Quality detection method for forming of automobile tail lamp mask

Abstract

The invention relates to the field of image processing, in particular to a quality detection method for forming an automobile tail lamp mask, and the method comprises the steps: collecting and preprocessing three-dimensional point cloud data of the automobile tail lamp mask; calculating the tail cover structure support degree of each three-dimensional point cloud; calculating the local curvature of each three-dimensional point cloud; and calculating a tail cover structure significance index of each three-dimensional point cloud by combining the tail cover structure support degree and the local curvature, screening the three-dimensional point cloud data of the automobile tail lamp mask based on the calculated tail cover structure significance index, and matching based on the screened three-dimensional point cloud data to realize the molding quality detection of the automobile tail lamp mask. According to the method, the three-dimensional point cloud data is screened based on the significant index of the tail cover structure, the key feature points are reserved, and the forming quality of the tail lamp mask can be rapidly and accurately detected.

Description

Technical Field

[0001] The present invention relates to the field of image processing. More specifically, the present invention relates to a quality inspection method for the molding of automotive taillight masks. Background Art

[0002] The automotive taillight mask is a key functional component in the taillight assembly. As a transparent or semi-transparent plastic protective cover covering the taillight bulb, reflector, and circuit, it plays multiple important roles. The taillight mask not only provides physical protection for the internal optical elements against the external environment but also enables light diffusion and directional reflection through precisely designed optical surfaces to ensure that the optical performance of the taillight meets regulatory requirements. Given the crucial role of the taillight mask in safety performance and visual quality, the accurate detection of its molding quality has great engineering significance.

[0003] Currently, high-precision three-dimensional detection technologies are commonly used in the industry for quality control. Among them, the ICP (Iterative Closest Point) registration algorithm based on point cloud data is widely used due to its excellent measurement accuracy. This algorithm iteratively matches the three-dimensional scan data of the sample to be measured with the standard CAD model to achieve shape deviation detection, providing a reliable quantitative evaluation basis for the molding quality of the taillight mask.

[0004] In the three-dimensional point cloud data of automotive taillight masks, there are significant differences in the point cloud density in different regions. It is worth noting that although the number of point clouds in the sparse region is small, due to its spatial distribution characteristics, these discrete points have a disproportionately important impact on the expression of the overall structural characteristics. However, traditional point cloud processing methods usually adopt strategies of uniform sampling or global optimization, failing to fully consider this non-uniform density distribution and its impact on structure reconstruction, resulting in distortion or missing of structural characteristics in the sparse region during the reconstruction process. Summary of the Invention

[0005] To solve the technical problem of insufficient recognition accuracy of the key areas of the taillight mask, the present invention provides the following technical solutions.

[0006] A quality inspection method for the molding of automotive taillight masks, comprising: Collecting three-dimensional point cloud data of the automotive taillight mask and performing preprocessing; Calculating the tail cover structure significance index, screening the three-dimensional point cloud data of the automotive taillight mask based on the calculated tail cover structure significance index, and performing ICP matching between the screened three-dimensional point cloud data and the three-dimensional point cloud data of the designed shape to evaluate the deviation between the actual shape and the designed shape of the automotive taillight mask, thereby realizing the molding quality inspection of the automotive taillight mask; The calculation process of the tail cover structure significance index is as follows: Calculate the support degree of the tail cover structure for each three-dimensional point cloud, where the support degree of the tail cover structure is used to reflect the change degree of the tail cover structure after deleting the three-dimensional point cloud; calculate the local curvature of each three-dimensional point cloud; select any one three-dimensional point cloud as the target point cloud, and determine the neighbor point set of the target point cloud; calculate the product of the support degree of the tail cover structure and the local curvature of each three-dimensional point cloud in the neighbor point set, and find the maximum and minimum values of this product. Take the difference between the product corresponding to the three-dimensional point cloud in the neighbor point set and the minimum value of the product as the first parameter, take the difference between the maximum value of the product in the neighbor point set and the product corresponding to the three-dimensional point cloud in the neighbor point set as the second parameter, and sum the ratios of the first parameter to the second parameter of each three-dimensional point cloud in the neighbor point set to obtain the significant index of the tail cover structure of the target point cloud.

