Impeller weld start and end point detection method, device, equipment and storage medium

Abstract

This application discloses a method, apparatus, equipment, and storage medium for detecting the starting point of an impeller weld. The method includes: firstly, scanning the edge of the chassis at the front of the weld; fitting the chassis plane with the data obtained from the scan; then projecting the points of the chassis edge onto the fitted plane; fitting a standard circle of the chassis edge using the points of the chassis edge on the fitted chassis plane; subsequently scanning the weld formed by the blade and the chassis plane; projecting the weld points onto the fitted chassis plane; and finally calculating the intersection point of the weld curve and the standard circle of the chassis edge using the weld points on the fitted chassis plane. This method obtains chassis plane data while rapidly scanning the chassis edge, and accurately calculates the coordinates of the weld endpoint using the inherent information of the impeller, resulting in high detection efficiency and good stability.

Description

Technical Field

[0001] This application relates to a weld inspection technology, specifically to a method, apparatus, equipment, and storage medium for detecting the starting point of a weld on a wind turbine impeller. Background Technology

[0002] Welding is a crucial task in modern industrial production, especially in impeller manufacturing. Impellers consist of a base and blades, and their production requires extensive welding to secure the blades to the base support. This demands high levels of skill and experience from the welders, but manual welding is inefficient. Therefore, using robots to weld impellers has become the mainstream choice. Compared to manual welding, robotic welding offers significant advantages, such as faster welding speed, more uniform and aesthetically pleasing welds, and higher weld quality. However, previously widely used teachable robots are now ill-suited for the diverse shapes and sizes of impellers and the complexity of weld seams. Therefore, robot-assisted vision guidance technology is now commonly used for impeller welding.

[0003] However, current vision guidance systems on robots cannot accurately detect weld seam endpoints, which poses a significant challenge to achieving weld seam wrap-around. In existing technologies, there are two main approaches to using vision-guided robots to achieve impeller weld seam wrap-around welding: one is using a 3D surface structured light camera. 3D surface structured light cameras can acquire a large amount of point cloud data and detect weld seam endpoints based on this data. However, they are susceptible to interference, and reconstructing the point cloud data is time-consuming, resulting in low efficiency. The other approach is using a 3D line structured light camera. 3D line structured light cameras are more efficient, but it is difficult to guarantee scanning all weld seams and detecting their endpoints. This is mainly because impeller weld seam endpoints are often affected by weld points or oil contamination. Even without such interference, accurately detecting weld seam endpoints is not easy. To accurately detect weld seam endpoints, the scanning speed needs to be reduced to increase the amount of data, which directly leads to a decrease in efficiency. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, this application provides a method, apparatus, device, and storage medium for rapidly and accurately detecting the end point of an impeller weld. The specific technical solution of this application is as follows:

[0005] In its first aspect, this application proposes a method for detecting the starting point of an impeller weld. This method is based on initial workpiece information obtained through a pre-scanning mode and includes the following steps:

[0006] Step S1: Scan a section of the chassis edge in front of the weld to obtain three-dimensional point cloud data of the chassis plane and the chassis edge;

[0007] Step S2: Based on the three-dimensional point cloud data of the chassis plane obtained in step S1, fit the chassis plane using the RANSAC algorithm. The fitted chassis plane corresponds to the actual chassis.

[0008] Step S3: Based on the three-dimensional point cloud data of the chassis edge obtained in step S1, project the three-dimensional points of the chassis edge onto the chassis plane fitted in step S2 to obtain the three-dimensional points of the chassis edge on the fitted chassis plane.

[0009] Step S4: Based on the three-dimensional points of the chassis edge on the fitted chassis plane obtained in step S3, fit a standard circle of the chassis edge using the RANSAC algorithm.

[0010] Step S5: Scan the entire weld seam according to the preset start and end coordinates obtained from the pre-scan to obtain the three-dimensional point cloud data of the weld seam;

[0011] Step S6: Based on the weld three-dimensional point cloud data obtained in step S5, project the weld three-dimensional coordinate points onto the chassis plane fitted in step S2.

[0012] Step S7: Use a curve fitting algorithm to fit the three-dimensional points of the weld obtained in step S6 to obtain the fitted weld curve.

[0013] Step S8: Calculate the intersection point of the standard circle of the chassis edge fitted in step S4 and the weld curve fitted in step S7. If the actual endpoint of the weld coincides with the intersection point, the intersection point is the coordinate of the starting endpoint of the weld. If the actual endpoint of the weld does not coincide with the intersection point, the position coordinates of the actual endpoint can be calculated based on the known distance.

