Aperture detection method, device and system for automobile injection molded part
By combining the ant colony algorithm and the supplementary index to optimize the aperture detection method of automotive injection molded parts, the problem of inaccurate detection caused by structural occlusion and defects of injection molded parts is solved, and efficient and accurate aperture detection is achieved.
Patent Information
- Application Number
- CN202511246357.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-02
- Publication Date
- 2025-10-10
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the existing technology of aperture detection of automobile injection molded parts, incomplete scanning caused by structural occlusion and defects of the injection molded parts themselves leads to inaccurate aperture detection results. It is difficult to distinguish between structural occlusion and injection molded part defect areas, affecting the detection accuracy.
By obtaining the local differences between the initial point cloud data and the standard CAD model, the ant colony algorithm and the supplementary index are used to optimize the supplementary scanning path. Combined with the principal curvature and the normal vector direction, supplementary scanning of the local defect area is performed to obtain the final point cloud data and perform aperture detection.
The accuracy of aperture detection is improved, the missed detection of tiny defects in traditional methods is avoided, the direction of the re-scan is ensured to match the real physical structure, the coverage and time cost of the re-scan process are optimized, and the reliability of the point cloud data is improved.
Smart Images
Figure CN120765640A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of injection molding part aperture detection, and in particular to a hole diameter detection method, device and system for an automobile injection molding part. BACKGROUND
[0002] During the injection molding process of an automobile injection molding part, material shrinkage, mold wear and the like can all cause changes in the size of the hole diameter, and the accuracy of the hole diameter directly affects the assembly accuracy with other components, and if there is a hole diameter error, it can cause assembly process failure, and cause poor electrical contact and the like, so it is necessary to detect the hole diameter of the automobile injection molding part. In the existing method, three-dimensional scanning technology is usually used to scan the automobile injection molding part to obtain three-dimensional data of the surface of the automobile injection molding part, but during scanning, due to the blocking of the structure of the injection molding part itself and the like, the three-dimensional structure of the injection molding part obtained can be missing, and in the actual detection process, the missing area usually needs to be supplemented by scanning, and the results of multiple scans are spliced to obtain the complete three-dimensional structure of the automobile injection molding part, and the hole diameter is detected.
[0003] During the supplementary scanning, it is difficult to distinguish between fixed defects caused by defects in the injection molding part itself and temporary defects caused by structural blocking, which can cause over-scan of the defect area of the injection molding part itself or premature termination of the scan, resulting in inaccurate injection molding part structure and affecting the accuracy of the hole diameter detection result. SUMMARY
[0004] The present application provides a hole diameter detection method, device and system for an automobile injection molding part to solve the existing problems.
[0005] The hole diameter detection method, device and system for an automobile injection molding part of the present application adopt the following technical solutions: One embodiment of the present application provides a hole diameter detection method for an automobile injection molding part, which comprises the following steps: obtaining initial point cloud data of the automobile injection molding part and a standard CAD model; obtaining a local difference between the data points in the initial point cloud data and the standard CAD model, merging adjacent data points based on the size of the local difference to obtain a plurality of local defect areas, obtaining a supplementary index of the local defect area using the local difference and the spatial distribution of the data points in the local defect area, determining a supplementary scanning direction for the local defect area based on the principal curvature direction of the data points in the local defect area and the normal vector of the corresponding local area in the standard CAD model, and obtaining a supplementary scanning path for the local defect area using the supplementary index and combining an ant colony algorithm; The local defect area is scanned by the scanning direction and scanning path, and the structural defect area and the final point cloud data including the structural defect area are obtained by the change of the local defect area before and after the scanning; The final point cloud data is visualized and used for aperture detection.
[0006] Furthermore, the specific method of obtaining the local differences between the data points in the initial point cloud data and the standard CAD model includes: Converting a standard CAD model of an automobile injection molded part into an STL format, obtaining a plurality of triangular facets of the standard CAD model of the automobile injection molded part in the STL format, and matching the standard CAD model in the STL format with initial point cloud data of the automobile injection molded part using an ICP algorithm; After matching the standard CAD model in STL format with the initial point cloud data of the automotive injection molded part, a number of triangular facets within a preset search radius between any data point in the initial point cloud data and the standard CAD model are obtained, and recorded as the facets to be tested of the data point; the normal vector of the facet to be tested is obtained, and based on the cosine value between the normal vector of the data point in the initial point cloud data and the normal vector of the facet to be tested, as well as the minimum distance between the data point and the corresponding facet to be tested, a difference index between the data point and the facet to be tested is obtained. The difference index is used to describe the local difference between the data point in the initial point cloud data and the standard CAD model.
[0007] Furthermore, the method of merging adjacent data points based on the size of the local difference to obtain a number of local defect areas includes: The surface patch to be tested corresponding to the data point when the difference index is minimum is used as the matching surface patch of the data point. The minimum value of all difference indices between each data point and all corresponding surface patches to be tested is obtained and recorded as the target difference index of the data point. The data point corresponding to the target difference index is greater than or equal to the preset difference threshold and is recorded as the matching difference point. The adjacent matching difference points in the initial point cloud data are merged to obtain several areas, which are recorded as local defect areas.