[0007] First, through the enhancement of three-dimensional point cloud features, that is, calculating the support degree of the structure and the local curvature of the three-dimensional point cloud. The support degree of the structure reflects the contribution degree of the point cloud to the overall shape of the tail cover, and the local curvature characterizes the surface geometric features (such as edges, corners, etc.). The product of the two highlights the points with both high geometric characteristics and structural importance. Further, combine the product of the two to calculate the significant index of the tail cover structure of the three-dimensional point cloud, quantify the structural characteristics of the taillight mask, and screen the three-dimensional point cloud data according to the significant index of the tail cover structure. Those points that have little impact on the taillight mask structure and are unimportant can be removed, and the key structural information points can be retained, which helps to focus on the key areas of the taillight mask. Using the screened key three-dimensional point cloud data for matching can more accurately detect the forming quality of the automotive taillight mask. By comparing the point cloud data of the actually produced taillight mask with the point cloud data of the standard model, problems such as deviations and defects that occur during the forming process, such as shape deformation and dimensional errors, can be discovered in a timely manner.

[0008] Preferably, the process of obtaining the support degree of the tail cover structure includes: Calculate the area of the region of the neighbor point set on three orthogonal projection planes, then remove the target point cloud from the neighbor point set to obtain a new point set, and calculate the area of the region of the new point set on three orthogonal projection planes; Calculate the sum of the area ratio differences between the area of the region of the new point set on three orthogonal projection planes and the area of the region of the neighbor point set on three orthogonal projection planes before removing the target point cloud, as the support degree of the tail cover structure of the target point cloud.

[0009] By calculating the area of the region of the neighbor point set on three orthogonal projection planes, the local structural characteristics around the target point cloud can be quantified. After removing the target point cloud, calculate the area of the region of the new point set and compare it with the area of the region of the original neighbor point set, and the impact of the target point cloud on the local structural stability can be evaluated.

[0010] By comparing the change in the projected area before and after removing the target point, the role of this point in maintaining the local geometry is reflected. If the area decreases significantly after removal, it indicates that this point is a key support point (such as an edge or a corner point); otherwise, it may be a redundant point (such as a point inside a flat area).

[0011] Preferably, the process of obtaining the support degree of the tail cover structure includes: Calculate the sum of the distances between the target point cloud and all the neighboring point clouds in the neighboring point set to obtain the support degree of the tail cover structure of the target point cloud.

[0012] Preferably, the process of obtaining the support degree of the tail cover structure further includes: Calculate the neighboring aggregation degree of the target point cloud based on the neighboring point set of the target point cloud, take the reciprocal of the neighboring aggregation degree as the weight, and take the product of the weight and the support degree of the tail cover structure as the adjusted support degree of the tail cover structure.

[0013] The neighboring aggregation degree can measure the aggregation degree of the target point cloud in the neighboring point set. The point cloud area with a high aggregation degree usually contributes more to the stability of the overall structure. By taking the reciprocal of the aggregation degree as the weight, the support degree of the tail cover structure can be adjusted so that the structural support degree can better reflect the importance of the key area.

[0014] Preferably, the process of obtaining the neighboring aggregation degree includes: Project the neighboring point set of the target point cloud onto three projection planes respectively, calculate the area of the projection region on each projection plane, take the ratio of the number of neighboring points to the area of the projection region as the aggregation degree component of each projection plane, and take the average value of the aggregation degree components of the three projection planes to obtain the neighboring aggregation degree.

[0015] By calculating the neighboring aggregation degree, the aggregation degree of the local area of the target point cloud can be quantified. For example, in the point cloud data obtained by 3D scanning, the point cloud distribution density and aggregation state of different object surfaces or different structures are different, and the neighboring aggregation degree can be used as a characteristic index to describe this local characteristic, helping to distinguish different point cloud regions, such as distinguishing flat surfaces, edges, and corner points, etc.

[0016] Preferably, the process of obtaining the neighboring aggregation degree includes: Project the neighboring point set of the target point cloud onto three projection planes respectively, calculate the area and aspect ratio of the minimum circumscribed rectangle on each projection plane, calculate the average value of the ratio of the area of the minimum circumscribed rectangle of all projection planes to the aspect ratio of the minimum circumscribed rectangle to obtain the neighboring aggregation degree.

[0017] Preferably, the screening process is: Taking the tail cover structure significance index of all the collected three-dimensional point cloud data as the input, processing it using a clustering algorithm to obtain multiple clusters, calculating the mean value of the tail cover structure significance index of all the point clouds in each cluster respectively, then sorting these mean values in ascending order, selecting the first set number of clusters and removing them to obtain the filtered three-dimensional point cloud data.

