracted data point to ensure that the color of the data point is completely consistent with the color corresponding to its own normal distance. In the above manner, the color assignment operation of all data points in the complete point cloud set is completed in sequence. The point cloud set composed of all data points assigned the corresponding colors is the colored point cloud of the workpiece to be polished.

[0035] Specifically, the corresponding point cloud set The coordinates of the first point are obtained by first scanning the surface of the workpiece to be polished to obtain surface point cloud data. Then, discrete noise points are filtered out from the surface point cloud data to obtain a preliminary filtered point cloud set. Next, the preliminary filtered point cloud set is optimized—the point cloud density is checked, new points are added to areas with low density according to the distribution pattern of surrounding points, and duplicate or too close points are removed from areas with high density. At the same time, the coordinates of all points are calibrated to finally obtain a complete point cloud set. The coordinates are the first point extracted from the complete point cloud set. The coordinates of the points.

[0036] Furthermore, the coordinates of the nearest point on the surface of the corresponding theoretical model are derived from first obtaining the theoretical geometric model of the workpiece to be polished, and then, in the standard model coordinate system obtained by transforming the spatial mapping relationship of the complete point cloud, targeting the first point in the complete point cloud. The point is located on the surface of the theoretical geometric model and its relationship to the first point is determined. The coordinates of the point that is spatially closest to the given point are the required coordinates; the corresponding theoretical model surface is at the th . The normal vector at point n is derived from the distance from the complete point cloud set on the surface of the determined theoretical model. After the nearest point, based on the surface geometric properties of the theoretical geometric model, calculate the vector perpendicular to the surface at this nearest point. This vector is the vector perpendicular to the surface at the nth point. The normal vector at each point.

[0037] Furthermore, the calculated distance can accurately reflect the first point in the complete point cloud set. The deviation between the position of each point on the surface of the workpiece to be polished and the position of the nearest point on the theoretical geometric model in the direction perpendicular to the surface of the theoretical model is the core basis for subsequent color coding of this distance according to the principle of visual distinction. This involves statistically analyzing the range of distance values ​​and dividing it into multiple sub-intervals, assigning a unique and visually distinguishable color to each sub-interval, establishing the association rule between distance and color, coloring the data points in the complete point cloud based on the color information to obtain a colored point cloud, and finally overlaying the colored point cloud with the theoretical geometric model to obtain a three-dimensional deviation distribution map.

[0038] Furthermore, the three-dimensional deviation distribution map can intuitively present the deviation between the surface of the workpiece to be polished and the theoretical geometric model. The calculation result of this distance provides accurate deviation data support for subsequent region segmentation and abnormal region identification based on the three-dimensional deviation distribution map. When the complete point cloud is concentrated... The spatial difference between the position of the point corresponding to the workpiece surface to be polished and the corresponding nearest point on the theoretical geometric model increases as the distance increases in the direction perpendicular to the surface of the theoretical model; when the point in the complete point cloud... The spatial difference between the position of the point corresponding to the workpiece surface to be polished and the position of the nearest corresponding point on the theoretical geometric model decreases as the distance decreases in the direction perpendicular to the surface of the theoretical model; when the point in the complete point cloud... When the position of the point on the workpiece surface to be polished completely coincides with the position of the nearest point on the theoretical geometric model in the direction perpendicular to the surface of the theoretical model, the calculated distance is zero.

[0039] Furthermore, this trend can accurately reflect the degree of agreement between the actual state of the workpiece surface to be polished and the theoretical geometric model. The larger the distance, the lower the degree of agreement at that position; the smaller the distance, the higher the degree of agreement; and a distance of zero indicates a perfect agreement. This provides a clear numerical basis for determining whether the workpiece surface needs polishing and for identifying the polishing area.