[0008] Furthermore, the method of obtaining a supplementary index of the local defect area by utilizing the local difference and the spatial distribution of data points in the local defect area includes the following specific methods: For any local defect area, the number of voxels contained in the local defect area, the average number of data points in all voxels, and the average target difference index of all data points in the target point cloud corresponding to the local defect area are obtained to obtain the supplementary index of the local defect area, wherein the number of voxels and the average target difference index are positively correlated with the supplementary index, and the average number of data points in all voxels is negatively correlated with the supplementary index.
[0009] Furthermore, the method of determining the scanning direction of the local defect area based on the principal curvature direction of the data points in the local defect area and the normal vector of the local area corresponding to the standard CAD model includes the following specific methods: The maximum principal curvature of each data point in any local defect area is obtained, and the maximum principal curvature direction of each data point is projected onto a two-dimensional plane. The direction of the two-dimensional plane is evenly divided into several angle intervals. All data points projected on the two-dimensional plane are traversed, and the interval with the largest number of data points is recorded as the principal curvature interval; the sum vector of the principal curvature direction of all data points in the principal curvature interval and the normal vector of the corresponding matching patch of each data point in the standard CAD model is recorded as the scanning direction of the local defect area.
[0010] Furthermore, the method of using the supplementary index in combination with the ant colony algorithm to obtain the scanning path for the local defect area includes the following specific methods: The center of gravity of each local defect area is projected onto the initial scanning plane, and its projected coordinates on the scanning plane are used as the nodes in the ant colony algorithm. The initial pheromone concentration of the ant colony algorithm is initialized and adjusted according to the supplementary index of each local defect area. The current position of the scanning probe is used as the starting point of the ant colony algorithm, and the projected coordinates of the local defect area with the smallest scanning priority index are used as the end point, thereby adjusting the initial pheromone concentration of the node corresponding to each local defect area. The adjusted initial pheromone concentration of each node is input into the ant colony algorithm to obtain the supplementary scanning path of all local defect areas.
[0011] Furthermore, the structural defect area and the final point cloud data including the structural defect area are obtained by analyzing the changes of the local defect area before and after the rescan, including the specific method of: According to the overlapping conditions of the local defect area before and after the supplementary scan, the supplementary scan matching index of the local defect area is obtained; According to the change of curvature distribution of local defect areas before and after rescanning and the overall level difference of target difference index of data points, combined with rescanning matching index, the structural defect index of local defect areas is obtained, thereby screening out the final point cloud data containing several structural defect areas.
[0012] Furthermore, the specific method of obtaining the structural defect index of the local defect area includes: For any local defect area, obtain the standard deviation of the principal curvature corresponding to all boundary points of the local defect area before the re-scanning, and record it as the front curvature parameter of the local defect area. Record the standard deviation of the principal curvature corresponding to all boundary points of the local defect area after the re-scanning as the back curvature parameter of the local defect area. The absolute value of the difference between the front curvature parameter and the back curvature parameter is recorded as the curvature change rate of the local defect area. The structural defect index of any local defect area is obtained based on the rescanning matching index, curvature change rate, and distribution difference of the target difference index of the data points included in the local defect area before and after the rescanning. The rescanning matching index is positively correlated with the structural defect index, and the curvature change rate and the distribution difference of the target difference index of the data points included in the local defect area before and after the rescanning are negatively correlated with the structural defect index.
[0013] A device for detecting the aperture of an automobile injection molded part comprises a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the device implements any one of the steps of a method for detecting the aperture of an automobile injection molded part.
[0014] A system for detecting the aperture of an automobile injection molded part adopts any one of the methods for detecting the aperture of an automobile injection molded part, and the system includes the following modules: Data acquisition module: used to obtain initial point cloud data and standard CAD models of automotive injection molded parts; The patching planning module is used to obtain local differences between data points in the initial point cloud data and the standard CAD model, merge adjacent data points based on the size of the local differences to obtain several local defect areas; use the local differences and the spatial distribution of data points in the local defect areas to obtain a supplementary index for the local defect areas; determine the patching direction for the local defect areas based on the principal curvature direction of the data points in the local defect areas and the normal vector of the corresponding local area in the standard CAD model; and use the supplementary index in combination with the ant colony algorithm to obtain the patching path for the local defect areas; Defect recognition module: used to scan the local defect area through the scanning direction and scanning path, and obtain the structural defect area and the final point cloud data containing the structural defect area based on the changes in the local defect area before and after the scanning; Aperture detection module: used to visualize the final point cloud data and perform aperture detection using the final point cloud data.