[0018] By processing the three-dimensional point cloud data with a clustering algorithm, dividing the similar point cloud data into the same cluster, and then screening and removing according to the mean value of the tail cover structure significance index, a large amount of duplicate or similar point cloud information can be removed, thereby reducing data redundancy; and the clustering algorithm can identify the outliers or noise points in the data, which usually have a lower tail cover structure significance index. By screening and removing these points, the quality and reliability of the data can be improved.

[0019] Preferably, the clustering algorithm is the DBSCAN clustering algorithm.

[0020] Preferably, the screening process is as follows: Removing the three-dimensional point cloud data corresponding to the tail cover structure significance index less than a preset threshold to obtain the filtered three-dimensional point cloud data.

[0021] Preferably, after arranging all the tail cover structure significance indexes in ascending order, they are divided into four equal parts, and the 75th percentile is selected as the threshold for screening.

[0022] The beneficial effects of the present invention are as follows: The present invention obtains the tail cover structure support degree by calculating the area and its change of the near-neighbor point set of the point cloud on three orthogonal projection planes, and further considers the adjustment of the tail cover structure support degree by the near-neighbor aggregation degree, comprehensively evaluates the importance of each point cloud in the taillight mask structure from multiple dimensions, further calculates the tail cover structure significance index and filters the three-dimensional point cloud data based on this, can effectively identify the areas with important structural features on the taillight mask, exclude the irrelevant or interfering point cloud data, make the subsequent matching detection more accurately focus on the key parts, thereby improving the accuracy of the forming quality detection, and can more sensitively detect the shape deviation, defects and other problems that may occur in the taillight mask during the forming process. Description of the Drawings

[0023] Figure 1 is the method flow chart of steps S1 - S2 in a quality detection method for an automobile taillight mask forming according to an embodiment of the present invention.

[0024] Figure 2 is the schematic diagram of steps S20 - S22 in a quality detection method for an automobile taillight mask forming according to an embodiment of the present invention. Detailed Embodiments

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments.

[0026] Referring to Figure 1 , a quality inspection method for forming an automotive taillight mask includes steps S1 - S2, which are specifically as follows: S1: Collect the three - dimensional point cloud data of the automotive taillight mask and perform pre - processing.

[0027] In one embodiment, a laser scanner is used to collect the three - dimensional point cloud data of the automotive taillight mask, and each point cloud point contains its coordinate information in space.

[0028] Due to factors such as environmental interference during the collection process, there may be noise, outliers and other noise points in the point cloud. Therefore, the Gaussian filtering algorithm can be used to denoise the collected three - dimensional point cloud data.

[0029] At the same time, collect the three - dimensional point cloud data of the automotive taillight mask with a standard shape structure according to the same working environment and steps as the template point cloud of the tail mask, which is used to match with the point cloud of the tail mask to be detected later.

[0030] S2: Calculate the structure significance index of the tail mask, screen the three - dimensional point cloud data of the automotive taillight mask based on the calculated structure significance index of the tail mask, and perform ICP matching between the screened three - dimensional point cloud data and the three - dimensional point cloud data of the designed shape to evaluate the deviation between the actual shape and the designed shape of the automotive taillight mask, so as to realize the forming quality inspection of the automotive taillight mask.

[0031] In the point cloud data of the automotive taillight mask, since the scanning direction of the laser scanner is vertically downward, the point cloud point density in different regions of the automotive taillight mask is different. In the sparse region of the point cloud, each point cloud point is like an isolated small wooden stick in a net, and its supporting effect on the whole structure is relatively large because there are no other point cloud points around to share the supporting task. For example, there may be fewer point cloud points at the edge of the mask due to scanning angle and other reasons, so these point cloud points are crucial for supporting the edge structure of the mask. On the contrary, in the dense region of the point cloud, the absence of a point cloud point can be compensated by other surrounding point cloud points, and the impact on the overall structure is small, so its structural support degree is relatively low.

[0032] In order to quantify this support degree and evaluate the significance of the structure corresponding to the point cloud points, a structure significance index of the tail mask is constructed. By constructing the structure significance index of each point cloud data, key point cloud points with a higher contribution degree to the structure of the automotive taillight mask can be identified, and these point cloud points are usually located in geometric feature regions such as the ridge lines and edges of the automotive taillight mask.