[0040] S4. Based on the three-dimensional deviation distribution map, the judgment criteria of the workpiece to be polished are divided into regions to obtain the initial abnormal region set of the workpiece to be polished. In this embodiment of the invention, the step of segmenting the workpiece to be polished based on the three-dimensional deviation distribution map to obtain an initial set of abnormal regions of the workpiece to be polished includes: Based on the theoretical geometric model, a deviation analysis is performed on the workpiece to be polished to obtain the judgment criteria for the workpiece to be polished. Data points that exceed the judgment benchmark in the three-dimensional deviation distribution map are extracted and marked to obtain the abnormal points of the workpiece to be polished; Spatial clustering is performed on the abnormal points to obtain the initial abnormal region set of the workpiece to be polished.

[0041] Specifically, when performing deviation analysis on the workpiece to be polished based on the theoretical geometric model, the design attributes of all key surfaces on the theoretical geometric model are first obtained. These attributes include standard design indicators such as surface contour shape, flatness requirements, and curvature parameters. Then, combined with the manufacturing precision specifications of the industry to which the workpiece belongs and the process capabilities of subsequent polishing processes, the maximum and minimum allowable deviation ranges of the workpiece surface are determined. For example, for curved surface workpieces, according to their design purpose, it is determined that the deviation between its surface and the theoretical curved surface must not be greater than a certain specific value and must not be less than a certain specific value. This determined deviation range is used as the standard for judging whether the workpiece surface points need further processing, and finally, the judgment criteria for the workpiece to be polished are obtained.

[0042] Furthermore, when extracting and marking data points that exceed the judgment benchmark in the three-dimensional deviation distribution map, the deviation value corresponding to each data point in the three-dimensional deviation distribution map is first checked one by one. This deviation value is the magnitude of the deviation between the workpiece surface position represented by the data point and the corresponding position in the theoretical geometric model. Then, the deviation value of each data point is compared with the deviation range in the judgment benchmark. If the deviation value of a certain data point is greater than the maximum deviation value of the judgment benchmark or less than the minimum deviation value of the judgment benchmark, it is determined that the data point exceeds the judgment benchmark. Subsequently, the data point is visually marked on the three-dimensional deviation distribution map. The marking method can use a different color or symbol than other data points. At the same time, the specific position information of the data point in the coordinate system is recorded to ensure that all data points that exceed the judgment benchmark are accurately identified and marked, and finally the abnormal points of the workpiece to be polished are obtained.

[0043] Furthermore, when performing spatial clustering on the anomalies, a spatial distance judgment rule is first set. This rule is as follows: if the distance between two anomalies in the X-axis direction, the Y-axis direction, and the Z-axis direction does not exceed a preset specific length, which is determined based on the size and surface features of the workpiece to be polished, ensuring that spatially adjacent and non-adjacent anomalies can be distinguished, then these two anomalies are determined to be spatially close to each other and belong to the same cluster. Next, an anomaly that has not been clustered is randomly selected from all anomalies as the starting point. Then, all other anomalies that satisfy the spatial distance judgment rule with the starting point are searched, and these anomalies and the starting point are grouped into a temporary cluster. After that, a starting point is selected again from the remaining unclustered anomalies, and the above search and classification operations are repeated until all anomalies are grouped into the corresponding clusters. Each cluster represents a spatially continuous initial anomaly region. All these initial anomaly regions are summarized to form a set, and finally, the initial anomaly region set of the workpiece to be polished is obtained.

[0044] S5. Reconstruct the initial abnormal region set to obtain the connected abnormal region of the workpiece to be polished; In this embodiment of the invention, the step of reconstructing the initial abnormal region set to obtain the connected abnormal region of the workpiece to be polished includes: The initial set of abnormal regions is integrated to obtain a preliminary combination of the abnormal points. A spatial layout analysis is performed on the preliminary assembly to obtain the spatial location and extent of the assembly. Based on the judgment criteria, the spatial position and range of the assembly are analyzed and screened to obtain the abnormal connectivity area of ​​the workpiece to be polished.