[0015] The beneficial effects of the technical solution of the present invention are: dynamically merging point cloud data through local differences between initial point cloud data and standard CAD models, converting discrete data points into defect areas that can be quantified, avoiding the missed detection of minor defects by traditional methods; by introducing the concept of supplementary index, integrating the dual indicators of spatial distribution and degree of difference, making the re-scanning priority determination more in line with actual engineering needs, and improving the re-scanning effect of automotive injection molded parts; in addition, combining the direction of the principal curvature and the normal vector of the CAD model, ensuring the matching of the re-scanning direction with the real physical structure, and combining the path optimization of the ant colony algorithm so that the re-scanning process takes into account both coverage and time cost, avoiding the problem of inaccurate aperture detection structure caused by missing data in the point cloud data due to unreasonable scanning direction when performing aperture detection on automotive injection molded parts, improving the reliability of the point cloud data of automotive injection molded parts, and thus improving the accuracy of subsequent aperture detection of automotive injection molded parts. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0017] Figure 1 A flowchart of a method for detecting apertures of automotive injection molded parts according to the present invention; Figure 2 The figure is a structural block diagram of an aperture detection system for automobile injection molded parts according to the present invention. DETAILED DESCRIPTION
[0018] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a method, device, and system for detecting the aperture of automotive injection molded parts, including its specific implementation, structure, features, and effectiveness. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
[0019] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0020] The following describes in detail a method, device and system for detecting the aperture of automobile injection molded parts provided by the present invention with reference to the accompanying drawings.
[0021] See also Figure 1, which shows a flowchart of a method for detecting the aperture of an automobile injection molded part provided by one embodiment of the present invention, the method comprising the following steps: Step S001: Acquire initial point cloud data and a standard CAD model of an automobile injection molded part.
[0022] It should be noted that, since the outer contours of automobile injection molded parts are usually variable and irregular, when using 3D scanning technology to obtain their 3D models, when performing aperture detection on the injection molded parts, the structural occlusion caused by the influence of the outer contour makes it impossible to effectively distinguish between scan omissions and defects in the injection molded parts themselves, resulting in inaccurate aperture detection results of the injection molded parts. Therefore, in order to improve the aperture detection results of the injection molded parts, it is necessary to perform supplementary scanning on them. During the supplementary scanning process, in order to improve the scanning effect and ensure that the local defective areas in the 3D scanning results of automobile injection molded parts can be accurately identified as temporary structural omission areas caused by structural occlusion in the final scanning results, the embodiment of the present invention chooses to plan the number and path of supplementary scanning.
[0023] Specifically, in order to implement the aperture detection method for automobile injection molded parts proposed in this embodiment, it is first necessary to collect point cloud data of the automobile injection molded parts. The specific process is as follows: Automotive injection molded parts often contain complex features such as deep holes, irregular-shaped holes, and densely distributed holes. When inspecting their pore sizes, if only local spot checks are performed, hidden holes or tiny pore defects may be missed. Therefore, 3D scanning of the entire injection molded part is usually required: First, obtain the standard CAD model of the automotive injection molded part.
[0024] Then, the automotive injection molded parts are fixed with a fixture to ensure that the scanning area is unobstructed. In addition, matte developer is sprayed on the highly reflective areas of the automotive injection molded parts (such as chrome holes).
[0025] Finally, a blue light 3D scanner was selected as the 3D scanning device, and a 3D scanning device with its own mobile light source was used to perform a single scan at an angle that can cover the largest feature surface of the automotive injection molded parts to obtain the initial point cloud data of the automotive injection molded parts in 3D space.
[0026] It should be noted that in the embodiment of the present invention, the area directly above the injection molded part is selected as the maximum feature surface, and vertical scanning is performed. The specific scanning angle is determined according to the placement angle of the maximum feature surface of the automobile injection molded part, and the embodiment of the present invention does not make any specific limitations.
[0027] Finally, the initial point cloud data obtained by scanning is Gaussian filtered and transmitted and stored via a wireless network or wired connection.
[0028] At this point, the standard CAD model and initial point cloud data of the automobile injection molded parts are obtained through the above method.
[0029] Step S002: Obtain local differences between data points in the initial point cloud data and the standard CAD model, and merge adjacent data points based on the size of the local differences to obtain several local defect areas; use the local differences and the spatial distribution of data points in the local defect area to obtain a supplementary index for the local defect area; determine the direction of the fill scan for the local defect area based on the principal curvature direction of the data points in the local defect area and the normal vector of the corresponding local area in the standard CAD model, and use the supplementary index in combination with the ant colony algorithm to obtain a fill scan path for the local defect area.
[0030] It should be noted that the point cloud data obtained from a single 3D scan of an automotive injection molded part may contain local defect areas. These local defect areas arise from local scan omissions caused by structural defects in the automotive injection molded part itself and temporary omissions during the scanning process. Structural defects in automotive injection molded parts are typically caused by temperature (e.g., melt temperature fluctuations and differences in cooling rates in different mold regions) and stress (e.g., injection pressure fluctuations), which lead to aperture deformation and local defects in the 3D scan results of the deformed areas. Temporary structural omissions during the scanning process are typically due to the limited coverage of a single scan, where the size of the automotive injection molded part exceeds the scanning range, or due to the complex structure of the automotive injection molded part, which results in structural obstruction during a single scan and causes scanning blind spots. Furthermore, since automotive injection molded parts are typically injection molded strictly according to a preset CAD model, i.e., the CAD model of the automotive injection molded part is known data, in order to accurately understand the specific abnormality of the automotive injection molded part, the embodiment of the present invention selects the difference between the point cloud data of the automotive injection molded part obtained from a single scan and the standard CAD model of the automotive injection molded part to obtain the local defect area.