[0033] Reference Figure 2 The acquisition of the significant index of the above-mentioned tail cover structure includes steps S20 - S22.

[0034] S20: Calculate the tail cover structure support degree of each 3D point cloud, and the tail cover structure support degree is used to reflect the change degree of the tail cover structure after deleting the 3D point cloud.

[0035] In one embodiment, any one of the 3D point cloud data after the above-mentioned S1 preprocessing is selected as the target point cloud, and the K nearest neighbor point clouds (in the embodiment of the present invention, the value of K is 10, and in other embodiments, the value can be adjusted according to specific situations) of the target point cloud are selected to construct the nearest neighbor point set of the target point cloud (then the number of point clouds in the nearest neighbor point set of the target point cloud is 11). The nearest neighbor point set is projected onto three orthogonal projection planes of X - Y, X - Z, and Y - Z. For each projection plane, the Alpha Shapes concave hull algorithm is used to connect the external contour points of the projected planar point cloud to form a region, and the area of the formed region is calculated by the shoelace formula.

[0036] The above-mentioned Alpha Shapes concave hull algorithm can generate a suitable concave polygon according to the point set to describe the shape of the point set, and the shoelace formula is used to calculate the area of the polygon.

[0037] According to the area of each projection plane calculated above, calculate the nearest neighbor aggregation degree of the target point cloud.

[0038] Exemplarily, the nearest neighbor aggregation degree of the above-mentioned target point cloud satisfies the relationship as:

[0039] In the formula, is the nearest neighbor aggregation degree of the target point cloud, is the area of the region after the nearest neighbor point set of the target point cloud is projected onto the th projection plane. represents the aggregation degree component of the th projection plane, which measures the point cloud density in the th projection direction.

[0040] The above reflects the local density size of the target point cloud by calculating the mean value of the areas of the regions where the nearest neighbor point set of the target point cloud is projected onto each projection plane. The larger the

[0041] value, the more extensive and relatively uniform the distribution of the nearest neighbor point set of the target point cloud on the projection plane, that is, the higher the aggregation degree. Exemplarily, another way to calculate the nearest neighbor aggregation degree of the above-mentioned target point cloud is also provided, that is, the relationship is satisfied as:

[0042] In the formula, is the neighbor aggregation degree of the target point cloud, is the area of the minimum circumscribed rectangle after the projection of the neighbor point set of the target point cloud on the th projection plane, is the aspect ratio of the length to the width of the minimum circumscribed rectangle after the projection of the neighbor point set of the target point cloud on the th projection plane.

[0043] The on the above single projection plane reflects the specific distribution characteristics of the point set on this plane. By comprehensively considering the situations on all projection planes, the value is calculated. Therefore, even if the value on a certain projection plane is large (indicating a long and narrow distribution), as long as the distribution on other projection planes is wide and uniform enough, the value can still be large, indicating a high aggregation degree of the overall point set.

[0044] Based on the above neighbor density of the target point cloud, the support degree of the tail cover structure of the target point cloud is further calculated. By deleting the target point cloud from the neighbor point set of the target point cloud, a new point set is obtained, and the area of the region after the projection of the new point set is recalculated. Further, the area ratio of the area of the region after the projection of the new point set and the area of the region after the projection of the neighbor point set before deletion is calculated to reflect the support degree of the target point cloud for the automotive taillight mask structure.

[0045] Exemplarily, when calculating the support degree of the tail cover structure of the target point cloud, the relational expression that is satisfied is:

[0046] In the formula, is the support degree of the tail cover structure of the target point cloud, is the area of the region after the projection of the neighbor point set of the target point cloud on the th projection plane, is the area of the region after the projection of the neighbor point set recalculated after deleting the target point cloud from the neighbor point set on the th projection plane.

[0047] Among them, by comparing and , the influence of the target point cloud on the distribution of the neighbor point set on the projection plane can be evaluated. represents the difference in the area ratio of the areas of the regions after the projection of the neighbor point set before and after deleting the target point cloud. If the corresponding projection areas change significantly after removing the target point cloud, then is large, indicating that the target point cloud has a great influence on the structural stability.

[0048] Exemplarily, another method for calculating the support degree of the tail cover structure of the target point cloud is also provided, that is, the relational expression is satisfied as:

[0049] In the formula, is the support degree of the tail cover structure of the target point cloud, represents the Euclidean distance between the target point cloud and its th nearest neighbor point cloud.

