Complex component form and position error detection method based on cross-view point cloud data

Abstract

The invention relates to the technical field of component form and position error detection, and discloses a complex component form and position error detection method based on cross-view-angle point cloud data, which comprises the following steps: 1, multi-view-angle point cloud data acquisition: scanning a target component from different view angles to obtain multi-view-angle point cloud data; acquiring a plurality of point cloud data sets; and step 2, point cloud data registration: registering the plurality of point cloud data sets to obtain a registered point cloud data set. According to the complex component form and position error detection method based on the cross-view point cloud data, key surfaces and feature areas of complex components can be fully covered through collection of the multi-view point cloud data, the data integrity is ensured, a registration algorithm based on feature matching and a high-precision point cloud processing algorithm are adopted, and the detection accuracy is improved. The registration precision and the accuracy of feature extraction are remarkably improved, so that the precision of form and position error detection is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of component geometric error detection, and particularly to a method for detecting geometric errors of complex components based on cross-view point cloud data. Background Art

[0002] In modern industrial manufacturing, the detection of geometric errors of complex components is an important link to ensure product quality. Traditional geometric error detection methods mainly rely on contact measurement (such as coordinate measuring machines) or non-contact measurement from a single view (such as laser scanners). However, these methods have the following deficiencies when dealing with complex components:

[0003] 1) Limitations of single-view data: Complex components often have geometric features with multiple views and multiple surfaces. The point cloud data from a single view is difficult to comprehensively reflect the true shape and position errors of the component.

[0004] 2) Low registration accuracy: Existing methods have insufficient registration accuracy when processing cross-view point cloud data, which easily leads to data alignment errors and affects subsequent error calculations.

[0005] 3) Incomplete feature extraction: Traditional methods often can only extract some geometric features during feature extraction and cannot comprehensively cover all key features of complex components.

[0006] 4) Inaccurate error calculation: Existing methods lack comprehensive analysis of multi-view data when calculating geometric errors, resulting in inaccurate error calculation results. Summary of the Invention

[0007] In view of the problems that the above-mentioned existing technologies also have deficiencies in terms of automation level, real-time feedback, and adaptability to complex scenarios and are difficult to meet the requirements of high-precision detection, the present invention is proposed.

[0008] Therefore, the object of the present invention is to provide a method for detecting geometric errors of complex components based on cross-view point cloud data, and its purpose is to: be able to comprehensively cover the key surfaces and feature regions of complex components, ensure data integrity, and at the same time adopt a registration algorithm based on feature matching and a high-precision point cloud processing algorithm, significantly improving the registration accuracy and the accuracy of feature extraction, thereby enhancing the accuracy of geometric error detection.

[0009] To solve the above technical problems, the present invention provides the following technical solution: A method for detecting geometric errors of complex components based on cross-view point cloud data, including the following steps:

[0010] Step 1: Acquisition of multi-view point cloud data

[0011] Scan the target component from different perspectives to obtain multiple point cloud data sets; the selection of the different perspectives should cover the key surfaces and feature regions of the target component to ensure the comprehensiveness of the data;

[0012] Step Two: Point cloud data registration

[0013] Register the multiple point cloud data sets to obtain a registered point cloud data set;

[0014] Step Three: Geometric feature extraction

[0015] Based on the registered point cloud data set, extract the geometric features of the target component; the geometric features include surface contours, edges, holes, depressions or texture features;

[0016] Step Four: Geometric and positional error calculation

[0017] Compare the extracted geometric features with the design model of the target component to calculate the geometric and positional error of the target component;

[0018] Step Five: Defect detection and classification

[0019] Based on the calculated geometric and positional error, perform defect detection and classification on the target component.

[0020] As a preferred solution of the method for detecting geometric and positional errors of complex components based on cross-perspective point cloud data according to the present invention, wherein: the acquisition device for the point cloud data includes, but is not limited to, industrial-grade 3D scanners, handheld 3D scanners or lidars, and the resolution of the point cloud data is not less than 0.1 mm to ensure the high precision of the data.