[0014] Preferably, step S2 specifically comprises:

[0015] Step S201: Based on the chassis planar 3D point cloud data obtained in step S1, select any three non-collinear points and calculate the equation of the plane they form. The formula used is:

[0016] Ax+By+Cz+D=0 (Formula 1)

[0017] Step S202: Calculate the distance d between all three-dimensional points of the chassis plane obtained in step S1 and the plane obtained in step S201. i And set a threshold distance d between a 3D point on the chassis plane and the plane to be determined. th When d i ≤d th If the point is in an inlier, it is considered an outlier; otherwise, it is considered an outlier. The number of inliers in the current total number of samples in the plane is recorded, and the distance d is...i The calculation formula is:

[0018] d i =|Ax i +By i +Cz i +D| (Formula 2)

[0019] Step S203: Repeat the above steps and select the model parameters corresponding to the plane with the most interior point data in the first iteration as the best fitting parameters.

[0020] Step S204: After the iteration ends, the selected optimal model parameters are the final plane equation parameters. Let n = (n x ,n y n z ) is the unit normal vector of the plane.

[0021] Preferably, step S4 specifically comprises:

[0022] Step S401: Based on the three-dimensional points of the chassis edge on the fitted chassis plane obtained in step S3, select three non-collinear points P1(x1, y1, z1), P2(x2, y2, z2), and P3(x3, y3, z3). According to formulas 3 and 4, calculate the vector passing through line segment P1P2. Line segment P2P3 vector Given the normal vector n of the fitted chassis plane, the vectors V1 and V2 corresponding to the perpendicular bisector L1 of line segment P1P2 and the perpendicular bisector L2 of line segment P2P3 are obtained respectively:

[0023]

[0024]

[0025] Step S402: Let the intersection point P0 = (x0, y0, z0) of lines L1 and L2 be the center of the standard circle of the chassis edge to be fitted, and let the radius of the standard circle of the chassis edge to be fitted be R, where the radius R is the distance from point P0 to P1.

[0026] Step S403: Calculate the parametric equation of the fitted standard circle of the chassis edge based on the center and radius obtained in step S402. Let the vector of line segment P0P1 be... After normalization, set it as According to Formula 5, the vector The cross product of the fitted chassis plane normal vector n and the obtained vector V is:

[0027]

[0028] After normalizing the vector V, set it as The parametric equation of the fitted standard circle of the chassis edge is shown in Formula 6, where the value of θ ranges from [0, 2π]:

[0029]

[0030] Step S404: Let d be the distance from all points on the fitted chassis edge plane to the center P0 = (x0, y0, z0). l Set a threshold distance d thresh When the distance d from the edge point of the chassis plane to the center P0 is... l The absolute value of the difference between the radius R of the fitted standard circle of the chassis edge and the radius |d l -R|≤d thresh When the point is on the fitted standard circle of the chassis edge, the number of chassis edge points on the fitted standard circle of the chassis edge is recorded, and the distance d is... l The calculation formula is:

[0031]

[0032] Step S405: Repeat the above steps to calculate the parametric equation of the chassis edge standard circle when the number of chassis edge points on the fitted chassis edge standard circle is the maximum.

[0033] A second aspect of this application discloses a device for detecting the starting point of an impeller weld, comprising:

[0034] The scanning unit is used to scan the impeller and weld seam to obtain three-dimensional point cloud data of the chassis plane, chassis edge and weld seam;

[0035] The fitting unit is used to fit the chassis plane, the standard circle of the chassis edge, and the weld curve based on the three-dimensional point cloud data obtained by the scanning unit. The fitted chassis plane corresponds to the actual chassis.

[0036] The projection unit is used to project the coordinates of the acquired three-dimensional points on the chassis edge and the three-dimensional points on the weld onto the fitted chassis plane.

[0037] The determining unit is used to determine whether the intersection of the actual endpoint of the weld and the calculated fitted standard circle of the chassis edge and the fitted weld curve coincides.

[0038] In a third aspect, this application provides a computer device comprising a memory and a processor, wherein the memory stores program instructions that, when executed by the processor, cause the processor to perform the steps in the impeller weld start-end point detection method described above.

[0039] In a fourth aspect, this application proposes a storage medium storing program instructions capable of implementing the impeller weld start-end detection method described above.