[0050] When is relatively large, it means that the distance between the target point cloud and its th nearest neighbor point cloud is relatively far. In terms of spatial distribution, it means that the target point cloud is sparser relative to its neighboring point clouds. In the structural analysis of the automotive taillight mask, if a certain point is sparser relative to its neighbors, it may play a more "independent" or "supporting" role in the structure.

[0051] In the target point cloud, different points have different contributions to the support degree of the tail cover structure due to their different positions, densities, or relationships with other points. By introducing weights, this difference can be quantified, so that when calculating the support degree of the tail cover structure, the actual contribution of each point can be more accurately reflected. The weights can be dynamically adjusted according to indicators such as the nearest neighbor aggregation degree, so as to more finely evaluate the stability of the tail cover structure. For example, in areas with a higher nearest neighbor aggregation degree, the density of points is larger, and the supporting effect on the tail cover structure may be stronger. Therefore, a smaller weight can be given (because the reciprocal of the nearest neighbor aggregation degree is smaller, but after multiplying by a larger basic support degree value of the tail cover structure, the overall contribution can still be maintained at a high level); while in areas with a lower nearest neighbor aggregation degree, the density of points is smaller, and the supporting effect on the tail cover structure may be weaker. Therefore, a larger weight is given to highlight its relative importance (in the overall calculation, although the contribution of a single point is small, the total impact of this area on the support degree of the tail cover structure may not be ignored after the weight adjustment).

[0052] In one embodiment, the reciprocal of the nearest neighbor aggregation degree of the target point cloud calculated above is used as the weight, and the product of this weight and the support degree of the tail cover structure calculated above is used as the adjusted support degree of the tail cover structure.

[0053] Exemplarily, calculating the adjusted support degree of the tail cover structure of the target point cloud, that is, the relational expression is satisfied as:

[0054] In the formula, is the adjusted support degree of the tail cover structure of the target point cloud, is the support degree of the tail cover structure of the target point cloud, is the reciprocal of the nearest neighbor aggregation degree of the target point cloud.

[0055] Similarly, according to the above operations, the support degree of the tail cover structure for all the collected three-dimensional point cloud data and the adjusted support degree of the tail cover structure can be calculated.

[0056] S21: Calculate the local curvature of each three-dimensional point cloud.

[0057] To further identify the geometric features of the automotive taillight mask, such as ridgelines, depressions, convex regions, etc., it is necessary to calculate the local curvature of the point cloud data. By using the least squares plane algorithm to perform surface fitting on all the point cloud data in the neighborhood point set of the point cloud data, a surface equation is obtained, and the surface equation is differentiated twice. The obtained second-order derivative is used as the local curvature of the point cloud data. The magnitude of the local curvature can reflect the smoothness of the local surface of the point cloud data. The point cloud with a larger local curvature may belong to the ridgeline or edge region of the automotive taillight mask, and the point cloud with a smaller local curvature may belong to the surface region of the automotive taillight mask.

[0058] S22: Calculate the product of the support degree of the tail cover structure and the local curvature of each three-dimensional point cloud in the neighborhood point set, find the maximum and minimum values of this product, take the difference between the product corresponding to the three-dimensional point cloud in the neighborhood point set and the minimum value of the product as the first parameter, take the difference between the maximum value of the product in the neighborhood point set and the product corresponding to the three-dimensional point cloud in the neighborhood point set as the second parameter, and then sum the ratios of the first parameter and the second parameter of each three-dimensional point cloud in the neighborhood point set to obtain the tail cover structure significant index of the target point cloud.

[0059] In the tail cover point cloud to be detected, a tail cover structure significant index is calculated by comprehensively considering the support degree of the tail cover structure and the local curvature of the point cloud. The support degree of the tail cover structure reflects the degree of support of the point cloud points for the structure of the automotive taillight mask. If a point is located in an important structural area of the taillight mask, such as a ridgeline or an edge, it may contribute more to the overall structure, so its support degree of the tail cover structure will be higher. By considering the support degree of the tail cover structure, the points that are important for the structure of the taillight mask can be highlighted; the local curvature describes the degree of curvature at the location of the point cloud points. In the automotive taillight mask, key structural areas such as ridgelines and edges usually have a large curvature change. By introducing the local curvature, these areas with large curvature changes can be captured, so as to identify the key structures of the taillight mask.