[0031] Specifically, in step S201 , local differences between data points in the initial point cloud data and the standard CAD model are obtained, and adjacent data points are merged based on the magnitude of the local differences to obtain a number of local defect regions.
[0032] As a preferred embodiment, the method for obtaining a local defect area includes: First, the standard CAD model of the automobile injection molded part is converted into STL format, and several triangular facets of the standard CAD model of the automobile injection molded part in STL format are obtained. The standard CAD model in STL format is matched with the initial point cloud data of the automobile injection molded part using the ICP algorithm.
[0033] Then, after matching the standard CAD model in STL format with the initial point cloud data of the automotive injection molded part, a number of triangular facets within a preset search radius between any data point in the initial point cloud data and the standard CAD model are obtained, and recorded as the facets to be tested of the data point; the normal vector of the facet to be tested is obtained, and the difference index between the data point and the facet to be tested is obtained based on the cosine value between the normal vector of the data point in the initial point cloud data and the normal vector of the facet to be tested, as well as the minimum distance between the data point and the corresponding facet to be tested.
[0034] It should be noted that the search radius of the adjacent triangles in this solution is set to 3 times the voxels of the current target point cloud.
[0035] As an optional embodiment, the specific calculation method of the difference index is: , where Indicates the point cloud data The data point and the corresponding The difference index of the tested patches; Indicates the point cloud data The data point and the corresponding The minimum distance between the tested patches; Indicates the point cloud data The normal vector of each data point; Indicates the point cloud data The data point corresponding to The normal vector of the face to be tested, represents the cosine function; represents the linear normalization function.
[0036] It should be noted that the difference index is used to describe the local difference between the data points in the initial point cloud data and the standard CAD model, reflecting the degree of structural difference between the data points in the point cloud data and the corresponding positions in the standard CAD model. If the minimum distance between the data point in the point cloud data and the surface to be tested is farther, and the normal vector difference between the two is greater, the data point will be less consistent with the structural contour of the injection molded part at the corresponding surface to be tested.
[0037] Finally, the surface patch to be tested corresponding to the data point when the difference index is minimum is used as the matching surface patch of the data point, and the minimum value of all difference indices between each data point and all corresponding surface patches to be tested is obtained, which is recorded as the target difference index of the data point. The data point corresponding to the target difference index is greater than or equal to the preset difference threshold and is recorded as the matching difference point. The adjacent matching difference points in the initial point cloud data are merged to obtain several areas, which are recorded as local defect areas.
[0038] It should be noted that, in the embodiment of the present invention, the difference threshold is preset to 0.6 based on experience, and can be adjusted according to actual conditions, and is not specifically limited in the embodiment of the present invention.
[0039] In step S202, a supplementary index of the local defect area is obtained by using the local difference and the spatial distribution of the data points in the local defect area; a supplementary scanning direction of the local defect area is determined based on the main curvature direction of the data points in the local defect area and the normal vector of the corresponding local area in the standard CAD model, and a supplementary scanning path for the local defect area is obtained by using the supplementary index in combination with the ant colony algorithm.
[0040] It should be noted that for some local defect areas in the initial point cloud data, supplementary scanning is required to prevent the local defect areas from representing temporary structural defects in the automotive injection molded part. Based on the above analysis, to avoid the formation of temporary defect areas due to factors such as occlusion in the injection molded part's structure, the supplementary scanning angle needs to be adjusted based on the specific shape of the defect area. Specifically, if a local defect area occupies a large space and has a low point cloud density (the likelihood of incomplete injection molded part point cloud acquisition in this local defect area is greater), and the greater the difference between the points in the local defect area and the standard CAD, the more necessary it is to perform supplementary scanning of this local defect area.
[0041] First, a supplementary index of the local defect region is obtained by using the quantity distribution of data points in a plurality of voxels and the target difference index of the data points in the local defect region.
[0042] As a preferred embodiment, the method for obtaining the supplementary index includes: For any local defect area, the number of voxels contained in the local defect area, the average number of data points in all voxels, and the average target difference index of all data points in the target point cloud corresponding to the local defect area are obtained to obtain the supplementary index of the local defect area, wherein the number of voxels and the average target difference index are positively correlated with the supplementary index, and the average number of data points in all voxels is negatively correlated with the supplementary index.
[0043] As an optional embodiment, the specific calculation method of the supplementary index is: ,in, Indicates the A supplementary index for a local defect area; Indicates the The number of voxels contained in a local defect area; Indicates the The average target difference index of all data points in the target point cloud corresponding to the local defect area; No. The average number of data points in all voxels contained in a local defect area; is the sigmoid normalization function.