[0021] As a preferred solution of the method for detecting geometric and positional errors of complex components based on cross-perspective point cloud data according to the present invention, wherein: the registration process includes a registration algorithm based on feature matching, extracts key feature points (such as edge points, corner points or surface texture points) in the point cloud data, and performs precise registration through the iterative closest point (ICP) algorithm.

[0022] As a preferred solution of the method for detecting geometric and positional errors of complex components based on cross-perspective point cloud data according to the present invention, wherein: the registration process further includes global registration and local registration to improve the registration accuracy and efficiency.

[0023] As a preferred solution of the method for detecting geometric and positional errors of complex components based on cross-perspective point cloud data according to the present invention, wherein: the feature extraction process uses point cloud processing algorithms, such as curvature analysis, normal estimation, surface reconstruction or edge detection, and performs noise reduction processing on the point cloud data to improve the accuracy of feature extraction.

[0024] As a preferred solution of the method for detecting the geometric and dimensional errors of complex components based on cross - perspective point cloud data according to the present invention, wherein: the geometric and dimensional errors include position errors, shape errors, size errors or surface flatness errors. The error calculation adopts the least - squares method, the maximum deviation method or a statistics - based method, and the design model is set as a 3D CAD model or a point cloud model for comparison with the actual point cloud data.

[0025] As a preferred solution of the method for detecting the geometric and dimensional errors of complex components based on cross - perspective point cloud data according to the present invention, wherein: a set error threshold is set, and the area exceeding the threshold is marked as a defective area. The marking method of the defective area includes color coding or highlighting. Classification is carried out according to the type of defect (such as holes, cracks, edge irregularities or surface unevenness) and severity (such as minor, medium, severe), and a quality report is generated or subsequent repair processes are guided.

[0026] As a preferred solution of the method for detecting the geometric and dimensional errors of complex components based on cross - perspective point cloud data according to the present invention, wherein: the selection of different perspectives is achieved by a pre - set scanning path or by automatically adjusting the perspective of the scanning device to ensure that all key features of the target component are covered. The planning of the scanning path is based on the geometric complexity and key feature distribution of the target component to optimize the data acquisition efficiency and coverage.

[0027] As a preferred solution of the method for detecting the geometric and dimensional errors of complex components based on cross - perspective point cloud data according to the present invention, wherein: the feature - matching - based registration algorithm further includes the screening and optimization of feature points to remove noise points and outliers and improve the robustness of registration. The registration process also includes the verification and correction of the registration result to ensure that the accuracy of the registered point cloud data set meets the detection requirements.

[0028] As a preferred solution of the method for detecting the geometric and dimensional errors of complex components based on cross - perspective point cloud data according to the present invention, wherein: the geometric feature extraction further includes the quantitative analysis of features, such as calculating parameters such as surface roughness, edge sharpness or hole size, to more comprehensively describe the geometric characteristics of the target component. The feature extraction process also includes the multi - scale analysis of features to identify geometric characteristics at different scales and improve the comprehensiveness of feature extraction.

[0029] Advantages of the present invention:

[0030] 1. In the present invention, through the acquisition of multi - perspective point cloud data, it is possible to comprehensively cover the key surfaces and feature areas of complex components, ensuring the integrity of the data. And by adopting a feature - matching - based registration algorithm and a high - precision point cloud processing algorithm, the registration accuracy and the accuracy of feature extraction are significantly improved, thereby enhancing the accuracy of geometric and dimensional error detection.

[0031] 2. The present invention supports automated processing, including automatic scanning, automatic registration, and automatic error calculation, and provides real-time feedback, facilitating quality control and process optimization during the production process. Moreover, through high-precision detection, it reduces the manual detection cost, improves the production efficiency, and simultaneously reduces the rework and scrap rates caused by quality defects. Description of the Drawings

[0032] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following-described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. Among them:

[0033] Figure 1 It is a schematic diagram of the overall process of the method for detecting the geometric and dimensional errors of complex components based on cross-view point cloud data of the present invention. Detailed Embodiments

[0034] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification.