[0060] By calculating the significant index of the tail cover structure, each point in the point cloud data of the automotive taillight mask can be quantitatively evaluated, so as to identify which points are more likely to be located in key structural areas such as the ridge lines and edges of the automotive taillight mask. These key structural areas are very important for aspects such as the design, manufacturing, and quality inspection of the automotive taillight mask. By calculating the significant index, the points in the point cloud data can be sorted or classified according to their structural significance, which helps subsequent analysis and processing, such as extracting the features of the taillight mask and detecting defects in the mask. It can be understood that for a three-dimensional point cloud, if the tail cover structure changes greatly after deleting the three-dimensional point cloud, it means that the three-dimensional point cloud can reflect the tail cover structure to a large extent, and the three-dimensional point cloud has a large tail cover structure support degree; local curvature can characterize the surface geometric features, and the larger the local curvature, the more complex the geometric features near the three-dimensional point cloud, and the three-dimensional point cloud can also reflect the tail cover structure to a large extent; to ensure the accuracy of the forming quality inspection, three-dimensional point clouds that can reflect the tail cover structure should be selected for matching. Therefore, the product of the tail cover structure support degree and the local curvature is used as the significant index of the tail cover structure to accurately evaluate the ability of the three-dimensional point cloud to reflect the tail cover structure.

[0061] Exemplarily, the above-mentioned significant index of the tail cover structure satisfies the following relationship:

[0062] In the formula, is the significant index of the tail cover structure of the target point cloud, is the tail cover structure support degree of the th three-dimensional point cloud in the neighbor point set, is the local curvature of the th three-dimensional point cloud, represents the minimum value of the product of the tail cover structure support degree and the local curvature in the neighbor point set, represents the maximum value of the product of the tail cover structure support degree and the local curvature in the neighbor point set, is the first parameter, is the second parameter. Among them, the 1 in the denominator is a hyperparameter to prevent the denominator from being zero.

[0063] According to the above operations, the significant index of the tail cover structure of all the collected three-dimensional point cloud data can be calculated. Points with a high significant index are more likely to be located in the structural feature areas (such as edges or ridge lines), while points with a low significant index are located on the smooth surface.

[0064] In one embodiment, the significant index of the tail cover structure of all the collected three-dimensional point cloud data is used as the input, and the DBSCAN clustering algorithm is used for processing. This clustering algorithm will divide the three-dimensional point cloud into different clustering clusters according to the distribution of the significant index of the tail cover structure of the three-dimensional point cloud.

[0065] Calculate the mean of the tail cover structure significant indices of all the point clouds in each cluster respectively, and then sort these means in ascending order. Select the top N clusters from the sorting result (N is a hyperparameter, which takes the value of 3 in the embodiments of the present invention and can be adjusted according to actual situations in other embodiments), and remove all the point clouds in these N clusters from all the collected point cloud data, so as to remove the point cloud data with relatively low tail cover structure significance, which may interfere with subsequent matching and forming quality detection.

[0066] In another embodiment, a simple threshold-based direct screening method is provided.

[0067] Exemplarily, traverse the tail cover structure significant indices of all the collected point cloud data, sort all the tail cover structure significant indices in ascending order and divide them into four equal parts, select the 75th percentile as the threshold for screening, and then remove the three-dimensional point cloud data corresponding to the tail cover structure significant index less than the threshold to obtain the screened three-dimensional point cloud data.

[0068] After the above screening steps, the screened three-dimensional point cloud data is obtained. Use the ICP algorithm to perform matching based on these data, and detect the forming quality of the automobile tail light mask through the matching result.

[0069] The ICP algorithm finds the best transformation relationship between two groups of point clouds in an iterative manner, so that the two groups of point clouds can be aligned as much as possible, thereby evaluating the deviation between the actual shape and the designed shape of the automobile tail light mask, and then judging its forming quality.