[0040] Beneficial effects:

[0041] This invention rapidly scans the chassis edge, obtaining both chassis edge data and chassis plane data simultaneously. The obtained data is then used to fit the chassis plane and the chassis edge circle. Furthermore, the inherent information of the impeller is utilized to accurately calculate the coordinates of the weld endpoints. Unlike existing technologies that require reconstructing point cloud data at the weld endpoints, this method improves detection efficiency and enhances the stability of the detection method, avoiding data deviations caused by external interference when reconstructing the point cloud data at the weld endpoints. Attached Figure Description

[0042] Figure 1 This is a flowchart of the weld start point detection method according to an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of the impeller structure according to an embodiment of the present invention;

[0044] Figure 3 This is a three-dimensional point cloud diagram of the chassis plane and chassis edge as described in the embodiments of the present invention;

[0045] Figure 4 This is a schematic diagram of the fitted chassis plan as described in an embodiment of the present invention;

[0046] Figure 5 This is a schematic diagram of the projection of the chassis edge points according to an embodiment of the present invention;

[0047] Figure 6 This is a schematic diagram of the method for fitting a standard circle along the edge of a chassis according to an embodiment of the present invention;

[0048] Figure 7 This is a schematic diagram of the standard circle of the fitted chassis edge as described in an embodiment of the present invention;

[0049] Figure 8 This is a schematic diagram of the weld point projection according to an embodiment of the present invention;

[0050] Figure 9 This is a schematic diagram of the fitted weld curve according to an embodiment of the present invention;

[0051] Figure 10 This is a schematic diagram of the functional modules of the impeller weld start point detection device according to an embodiment of the present invention.

[0052] Figure 11 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention.

[0053] Figure 12 This is a schematic diagram of the impeller structure according to an embodiment of the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] Example 1

[0056] The first aspect of this application provides a method for detecting the outer endpoint of a weld formed by a blade and a chassis using a 3D line structured light camera. This detection method is based on initial workpiece information obtained in a pre-scanning mode, and refers to... Figure 1 The steps include the following:

[0057] Step S1, as follows Figure 2 As shown, the impeller consists of a front disc cover 201, a chassis 202, and blades 203. In this embodiment, by scanning a section of the chassis edge in front of the weld between the blades 203 and the chassis 202, three-dimensional point cloud data of the chassis plane and chassis edge in the area where the weld is located are obtained. The obtained three-dimensional point cloud data of the chassis plane and chassis edge are shown in the figure. Figure 3 As shown;

[0058] Step S2: Based on the 3D point cloud data of the chassis plane obtained in Step S1, a chassis plane is fitted using the RANSAC (Random Sample Consensus) algorithm. The fitted chassis plane corresponds to the actual chassis plane where the impeller chassis is located. In this embodiment, the specific steps for fitting the chassis plane using the RANSAC algorithm are as follows:

[0059] Step S201: Based on the chassis planar 3D point cloud data obtained in step S1, select any three non-collinear points and calculate the equation of the plane they form using Formula 1:

[0060] Ax+By+Cz+D=0 (Formula 1)

[0061] Step S202: Calculate the distance d between all three-dimensional points of the chassis plane obtained in step S1 and the plane obtained in step S201. i And set a fixed distance d between a three-dimensional point on the chassis plane and the plane obtained in step S201. th As a threshold distance, if d i ≤d thIf the value is within the model, it is considered an inlier; otherwise, it is considered an outlier. The number of inliers in the current total number of samples in the plane is recorded based on the above data. The distance d... i The calculation formula is:

[0062] d i =|Ax i +By i +Cz i +D| (Formula 2)

[0063] Step S203: Repeat the above steps, and at the end of each iteration, decide whether to stop the iteration based on the expected error rate, the ratio of the number of inliers to the total number of samples, and the current iteration number. Select the model parameters corresponding to the plane with the most inliers in the iteration as the best fitting parameters.

[0064] Step S204: After the iteration ends, the selected optimal model parameters are the final plane equation parameters, such as... Figure 4 As shown, this plane is the fitted chassis plane, and let n = (n x n y n z ) is the unit normal vector of the plane.

[0065] Step S3: Based on the 3D point cloud data of the chassis edge obtained in step S1, project the obtained 3D points of the chassis edge onto the chassis plane fitted in step S2 to obtain the 3D point data of the chassis edge on the chassis plane, such as... Figure 5 As shown in the figure, the darker point 502 is the original three-dimensional point of the chassis edge, and the brighter point 501 is the three-dimensional point projected onto the fitted chassis plane. The three-dimensional point of the chassis edge is not the same as the three-dimensional point of the chassis edge on the projected chassis plane, but the two partially overlap in spatial position.