[0045] Specifically, the initial abnormal region set is first defined as containing multiple independent initial abnormal regions, each consisting of spatially adjacent abnormal points. Next, it is determined whether any two initial abnormal regions meet the integration conditions. Specifically, any abnormal point on the edge of the first initial abnormal region is selected, and then any abnormal point on the edge of the second initial abnormal region is selected. The distances between these two abnormal points in the X, Y, and Z axes of the standard model coordinate system are measured. If the distances in all three directions are less than the specific length determined based on the overall dimensions of the workpiece to be polished, then it is determined that these two initial abnormal regions need to be integrated, and all the abnormal points contained in them are grouped into a whole. This operation is repeated, and all initial abnormal regions in the initial abnormal region set are judged and integrated one by one. The final whole formed by each integrated abnormal point is the preliminary combination of the abnormal points.

[0046] Furthermore, using the standard model coordinate system as a reference frame, coordinates of all anomaly points contained in each preliminary assembly are collected, and the X, Y, and Z coordinates of each anomaly point are recorded. Then, the maximum and minimum values ​​of the X, Y, and Z coordinates of all anomaly points in each preliminary assembly are calculated. The spatial position of the assembly is determined by calculating the average of the extreme values ​​of the X, Y, and Z coordinates. The coordinate point corresponding to this average value is the spatial position of the preliminary assembly in the standard model coordinate system. The range of the assembly is determined by calculating the difference between the maximum and minimum values ​​of the X, Y, and Z coordinates. These three differences correspond to the extension length of the preliminary assembly in the X, Y, and Z axis directions, respectively. Through the above methods, the spatial position and range of the assembly can be obtained.

[0047] Furthermore, the previously determined judgment criteria for the workpiece to be polished are retrieved first. Then, for each preliminary assembly, the deviation values ​​of all abnormal points contained therein are checked one by one to confirm whether the deviation value of each abnormal point exceeds the deviation range of the judgment criteria. At the same time, the spatial position and range of the preliminary assembly are checked to see if they are within the area of ​​the workpiece to be polished that needs to be polished. If a preliminary assembly meets the two conditions of "all abnormal point deviations exceed the judgment criteria" and "spatial position and range are within the area to be polished", then the preliminary assembly is retained. All preliminary assemblies that meet the conditions are summarized, and each preliminary assembly that meets the conditions is the connected abnormal area of ​​the workpiece to be polished.

[0048] S6. Perform boundary consistency optimization on the connected abnormal region to obtain the precise boundary of the polishing allowance of the workpiece to be polished; In this embodiment of the invention, the step of optimizing the boundary consistency of the connected abnormal region to obtain the precise boundary of the polishing allowance of the workpiece to be polished includes: Spatially sort the data points on the periphery of the connectivity anomaly region to obtain the contour point sequence of the connectivity anomaly region; Boundary reconstruction is performed on the contour point sequence to obtain the contour curve of the connected anomaly region; The contour curve is refined to obtain the precise boundary of the polishing allowance of the workpiece to be polished.

[0049] Specifically, based on the spatial location and extent of the connectivity anomaly region, the data points surrounding this region are first identified: Each data point within the connectivity anomaly region is examined for its adjacent points. Adjacent points are defined as those whose distance from the data point in the X, Y, and Z axes of the standard model coordinate system does not exceed the average point spacing of the point cloud on the surface of the workpiece to be polished. If at least one of the adjacent points of a data point is a non-anomaly point not belonging to the connectivity anomaly region, then that data point is determined to be an outer data point of the connectivity anomaly region. After identifying the outer data points, they are spatially sorted: using the positive X-axis direction of the standard model coordinate system as the starting reference direction, all outer data points are initially sorted by their X-axis coordinate values ​​from smallest to largest; for outer data points with the same X-axis coordinate values, they are sorted by their Y-axis coordinate values ​​from smallest to largest; for outer data points with the same X-axis and Y-axis coordinate values, they are sorted by their Z-axis coordinate values ​​from smallest to largest. Through this ordered arrangement, the scattered outer data points are linked together into a continuous point sequence, obtaining the contour point sequence of the connectivity anomaly region.