[0044] It should be noted that the supplementary index is used to describe the priority of supplementary scanning in the corresponding local defect area. The larger the value of the supplementary index, the higher the priority of supplementary scanning; the larger the value of the voxel number, the larger the space occupied by the local defect area; the smaller the mean number of data points in the voxel, the smaller the point cloud space density in the corresponding local defect area.
[0045] Then, the complementary scanning direction of any local defect area is determined.
[0046] It should be noted that for a single local defect area, the scanning direction of each local defect area is adjusted based on the main curvature direction with the highest coverage of the local defect area (this direction reflects the geometric deformation trend of the defect area) and the normal vector direction of multiple triangular facets in the CAD model corresponding to the local defect area (this direction reflects the theoretical orientation of the automobile injection molding structure corresponding to the area) to ensure that the scanning incident angle is aligned with the defect geometric features and reduce the projection blind area caused by angle deviation.
[0047] As an optional embodiment, the method for obtaining the supplementary scanning direction is: The maximum principal curvature of each data point in any local defect area is obtained, and the maximum principal curvature direction of each data point is projected onto a two-dimensional plane. The direction of the two-dimensional plane is evenly divided into several angle intervals. All data points projected on the two-dimensional plane are traversed, and the interval with the largest number of data points is recorded as the principal curvature interval; the sum vector of the principal curvature direction of all data points in the principal curvature interval and the normal vector of the corresponding matching patch of each data point in the standard CAD model is recorded as the scanning direction of the local defect area.
[0048] It should be noted that if the scanning incident direction exceeds the maximum rotation direction of the scanning device during scanning, the maximum rotation direction of the scanning device will be used as the scanning incident direction; in addition, in the embodiment of the present invention, the preset angle interval is divided at an interval of 10°, which can be adjusted according to actual conditions, and the embodiment of the present invention does not make specific limitations.
[0049] Finally, based on the priority index of the supplementary scans of all local defect areas obtained after the initial scan, the ant colony algorithm is used to plan the supplementary scanning path of the scanning equipment.
[0050] As a preferred embodiment, the specific process of the supplementary sweep path planning includes: The center of gravity of each local defect area is projected onto the initial scanning plane, and its projected coordinates on the scanning plane are used as the nodes in the ant colony algorithm. The initial pheromone concentration of the ant colony algorithm is initialized and adjusted according to the supplementary index of each local defect area. The current position of the scanning probe is used as the starting point of the ant colony algorithm, and the projected coordinates of the local defect area with the smallest scanning priority index are used as the end point, thereby adjusting the initial pheromone concentration of the node corresponding to each local defect area. The adjusted initial pheromone concentration of each node is input into the ant colony algorithm to obtain the supplementary scanning path of all local defect areas.
[0051] As an optional embodiment, the specific calculation method of the adjusted initial pheromone concentration is: ,in, represents the initial pheromone concentration of the node corresponding to the jth region after adjustment, is the initial pheromone concentration of the node before adjustment; Indicates the The supplementary index of a local defect area.
[0052] It should be noted that, in the embodiment of the present invention, the initial pheromone concentration of the nodes corresponding to each local defect area before adjustment is initialized to 1, which can be set according to actual conditions and is not specifically limited in the embodiment of the present invention; in order to make the scanning device give priority to covering high-priority areas when performing supplementary scanning on local defect areas of automobile injection molded parts, the larger the priority index of a certain area, the greater the corresponding initial pheromone concentration.
[0053] At this point, the scanning paths for all local defect areas are obtained through the above method.
[0054] Step S003: re-scan the local defect area using the re-scanning direction and re-scanning path, and obtain the structural defect area and the final point cloud data including the structural defect area based on the changes in the local defect area before and after the re-scanning.
[0055] It should be noted that the local defect regions obtained in the initial scan may be defects in the injection molded part itself or temporary structural defects caused by scanning blind spots. Temporary structural defects can be corrected using supplemental scans after changing the angle. This means that the morphology of the region before and after the supplemental scan differs significantly, while the morphology of local defect regions caused by intrinsic structural defects is relatively stable. Therefore, by comparing the changes in the local defect regions before and after the supplemental scan, it is possible to distinguish between intrinsic structural defects and temporary structural defects in automotive injection molded parts. The local defect regions obtained before and after the supplemental scan are matched. Local defect regions corresponding to temporary structural defects in automotive injection molded parts may disappear or shrink after the supplemental scan, or they may form new scanning blind spots, causing the local defect regions to actually increase in size. However, local defect regions caused by intrinsic structural defects in the injection molded part are relatively stable in morphology. Therefore, if any local defect region has a high overlap ratio before and after the supplemental scan, minimal positional change, and minimal change in the standard deviation of the principal curvature at each point on the region boundary (the region's structural morphology is relatively stable), and the difference between each point and the standard CAD model is relatively small (the correlation between the current defect and the designed structure is relatively stable), then the region is more likely to be a structural defect in the automotive injection molded part, and a rescan is less necessary.