[0035] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0036] Embodiment

[0037] Referring to Figure 1 , an embodiment of the present invention provides a method for detecting the geometric and dimensional errors of complex components based on cross-view point cloud data. This method for detecting the geometric and dimensional errors of complex components based on cross-view point cloud data includes the following steps:

[0038] Step 1: Acquisition of multi-view point cloud data

[0039] Scan the target component from different perspectives to obtain multiple point cloud data sets. The selection of different perspectives should cover the key surfaces and feature areas of the target component to ensure the comprehensiveness of the data. The acquisition devices for the point cloud data include, but are not limited to, industrial-grade 3D scanners, handheld 3D scanners, or lidars, and the resolution of the point cloud data is not less than 0.1 mm to ensure the high precision of the data.

[0040] Step 2: Registration of point cloud data

[0041] Register the multiple point cloud data sets to obtain the registered point cloud data set. The registration process includes a registration algorithm based on feature matching, extracting key feature points (such as edge points, corner points or surface texture points) in the point cloud data, and performing precise registration through the Iterative Closest Point (ICP) algorithm. The registration process also includes global registration and local registration to improve the registration accuracy and efficiency.

[0042] Step 3: Geometric feature extraction

[0043] Based on the registered point cloud data set, extract the geometric features of the target component. The geometric features include surface contour, edges, holes, depressions or texture features. The feature extraction process uses point cloud processing algorithms, such as curvature analysis, normal estimation, surface reconstruction or edge detection, and performs noise reduction processing on the point cloud data to improve the accuracy of feature extraction.

[0044] Step 4: Geometric and dimensional error calculation

[0045] Compare the extracted geometric features with the design model of the target component to calculate the geometric and dimensional errors of the target component. The geometric and dimensional errors include position error, shape error, size error or surface flatness error. The error calculation uses the least squares method, the maximum deviation method or a statistics-based method, and sets the design model as a 3D CAD model or a point cloud model for comparison with the actual point cloud data.

[0046] Step 5: Defect detection and classification

[0047] According to the calculated geometric and dimensional errors, perform defect detection and classification on the target component. Set an error threshold, and mark the area exceeding the threshold as a defect area. The marking method for the defect area includes color coding or highlighting. Classify according to the type of defect (such as holes, cracks, edge irregularities or surface unevenness) and severity (such as minor, medium, severe), and generate a quality report or guide subsequent repair processes.

[0048] The selection of different perspectives is achieved by a pre-set scanning path or automatically adjusting the perspective of the scanning device to ensure coverage of all key features of the target component. The planning of the scanning path is based on the geometric complexity and key feature distribution of the target component to optimize the data acquisition efficiency and coverage.

[0049] The registration algorithm based on feature matching also includes screening and optimization of feature points to remove noise points and outliers and improve the robustness of registration. The registration process also includes verification and correction of the registration results to ensure that the accuracy of the registered point cloud data set meets the detection requirements.

[0050] The geometric feature extraction also includes quantitative analysis of the features, such as calculating parameters such as surface roughness, edge sharpness, or hole size, to more comprehensively describe the geometric characteristics of the target component. The feature extraction process also includes multi-scale analysis of the features to identify geometric characteristics at different scales and improve the comprehensiveness of feature extraction.

[0051] The form and position error calculation also includes statistical analysis of the errors, such as calculating the mean, standard deviation, or distribution of the errors, to evaluate the overall quality level of the target component. The error calculation also includes visual display of the errors, such as generating an error heat map or a three-dimensional error model, so that users can more intuitively understand the error distribution.

[0052] The results of the defect classification can be used to generate a visual report, including an error distribution map of the three-dimensional model, a marked map of the defect area, or an error statistical chart, so that users can intuitively understand the detection results. The defect classification also includes an assessment of the reparability of the defects to provide guidance for subsequent repair processes.

[0053] The method is applicable to different types of complex components, including but not limited to mechanical parts, building components, aerospace components, or automotive components, and the corresponding detection parameters and algorithms can be adjusted according to different types of components. The adjustment includes parameter settings of the acquisition device, selection of the registration algorithm, and optimization of the error calculation method. Moreover, the method also includes automated processing of the detection process, including automatic scanning, automatic registration, automatic feature extraction, and automatic error calculation, to improve the detection efficiency and reduce manual intervention. The automated processing also includes real-time feedback of the detection results, which is used to guide quality control and process optimization in the production process and supports integration with the Manufacturing Execution System (MES).