[0050] Furthermore, starting from the first point in the contour point sequence, adjacent contour points are connected sequentially with straight line segments to form a preliminary polygonal contour. Next, the preliminary polygonal contour is smoothed: using the two endpoints of each straight line segment as a reference, and referring to the positions of one contour point before and after these endpoints, the direction of the straight line segment is adjusted so that the adjusted line segment better conforms to the actual distribution trend of the surrounding data points. For example, when three adjacent contour points exhibit a slightly curved distribution, the straight line segment connecting the first two points is adjusted to a line segment with a slight curvature, and the two ends of this line segment smoothly connect to the preceding and following line segments, respectively. This operation is repeated, adjusting each line segment formed by connecting all adjacent points in the contour point sequence one by one, ultimately forming a continuous, smooth curve that completely encloses the connected anomaly region, thus obtaining the contour curve of the connected anomaly region.

[0051] Furthermore, following the contour curve, the fit between the curve and the corresponding peripheral data points is checked segment by segment: Any segment of the contour curve with a length three times the average point spacing of the point cloud on the surface of the workpiece to be polished is selected. The distances from all points on this segment to their corresponding peripheral data points are calculated. If the average distance of a certain segment exceeds the average distribution error of the peripheral data points, the segment is corrected: using the corresponding peripheral data points as a reference, the curvature of the curve segment is fine-tuned to make the segment closer to these peripheral data points, while ensuring that the corrected segment remains continuous with adjacent segments without obvious breaks or abrupt turns. After checking and correcting all curve segments, the integrity and continuity of the contour curve are confirmed as a whole, ensuring that it accurately encloses the connected abnormal areas without deviation, thus obtaining the precise boundary of the polishing allowance for the workpiece to be polished.

[0052] S7. Based on the precise boundary of the polishing allowance, generate a polishing allowance detection report for the workpiece to be polished.

[0053] In this embodiment of the invention, generating a polishing allowance detection report for the workpiece to be polished based on the precise boundary of the polishing allowance includes: Based on the precise boundary of the polishing allowance, a three-dimensional depth analysis is performed on the polishing allowance area of ​​the workpiece to be polished to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model. Based on the theoretical geometric model, the polishing allowance area is geometrically mapped to obtain the position and contour shape of the polishing allowance area; A polishing allowance inspection report for the workpiece to be polished is generated based on the maximum indentation depth value, the location, and the contour shape.

[0054] The step of performing three-dimensional depth analysis on the polishing allowance area of ​​the workpiece to be polished based on the precise boundary of the polishing allowance to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model includes: Based on the precise boundary of the polishing allowance, spatial deviation analysis is performed on the point cloud data within the polishing allowance area to obtain the positional difference between the point cloud data and the precise boundary of the polishing allowance. The indentation depth of the polishing allowance area is calculated based on the positional difference to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model. The formula for calculating the indentation depth of the polishing allowance area is as follows: in, Indicates the maximum indentation depth value. The reference height value representing the theoretical reference surface. Indicates the first The height measurement value of a point cloud data point. This represents the total number of point cloud data within the polishing allowance area; The formula for calculating the weighted maximum depth of the polishing allowance area is as follows: This represents the maximum depth value of the weighted region. Indicates the number of sub-regions. Indicates the first Sub-regions Indicates the first The area value of each sub-region Indicates the first The height measurement value of the point cloud data.

[0055] Record the distribution location of the maximum indentation depth value in the polishing allowance area.