[0070] It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. A quality inspection method for the molding of an automobile taillight mask, characterized in that Including: Collecting three-dimensional point cloud data of the automotive taillight mask and performing preprocessing; Calculating the significant index of the tail mask structure, screening the three-dimensional point cloud data of the automotive taillight mask based on the calculated significant index of the tail mask structure, and performing ICP matching between the screened three-dimensional point cloud data and the three-dimensional point cloud data of the designed shape to evaluate the deviation between the actual shape and the designed shape of the automotive taillight mask, so as to realize the forming quality detection of the automotive taillight mask; The calculation process of the significant index of the tail mask structure is as follows: calculating the tail mask structure support degree of each three-dimensional point cloud, where the tail mask structure support degree is used to reflect the change degree of the tail mask structure after deleting the three-dimensional point cloud; calculating the local curvature of each three-dimensional point cloud; Selecting any one three-dimensional point cloud as the target point cloud, determining the near neighbor point set of the target point cloud; calculating the product of the tail mask structure support degree and the local curvature of each three-dimensional point cloud in the near neighbor point set, and finding the maximum value and the minimum value of the product. Taking the difference between the product corresponding to the three-dimensional point cloud in the near neighbor point set and the minimum value of the product as the first parameter, taking the difference between the maximum value of the product in the near neighbor point set and the product corresponding to the three-dimensional point cloud in the near neighbor point set as the second parameter, and summing the ratios of the first parameter and the second parameter of each three-dimensional point cloud in the near neighbor point set to obtain the significant index of the tail mask structure of the target point cloud.

2. The quality inspection method for forming an automobile taillight mask according to claim 1, characterized in that, The obtaining process of the tail mask structure support degree includes: Calculating the area of the region of the near neighbor point set on three orthogonal projection planes, then removing the target point cloud from the near neighbor point set to obtain a new point set, and calculating the area of the region of the new point set on three orthogonal projection planes; Calculating the sum of the area ratio differences between the area of the region of the new point set on three orthogonal projection planes and the area of the region of the near neighbor point set on three orthogonal projection planes before removing the target point cloud, as the tail mask structure support degree of the target point cloud.

3. A quality inspection method for forming an automotive taillight mask according to claim 1, characterized in that, The obtaining process of the tail mask structure support degree includes: Calculating the sum of the distances between the target point cloud and all the near neighbor point clouds in the near neighbor point set to obtain the tail mask structure support degree of the target point cloud.

4. A quality inspection method for forming an automobile taillight mask according to claim 2 or 3, characterized in that, The obtaining process of the tail mask structure support degree further includes: Calculating the near neighbor aggregation degree of the target point cloud based on the near neighbor point set of the target point cloud, taking the reciprocal of the near neighbor aggregation degree as the weight, and taking the product of the weight and the tail mask structure support degree as the adjusted tail mask structure support degree.

5. A quality inspection method for forming an automobile taillight mask according to claim 4, characterized in that, The obtaining process of the near neighbor aggregation degree includes: Projecting the near neighbor point set of the target point cloud onto three projection planes respectively, calculating the projected area of each projection plane, taking the ratio of the number of near neighbor points and the projected area as the aggregation degree component of each projection plane, and taking the average value of the aggregation degree components of the three projection planes to obtain the near neighbor aggregation degree.

6. A quality inspection method for forming an automobile taillight mask according to claim 4, characterized in that, The obtaining process of the near neighbor aggregation degree includes: Projecting the near neighbor point set of the target point cloud onto three projection planes respectively, calculating the area and the aspect ratio of the minimum circumscribed rectangle of each projection plane, and calculating the average value of the ratios of the area of the minimum circumscribed rectangle of all projection planes to the aspect ratio of the minimum circumscribed rectangle to obtain the near neighbor aggregation degree.

7. A quality inspection method for forming an automotive tail light mask according to claim 5, characterized in that, The screening process is: Taking the significant index of the fairing structure of all the collected three-dimensional point cloud data as the input, processing it using a clustering algorithm to obtain multiple clustering clusters, calculating the mean of the significant index of the fairing structure of all the point clouds in each clustering cluster respectively, then sorting these means in ascending order, selecting the first set number of clustering clusters and removing them to obtain the filtered three-dimensional point cloud data.

8. A quality inspection method for the molding of an automobile tail lamp mask according to claim 7, characterized in that, The clustering algorithm is the DBSCAN clustering algorithm.

9. A quality inspection method for forming an automobile tail light mask according to claim 5, characterized in that, The screening process is as follows: Removing the three-dimensional point cloud data corresponding to the significant index of the fairing structure less than a preset threshold to obtain the filtered three-dimensional point cloud data.

10. A quality inspection method for forming an automotive taillight mask according to claim 9, characterized in that, After arranging all the significant indices of the fairing structure in ascending order, dividing them into four equal parts and selecting the 75th percentile as the threshold for screening.

Images