[0054] During use, by collecting multi-viewpoint cloud data, it is possible to comprehensively cover the key surfaces and feature areas of complex components, ensuring data integrity. Moreover, by using a registration algorithm based on feature matching and a high-precision point cloud processing algorithm, the registration accuracy and the accuracy of feature extraction are significantly improved, thereby enhancing the accuracy of form and position error detection.

[0055] 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 them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for detecting geometric and position errors of complex components based on cross - perspective point cloud data, characterized in that It includes the following steps: Step 1: Multi-viewpoint cloud data acquisition Scan the target component from different viewpoints to obtain multiple point cloud data sets; Step 2: Point cloud data registration Register the multiple point cloud data sets to obtain a registered point cloud data set; Step 3: Geometric feature extraction Extract the geometric features of the target component based on the registered point cloud data set; Step 4: Geometric and dimensional error calculation Compare the extracted geometric features with the design model of the target component to calculate the geometric and dimensional errors of the target component; Step 5: Defect detection and classification Perform defect detection and classification on the target component according to the calculated geometric and dimensional errors.

2. The complex component geometric error detection method based on cross-view point cloud data according to claim 1, wherein: The acquisition device for the point cloud data includes but is not limited to industrial-grade 3D scanners, handheld 3D scanners, or lidars, and the resolution of the point cloud data is not less than 0.1 mm to ensure high-precision data.

3. The complex component geometric error detection method based on cross-view point cloud data according to claim 2, wherein: The registration process includes a registration algorithm based on feature matching, extracts key feature points in the point cloud data, and performs precise registration through the iterative closest point algorithm.

4. The method for detecting the geometric error of complex components based on cross-view point cloud data according to claim 3, wherein: The registration process also includes global registration and local registration to improve the registration accuracy and efficiency.

5. The method for detecting the geometric error of complex components based on cross-view point cloud data according to claim 4, characterized in that: The feature extraction process uses point cloud processing algorithms such as curvature analysis, normal estimation, surface reconstruction, or edge detection, and performs noise reduction processing on the point cloud data to improve the accuracy of feature extraction.

6. The method for detecting the geometric error of complex components based on cross-view point cloud data according to claim 5, characterized in that: The geometric and dimensional errors include position errors, shape errors, size errors, or surface flatness errors. The error calculation uses the least squares method, the maximum deviation method, or a statistics-based method, and the design model is set as a 3D CAD model or a point cloud model for comparison with the actual point cloud data.

7. The complex component geometric error detection method based on cross-view point cloud data according to claim 6, characterized in that: Set an error threshold, mark the area exceeding the threshold as a defect area, and the marking method for the defect area includes color coding or highlighting. Classify according to the type and severity of the defect and generate a quality report or guide subsequent repair processes.

8. The method for detecting the geometric error of complex components based on cross-view point cloud data according to claim 7, characterized in that: The selection of different viewpoints is achieved by a pre-set scanning path or automatically adjusting the viewpoint of the scanning device to ensure coverage of all key features of the target component. The scanning path is planned based on the geometric complexity and key feature distribution of the target component to optimize the data acquisition efficiency and coverage.

9. The method for detecting the geometric error of complex components based on cross-view point cloud data according to claim 8, wherein: The registration algorithm based on feature matching also includes screening and optimization of feature points to remove noise points and outliers and improve the robustness of registration. The registration process also includes verification and correction of the registration result to ensure that the accuracy of the registered point cloud data set meets the detection requirements.

10. The method for detecting the geometric error of complex components based on cross-view point cloud data according to claim 9, characterized in that: The geometric feature extraction also includes quantitative analysis of features such as calculating parameters such as surface roughness, edge sharpness, or hole size to more comprehensively describe the geometric characteristics of the target component. The feature extraction process also includes multi-scale analysis of features to identify geometric characteristics at different scales and improve the comprehensiveness of feature extraction.

Images