[0056] Specifically, in step S301, a supplementary scanning matching index of the local defect area is obtained according to the overlapping conditions of the local defect area before and after the supplementary scanning.
[0057] First, several local defect areas in the supplementary scanning results are obtained, and the several local defect areas in the supplementary scanning results and the local defect areas in the initial point cloud data are mapped to the surface of the standard CAD model by normal projection.
[0058] Then, any local defect area on the surface of the standard CAD model belonging to the initial point cloud data is recorded as the target local defect area, and the overlapping area between the target local defect area and any local defect area after re-scanning on the surface of the standard CAD model, as well as the Euclidean distance between the coordinates of the center of gravity of the area are obtained to obtain the matching index between the target local defect area and the local defect area after re-scanning, wherein the overlapping area is positively correlated with the matching index, and the Euclidean distance is negatively correlated with the matching index.
[0059] As an optional embodiment, the specific calculation method of the matching index is: ,in, Indicates the target local defect area and the Matching index between the local defect areas after rescanning; Indicates the target local defect area and the The overlapping area of the local defect area after the area is scanned on the surface of the standard CAD model; Indicates the target local defect area and the The Euclidean distance between the coordinates of the center of gravity of the local defect area after the area is scanned on the surface of the standard CAD model.
[0060] It should be noted that the matching index is used to describe the degree of matching correspondence between the local defect area and the local defect area after rescanning.
[0061] Finally, the maximum value of the matching index between the target local defect area and all the local defect areas after rescanning is recorded as the rescanning matching index of the target local defect area.
[0062] In step S302, the structural defect index of the local defect area is obtained based on the change in the curvature distribution of the local defect area before and after the rescan and the overall level difference of the target difference index of the data points, combined with the rescan matching index, so as to screen out the final point cloud data containing several structural defect areas.
[0063] First, the structural defect index of any local defect area is obtained.
[0064] As a preferred embodiment, the method for obtaining the structural defect index of the local defect area includes: 1) For any local defect area, obtain the standard deviation of the principal curvature corresponding to all boundary points of the local defect area before rescanning, and record it as the front curvature parameter of the local defect area; record the standard deviation of the principal curvature corresponding to all boundary points of the local defect area after rescanning as the back curvature parameter of the local defect area; and record the absolute value of the difference between the front curvature parameter and the back curvature parameter as the curvature change rate of the local defect area.
[0065] 2) The structural defect index of any local defect area is obtained based on the scanning matching index, curvature change rate, and the distribution difference of the target difference index of the data points included in the local defect area before and after the scanning. The scanning matching index is positively correlated with the structural defect index, and the curvature change rate and the distribution difference of the target difference index of the data points included in the local defect area before and after the scanning are negatively correlated with the structural defect index.
[0066] As an optional embodiment, the specific calculation method of the structural defect index is: , where Indicates the Structural defect index of a local defect area; Indicates the The scan matching index of the local defect area; Indicates the The curvature change rate of a local defect area; Indicates the The average target difference index of all data points in the initial target point cloud for the local defect area, Indicates the The average target difference index of all data points in the local defect area after rescanning; represents the absolute value function of difference; represents the linear normalization function; Indicates the first parameter of the preset.
[0067] Then, the local defect area corresponding to the structural defect index greater than the preset second parameter is recorded as the structural defect area. The structural defect area is no longer scanned. The scanning angle and scanning path of the local defect area outside the structural defect area are re-acquired to obtain the final point cloud data of the automobile injection molded part.
[0068] It should be noted that the first parameter is used to avoid the situation where the denominator of the formula is 0. In the embodiment of the present invention, the first parameter is preset to 0.01, and the second parameter is preset to 0.6 based on experience. The values of the first parameter and the second parameter can be adjusted according to actual conditions, and the embodiment of the present invention does not make specific limitations.
[0069] It should be noted that if the number of re-scans of any local defective area exceeds a preset threshold during the re-scanning process, the local defective area will be visually marked and a warning message will be sent to facilitate timely manual operation to stop the scan.
[0070] At this point, the final point cloud data of the automobile injection molded parts is obtained through the above method.
[0071] Step S004: Visualize the final point cloud data and use the final point cloud data to perform aperture detection.
[0072] Specifically, first, the final point cloud data is visualized and the structural defect areas in the final point cloud data are marked.
[0073] Then, in the final point cloud data, all point cloud subsets corresponding to the hole structures (such as screw holes and positioning holes) in the standard CAD model are extracted. The center coordinates of the holes in the standard CAD model are projected into the point cloud space. A cylindrical ROI (region of interest) is constructed with the projection point as the center and a preset radius. For each point cloud in the ROI, the cylindrical model is fitted using the random sampling consensus (RANSAC) algorithm. If the deviation between the fitted cylindrical radius and the standard CAD aperture is greater than the preset aperture deviation threshold, it is marked as a deformed hole.
[0074] It should be noted that in the embodiment of the present invention, the preset radius is set to the corresponding aperture size in the standard CAD model plus 3 times the point cloud density. In addition, the aperture deviation threshold is preset to 5%. The radius and aperture deviation threshold can be adjusted according to actual conditions, and the embodiment of the present invention does not make specific limitations.