[0056] Specifically, based on the precise boundary of the polishing allowance, the polishing allowance area of ​​the workpiece to be polished is first determined. This area is the workpiece surface area corresponding to the point cloud data enclosed by the precise boundary of the polishing allowance. Then, the theoretical reference surface on the theoretical geometric model corresponding to the polishing allowance area is identified, that is, the surface in the theoretical geometric model whose position and shape match the polishing allowance area. Next, the height measurement values ​​of all point cloud data in the polishing allowance area are extracted one by one, and the reference height value of the theoretical reference surface is obtained at the same time. The difference between the height measurement value of each point cloud data and the reference height value of the theoretical reference surface is calculated, and all calculated differences are recorded. The largest value is selected from these differences. This value is the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model. At the same time, the specific position of the point cloud data corresponding to the maximum indentation depth value in the polishing allowance area is recorded.

[0057] Furthermore, using the coordinate system of the theoretical geometric model as the reference coordinate system, the coordinates of all contour points on the precise boundary of the polishing allowance are first transformed to this reference coordinate system to ensure that the coordinates of the contour points are consistent with the coordinate system of the theoretical geometric model. Then, the region on the theoretical geometric model that overlaps with the coordinate range of the transformed contour points is found. By comparing the coordinates of the contour points with the coordinate range of each surface of the theoretical geometric model, the specific location of the polishing allowance region on the theoretical geometric model is determined. For example, it is clear that the region corresponds to a certain functional surface of the theoretical geometric model, is close to which structural feature, and the specific coordinate interval. At the same time, the transformed contour points are connected in sequence to form a closed figure. The shape of this figure is the contour shape of the polishing allowance region. Through the above operations, the location and contour shape of the polishing allowance region are obtained.

[0058] Furthermore, the basic information of the workpiece to be polished is first collected, including the workpiece model, serial number, production batch, the time of this inspection, and the model of the point cloud scanning equipment used. Then, the location, contour shape, and maximum depression depth value of the previously obtained polishing allowance area are organized into structured content. This content is then organized in the order of "workpiece basic information - polishing allowance area location - polishing allowance area contour shape - maximum depression depth value and distribution" to ensure that the information in each part is accurate, clear, and without omissions or ambiguities. Finally, a document containing all the necessary information is formed, which is the polishing allowance inspection report of the workpiece to be polished.

[0059] Specifically, the range of the polishing allowance region is defined based on the precise boundary of the polishing allowance. This range is the spatial region corresponding to all point cloud data enclosed by the precise boundary of the polishing allowance. Then, each point cloud data is extracted one by one from this region. Taking the contour points of the precise boundary of the polishing allowance as a reference, for each extracted point cloud data, a line is constructed in the standard model coordinate system connecting the point cloud data to the nearest contour point. The length of this line in the direction perpendicular to the corresponding surface of the theoretical geometric model is measured. This length is the positional difference between the point cloud data and the precise boundary of the polishing allowance. Through the above operations, the positional difference corresponding to all point cloud data in the region is obtained one by one, thus obtaining the positional difference between the point cloud data and the precise boundary of the polishing allowance.

[0060] Furthermore, firstly, a theoretical reference surface corresponding to the polishing allowance area on the theoretical geometric model is determined, and the reference height of the theoretical reference surface in the standard model coordinate system is obtained. Then, for each point cloud data, the positional difference between it and the precise boundary of the polishing allowance is combined with the offset of the point cloud data in the reference height direction of the theoretical reference surface to calculate the concavity depth of each point cloud data relative to the theoretical reference surface. After completing the calculation of the concavity depth of all point cloud data, the largest value is selected from all the concavity depth values. This value is the maximum concavity depth value of the polishing allowance area relative to the theoretical geometric model.

[0061] Furthermore, the point cloud data corresponding to the maximum indentation depth value is found, and the X-axis, Y-axis, and Z-axis coordinates of the point cloud data are read in the standard model coordinate system. At the same time, the relative position of the point cloud data in the polishing allowance area is observed. For example, it is determined whether it is close to a certain contour edge of the precise boundary of the polishing allowance, whether it is in the center or edge position of the polishing allowance area, etc. The coordinate information and relative position description of the point cloud data are recorded together to form a complete record of the location of the maximum indentation depth value, and the distribution position of the maximum indentation depth value in the polishing allowance area is obtained.

[0062] Specifically, the reference height value of the theoretical reference surface is obtained by first determining the polishing allowance area of ​​the workpiece to be polished based on the precise boundary of the polishing allowance. This area is the workpiece surface area corresponding to the point cloud data enclosed by the precise boundary of the polishing allowance. Then, a theoretical reference surface that matches the position and shape of this polishing allowance area is found on the theoretical geometric model. Finally, the reference height value obtained from this theoretical reference surface is the reference height value of the theoretical reference surface. The height measurement value of each point cloud data point is obtained by extracting all point cloud data points within a defined polishing allowance area, and for each extracted point cloud data point, reading its height measurement value in the standard model coordinate system. The height measurement value of the first point cloud data point is then calculated. The height direction measurement value corresponding to the extracted point cloud data is the [value of the]th [item]. The height measurement value of each point cloud data; the total number of point cloud data in the polishing allowance area is obtained by counting all the extracted point cloud data in the polishing allowance area one by one, and calculating the total number of point cloud data in the area. This total number is the total number of point cloud data in the polishing allowance area.

[0063] The calculation process involves first subtracting the measured height of the corresponding point cloud data from the reference height value of the theoretical reference surface for each point cloud data within the polishing allowance area, thus obtaining the height difference for each point cloud data. Then, the largest difference is selected from all these height differences, and this largest difference is the maximum indentation depth value.

[0064] The maximum indentation depth value can accurately reflect the deepest indentation of the polishing allowance area relative to the theoretical geometric model in the height direction. This value is one of the key data used to generate the polishing allowance inspection report of the workpiece to be polished after the position and contour shape of the polishing allowance area are obtained by geometric mapping based on the theoretical geometric model. The maximum indentation depth value information in the inspection report can provide a clear depth reference for subsequent polishing processes and help determine the depth standard that the polishing operation needs to achieve.

[0065] The smaller the height measurement value of one or more point cloud data within the polishing allowance area, the larger the height difference obtained by subtracting the height measurement value of these point cloud data from the reference height value of the theoretical reference surface. The higher the probability of the maximum value appearing among all height differences, the greater the final maximum depression depth value will be. When the height measurement values ​​of all point cloud data within the polishing allowance area are larger and closer to the reference height value of the theoretical reference surface, the height difference corresponding to each point cloud data is smaller, the maximum value among all height differences will also decrease, and the final maximum depression depth value will also decrease. When the height measurement values ​​of all point cloud data within the polishing allowance area are equal to the reference height value of the theoretical reference surface, the height difference corresponding to each point cloud data is zero, and the maximum depression depth value obtained at this time is also zero. This indicates that the polishing allowance area completely matches the theoretical reference surface in the height direction, there is no depression, and no polishing treatment is needed for the depression.

[0066] Specifically, the number of sub-regions is obtained by first dividing the polishing allowance area into multiple independent sub-regions according to spatial distribution or point cloud density, and then counting each of these sub-regions one by one; The first sub-region is the one obtained after dividing the polishing allowance area. A small, independent region containing point cloud data within that region; the first The area value of each sub-region is calculated in the standard model coordinate system. The numerical value obtained by the area of ​​the space enclosed by the boundary contours of each sub-region; the first The height measurement value of the point cloud data is at the [number]th ...]. Extract point cloud data one by one from each sub-region, and read the first... The measured values ​​of the height direction of a point cloud in the standard model coordinate system.

[0067] Furthermore, the reference height value of the theoretical reference surface is obtained from the theoretical reference surface after the polishing allowance area is determined by the precise boundary of the polishing allowance.