[0075] Through the above steps, the aperture detection of automobile injection molded parts is completed.
[0076] In another embodiment of the invention, a device for detecting the aperture of an automobile injection-molded part is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of any one of the methods for detecting the aperture of an automobile injection-molded part are implemented.
[0077] The present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing related hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above-mentioned various method embodiments. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.
[0078] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or any conventional processor. It is worth noting that the processor may be a processor that supports the Advanced RISC Machines (ARM) architecture.
[0079] See also Figure 2, which shows a structural block diagram of an aperture detection system for automotive injection molded parts provided by one embodiment of the present invention; the aperture detection system for automotive injection molded parts implements steps S001 to S004 of an aperture detection method for automotive injection molded parts, and the system includes the following modules: Data acquisition module: used to obtain initial point cloud data and standard CAD models of automotive injection molded parts; The patching planning module is used to obtain local differences between data points in the initial point cloud data and the standard CAD model, merge adjacent data points based on the size of the local differences to obtain several local defect areas; use the local differences and the spatial distribution of data points in the local defect areas to obtain a supplementary index for the local defect areas; determine the patching direction for the local defect areas based on the principal curvature direction of the data points in the local defect areas and the normal vector of the corresponding local area in the standard CAD model; and use the supplementary index in combination with the ant colony algorithm to obtain the patching path for the local defect areas; Defect recognition module: used to scan the local defect area through the scanning direction and scanning path, and obtain the structural defect area and the final point cloud data containing the structural defect area based on the changes in the local defect area before and after the scanning; Aperture detection module: used to visualize the final point cloud data and perform aperture detection using the final point cloud data.
[0080] This embodiment dynamically merges point cloud data through local differences between the initial point cloud data and the standard CAD model, converts discrete data points into defect areas that can be quantified, and avoids the traditional method of missing small defects. By introducing the concept of supplementary index and integrating the dual indicators of spatial distribution and degree of difference, the repair priority determination is more in line with actual engineering needs, thereby improving the repair scanning effect of automotive injection molded parts. In addition, the combination of the main curvature and the direction of the CAD model normal vector ensures the matching of the repair scanning direction with the real physical structure. The path optimization of the ant colony algorithm is combined to ensure that the repair scanning process takes into account both coverage and time cost, thereby avoiding the problem of inaccurate aperture detection structure caused by missing data in the point cloud data due to unreasonable scanning direction when performing aperture detection on automotive injection molded parts. The reliability of the point cloud data of automotive injection molded parts is improved, thereby improving the accuracy of subsequent aperture detection of automotive injection molded parts.
[0081] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for detecting the aperture of an automobile injection molded part, characterized in that: The method comprises the following steps: Obtain initial point cloud data and standard CAD models of automotive injection molded parts; Obtaining local differences between data points in the initial point cloud data and the standard CAD model, merging adjacent data points based on the magnitude of the local differences to obtain several local defect regions; utilizing the local differences and the spatial distribution of data points in the local defect regions to obtain a supplementary index for the local defect regions; determining a patching direction for the local defect regions based on the principal curvature direction of the data points in the local defect regions and the normal vector of the corresponding local region in the standard CAD model; and utilizing the supplementary index in combination with an ant colony algorithm to obtain a patching path for the local defect regions; The local defect area is scanned by the scanning direction and scanning path, and the structural defect area and the final point cloud data including the structural defect area are obtained by the change of the local defect area before and after the scanning; The final point cloud data is visualized and used for aperture detection.
2. The method for detecting the aperture of an automobile injection molded part according to claim 1, wherein: The specific method of obtaining the local differences between the data points in the initial point cloud data and the standard CAD model includes: Converting a standard CAD model of an automobile injection molded part into an STL format, obtaining a plurality of triangular facets of the standard CAD model of the automobile injection molded part in the STL format, and matching the standard CAD model in the STL format with initial point cloud data of the automobile injection molded part using an ICP algorithm; After matching the standard CAD model in STL format with the initial point cloud data of the automotive injection molded part, a number of triangular facets within a preset search radius between any data point in the initial point cloud data and the standard CAD model are obtained, and recorded as the facets to be tested of the data point; the normal vector of the facet to be tested is obtained, and based on the cosine value between the normal vector of the data point in the initial point cloud data and the normal vector of the facet to be tested, as well as the minimum distance between the data point and the corresponding facet to be tested, a difference index between the data point and the facet to be tested is obtained. The difference index is used to describe the local difference between the data point in the initial point cloud data and the standard CAD model.
3. The method for detecting the aperture of an automobile injection molded part according to claim 2, wherein: The specific method of merging adjacent data points based on the size of the local difference to obtain a number of local defect areas includes: The surface patch to be tested corresponding to the data point when the difference index is minimum is used as the matching surface patch of the data point. The minimum value of all difference indices between each data point and all corresponding surface patches to be tested is obtained and recorded as the target difference index of the data point. The data point corresponding to the target difference index is greater than or equal to the preset difference threshold and is recorded as the matching difference point. The adjacent matching difference points in the initial point cloud data are merged to obtain several areas, which are recorded as local defect areas.