[0068] Furthermore, the calculation first multiplies the area of ​​each sub-region by the maximum value of the theoretical reference surface height minus the measured point cloud height. Then, the products of all sub-regions are summed to obtain the numerator. Simultaneously, the area values ​​of all sub-regions are summed to obtain the denominator. Finally, the numerator is divided by the denominator, and the result is the weighted maximum depth value of the region. This value, combined with the sub-region area weights, more accurately reflects the overall recess depth of the polishing allowance area. It is crucial data for generating polishing allowance inspection reports and provides a comprehensive depth reference for subsequent polishing.

[0069] Furthermore, when the area value of a certain sub-region is larger, and the maximum value of the theoretical reference surface height value minus the point cloud height measurement value within that region is larger, the maximum depth value of the weighted region will increase while the conditions of other sub-regions remain unchanged. When the area values ​​of all sub-regions are small, and the maximum value of the theoretical reference surface height value minus the point cloud height measurement value within each region is small, the maximum depth value of the weighted region will be smaller. When the maximum value of the theoretical reference surface height value minus the point cloud height measurement value within all sub-regions is zero, the numerator is zero, and the maximum depth value of the weighted region is also zero. At this time, each sub-region of the polishing allowance region matches the theoretical reference surface, and no concave polishing is required.

[0070] Furthermore, this trend clearly reflects the severity of the depression in the polishing allowance area. Based on the change in the maximum depression depth, staff can quickly determine whether the area needs polishing and the focus of the polishing operation.

[0071] In the several embodiments provided by this invention, it should be understood that the disclosed method can be implemented in other ways.

[0072] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0073] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, and technology that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A rapid method for detecting polishing allowance based on point cloud reconstruction, characterized in that, The method includes: S1. Perform noise filtering on the surface point cloud data of the workpiece to be polished to obtain the complete point cloud set of the workpiece to be polished. S2. Based on the complete point cloud, perform feature matching on the theoretical geometric model of the workpiece to be polished to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model; S3. Based on the spatial mapping relationship, the theoretical geometric model is compared for deviation to obtain a three-dimensional deviation distribution map of the workpiece to be polished; S4. Based on the three-dimensional deviation distribution map, the judgment criteria of the workpiece to be polished are divided into regions to obtain the initial abnormal region set of the workpiece to be polished. S5. Reconstruct the initial abnormal region set to obtain the connected abnormal region of the workpiece to be polished; S6. Perform boundary consistency optimization on the connected abnormal region to obtain the precise boundary of the polishing allowance of the workpiece to be polished; S7. Based on the precise boundary of the polishing allowance, generate a polishing allowance detection report for the workpiece to be polished.

2. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 1, characterized in that, The noise filtering of the surface point cloud data of the workpiece to be polished to obtain the complete point cloud set of the workpiece includes: Scan the surface of the workpiece to be polished to obtain the surface point cloud data of the workpiece; Discrete noise points are filtered out from the surface point cloud data to obtain a preliminary filtered point cloud set of the workpiece to be polished; The initial filtered point cloud set is optimized to obtain the complete point cloud set of the workpiece to be polished.

3. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 1, characterized in that, The step of performing feature matching on the theoretical geometric model of the workpiece to be polished based on the complete point cloud to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model includes: Feature extraction is performed on the complete point cloud to obtain the feature points of the workpiece to be polished; Spatial registration is performed between the feature points and the feature points of the theoretical geometric model in the workpiece to be polished to obtain the spatial correspondence between the feature points and the feature points of the theoretical geometric model. The spatial correspondence is iteratively registered to obtain the spatial mapping relationship between the complete point cloud and the theoretical geometric model.

4. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 1, characterized in that, The step of comparing the deviations of the theoretical geometric model based on the spatial mapping relationship to obtain a three-dimensional deviation distribution map of the workpiece to be polished includes: A spatial mapping relationship transformation is performed on the complete point cloud to obtain the standard model coordinate system of the workpiece to be polished; The normal distance from the complete point cloud to the theoretical geometric model is calculated based on the standard model coordinate system. The formula for calculating the normal distance is as follows: in, Indicates the first Normal distance of each point The first point in the point cloud set The coordinates of the points This represents the coordinates of the nearest point on the surface of the theoretical model. The theoretical model surface represents the first... Normal vectors at each point; A three-dimensional deviation distribution map of the workpiece to be polished is generated based on the normal distance.

5. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 4, characterized in that, The step of generating a three-dimensional deviation distribution map of the workpiece to be polished based on the normal distance includes: According to the principle of visual distinction, the normal distance is color-coded to obtain the color information corresponding to the normal distance; The data points in the complete point cloud are colored based on the color information to obtain the colored point cloud of the workpiece to be polished. The colored point cloud is superimposed on the theoretical geometric model to obtain the three-dimensional deviation distribution map.

6. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 1, characterized in that, The process of segmenting the workpiece to be polished based on the three-dimensional deviation distribution map to obtain an initial set of abnormal regions for the workpiece includes: Based on the theoretical geometric model, a deviation analysis is performed on the workpiece to be polished to obtain the judgment criteria for the workpiece to be polished. Data points that exceed the judgment benchmark in the three-dimensional deviation distribution map are extracted and marked to obtain the abnormal points of the workpiece to be polished; Spatial clustering is performed on the abnormal points to obtain the initial abnormal region set of the workpiece to be polished.

7. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 1, characterized in that, The process of reconstructing the initial set of anomalous regions to obtain the connected anomalous regions of the workpiece to be polished includes: The initial set of abnormal regions is integrated to obtain a preliminary combination of the abnormal points. A spatial layout analysis is performed on the preliminary assembly to obtain the spatial location and extent of the assembly. Based on the judgment criteria, the spatial position and range of the assembly are analyzed and screened to obtain the abnormal connectivity area of ​​the workpiece to be polished.

8. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 1, characterized in that, The step of optimizing the boundary consistency of the connected abnormal regions to obtain the precise boundary of the polishing allowance of the workpiece to be polished includes: Spatially sort the data points on the periphery of the connectivity anomaly region to obtain the contour point sequence of the connectivity anomaly region; Boundary reconstruction is performed on the contour point sequence to obtain the contour curve of the connected anomaly region; The contour curve is refined to obtain the precise boundary of the polishing allowance of the workpiece to be polished.

9. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 1, characterized in that, The step of generating a polishing allowance inspection report for the workpiece to be polished based on the precise boundary of the polishing allowance includes: Based on the precise boundary of the polishing allowance, a three-dimensional depth analysis is performed on the polishing allowance area of ​​the workpiece to be polished to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model. Based on the theoretical geometric model, the polishing allowance area is geometrically mapped to obtain the position and contour shape of the polishing allowance area; A polishing allowance inspection report for the workpiece to be polished is generated based on the maximum indentation depth value, the location, and the contour shape.

10. The method for rapid detection of polishing allowance based on point cloud reconstruction as described in claim 9, characterized in that, The step of performing three-dimensional depth analysis on the polishing allowance area of ​​the workpiece to be polished based on the precise boundary of the polishing allowance to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model includes: Based on the precise boundary of the polishing allowance, spatial deviation analysis is performed on the point cloud data within the polishing allowance area to obtain the positional difference between the point cloud data and the precise boundary of the polishing allowance. The indentation depth of the polishing allowance area is calculated based on the positional difference to obtain the maximum indentation depth value of the polishing allowance area relative to the theoretical geometric model. The formula for calculating the indentation depth of the polishing allowance area is as follows: in, Indicates the maximum indentation depth value. The reference height value representing the theoretical reference surface. Indicates the first The height measurement value of a point cloud data point. This represents the total number of point cloud data within the polishing allowance area; The formula for calculating the weighted maximum depth of the polishing allowance area is as follows: This represents the maximum depth value of the weighted region. Indicates the number of sub-regions. Indicates the first Sub-regions Indicates the first The area value of each sub-region Indicates the first The height measurement value of the point cloud data. Record the distribution location of the maximum indentation depth value in the polishing allowance area.

Images