4. The method for detecting the aperture of an automobile injection molded part according to claim 3, wherein: The method of obtaining the supplementary index of the local defect area by utilizing the local difference and the spatial distribution of the data points in the local defect area includes: For any local defect area, the number of voxels contained in the local defect area, the average number of data points in all voxels, and the average target difference index of all data points in the target point cloud corresponding to the local defect area are obtained to obtain the supplementary index of the local defect area, wherein the number of voxels and the average target difference index are positively correlated with the supplementary index, and the average number of data points in all voxels is negatively correlated with the supplementary index.
5. The method for detecting the aperture of an automobile injection molded part according to claim 1, wherein: The method of determining the scanning direction of the local defect area based on the principal curvature direction of the data points in the local defect area and the normal vector of the local area corresponding to the standard CAD model includes the following specific methods: Obtain the maximum principal curvature of each data point in any local defect area, and project the maximum principal curvature direction of each data point onto a two-dimensional plane. Divide the direction of the two-dimensional plane into several angle intervals, traverse all data points projected on the two-dimensional plane, and record the interval with the largest number of data points as the principal curvature interval; The sum vector of the principal curvature directions of all data points within the principal curvature interval and the normal vectors of the corresponding matching patches of each data point in the standard CAD model is recorded as the scanning direction for the local defect area.
6. The method for detecting the aperture of an automobile injection molded part according to claim 1, wherein: The method of using the supplementary index in combination with the ant colony algorithm to obtain the scanning path for the local defect area includes the following specific methods: The center of gravity of each local defect area is projected onto the initial scanning plane, and its projected coordinates on the scanning plane are used as the nodes in the ant colony algorithm. The initial pheromone concentration of the ant colony algorithm is initialized and adjusted according to the supplementary index of each local defect area. The current position of the scanning probe is used as the starting point of the ant colony algorithm, and the projected coordinates of the local defect area with the smallest scanning priority index are used as the end point, thereby adjusting the initial pheromone concentration of the node corresponding to each local defect area. The adjusted initial pheromone concentration of each node is input into the ant colony algorithm to obtain the supplementary scanning path of all local defect areas.
7. The method for detecting the aperture of an automobile injection molded part according to claim 1, wherein: The method of obtaining the structural defect area and the final point cloud data including the structural defect area by scanning the changes of the local defect area before and after the rescan is as follows: According to the overlapping conditions of the local defect area before and after the supplementary scan, the supplementary scan matching index of the local defect area is obtained; According to the change of curvature distribution of local defect areas before and after rescanning and the overall level difference of target difference index of data points, combined with rescanning matching index, the structural defect index of local defect areas is obtained, thereby screening out the final point cloud data containing several structural defect areas.
8. A method for detecting aperture of automobile injection molded parts according to claim 7, characterized in that: The specific method of obtaining the structural defect index of the local defect area includes: For any local defect area, obtain the standard deviation of the principal curvature corresponding to all boundary points of the local defect area before the re-scanning, and record it as the front curvature parameter of the local defect area. Record the standard deviation of the principal curvature corresponding to all boundary points of the local defect area after the re-scanning as the back curvature parameter of the local defect area. The absolute value of the difference between the front curvature parameter and the back curvature parameter is recorded as the curvature change rate of the local defect area. The structural defect index of any local defect area is obtained based on the rescanning matching index, curvature change rate, and distribution difference of the target difference index of the data points included in the local defect area before and after the rescanning. The rescanning matching index is positively correlated with the structural defect index, and the curvature change rate and the distribution difference of the target difference index of the data points included in the local defect area before and after the rescanning are negatively correlated with the structural defect index.
9. A device for detecting the aperture of an automobile injection molded part, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for detecting the aperture of an automobile injection molded part according to any one of claims 1 to 8 are implemented.
10. A system for detecting the aperture of automobile injection molded parts, using the method for detecting the aperture of automobile injection molded parts according to any one of claims 1 to 8, characterized in that: The system includes the following modules: Data acquisition module: used to obtain initial point cloud data and standard CAD models of automotive injection molded parts; The patching planning module is used to obtain local differences between data points in the initial point cloud data and the standard CAD model, merge adjacent data points based on the size of the local differences to obtain several local defect areas; use the local differences and the spatial distribution of data points in the local defect areas to obtain a supplementary index for the local defect areas; determine the patching direction for the local defect areas based on the principal curvature direction of the data points in the local defect areas and the normal vector of the corresponding local area in the standard CAD model; and use the supplementary index in combination with the ant colony algorithm to obtain the patching path for the local defect areas; Defect recognition module: used to scan the local defect area through the scanning direction and scanning path, and obtain the structural defect area and the final point cloud data containing the structural defect area based on the changes in the local defect area before and after the scanning; Aperture detection module: used to visualize the final point cloud data and perform aperture detection using the final point cloud data.