[0066] Step S4: Based on the three-dimensional points of the chassis edge on the fitted chassis plane obtained in step S3, such as... Figure 6 The RANSAC algorithm is used to fit a standard circle along the chassis edge. The specific steps are as follows:

[0067] Step S401: Based on the three-dimensional points along the chassis edge on the chassis plane obtained in step S3, select any three non-collinear points P1(x1, y1, z1), P2(x2, y2, z2), and P3(x3, y3, z3). According to formulas 3 and 4, calculate the vector passing through line segment P1P2. Line segment P2P3 vector Given the normal vector n of the fitted chassis plane, the vectors V1 and V2 corresponding to the perpendicular bisector L1 of line segment P1P2 and the perpendicular bisector L2 of line segment P2P3 are obtained respectively:

[0068]

[0069]

[0070] Step S402: Let the intersection point P0 = (x0, y0, z0) of lines L1 and L2 be the center of the standard circle of the chassis edge to be fitted, and let the radius of the standard circle of the chassis edge to be fitted be R, where the radius R is the distance from point P0 to P1.

[0071] Step S403: Calculate the parametric equation of the fitted standard circle of the chassis edge based on the center and radius obtained in step S402. Let the vector of line segment P0P1 be... After normalization, set it as According to Formula 5, the vector The cross product of the fitted chassis plane normal vector n and the obtained vector V is:

[0072]

[0073] After normalizing the vector V, set it as The parametric equation of the fitted standard circle of the chassis edge is shown in Formula 6, where the value of θ ranges from [0, 2π]:

[0074]

[0075] Step S404: Let d be the distance from all points on the chassis edge of the chassis plane to the center P0 = (x0, y0, z0). l Set a threshold distance d thresh When the distance d from the edge point of the chassis on the fitted chassis plane to the center P0 is... l The absolute value of the difference between the radius R of the fitted standard circle of the chassis edge and the radius |d l -R|≤d thresh When the point is on the fitted standard circle of the chassis edge, the number of chassis edge points on the fitted standard circle of the chassis edge is recorded, and the distance d is... l The calculation formula is:

[0076]

[0077] Step S405: Repeat the above steps to calculate the parametric equation of the fitted chassis edge standard circle when the number of chassis edge points on the fitted chassis edge standard circle is the maximum. The fitted chassis edge standard circle is as follows: Figure 7 As shown.

[0078] Step S5: Scan the entire weld seam according to the preset start and end coordinates obtained from the pre-scan to obtain the three-dimensional point cloud data of the weld seam;

[0079] Step S6: Based on the weld 3D point cloud data obtained in step S5, project the obtained weld 3D point coordinates onto the chassis plane fitted in step S2. The projected weld points 801 are as follows: Figure 8 As shown;

[0080] Step S7, as follows Figure 9 As shown, the three-dimensional points of the weld obtained in step S6 are fitted using a curve fitting algorithm to obtain the fitted weld curve 901.

[0081] Step S8: Calculate the intersection point of the standard circle of the chassis edge fitted in step S4 and the weld curve fitted in step S7. If the actual endpoint of the weld coincides with the intersection point, then the intersection point is the coordinate of the weld start endpoint. When there are two or more intersection points, only the intersection point closest to the first point of the weld is taken as the final weld start endpoint. If the actual endpoint of the weld does not coincide with the intersection point, the position coordinates of the actual endpoint can be calculated based on the known distance. Due to the uniformity of impeller machining height, the weld distance value of the same type of impeller only needs to be set once. Here, the "known distance" can be determined based on the distance from the actual impeller blade weld start endpoint to the chassis edge circle.

[0082] Figure 10 This is a functional module diagram of an impeller weld start-end detection device according to the second aspect of an embodiment of this application. Figure 10 As shown, the impeller weld start point detection device 10 includes a scanning unit 11, a fitting unit 12, a projection unit 13, and a confirmation unit 14.

[0083] The scanning unit 11 is used to scan the impeller and weld seam to obtain three-dimensional point cloud data of the chassis plane, chassis edge and weld seam.

[0084] The fitting unit 12 is used to fit the chassis plane, the standard circle of the chassis edge, and the weld curve based on the three-dimensional point cloud data obtained by the scanning unit 11. The fitted chassis plane corresponds to the actual chassis.

[0085] Projection unit 13 is used to project the coordinates of the acquired three-dimensional points of the chassis edge and the three-dimensional points of the weld onto the fitted chassis plane.

[0086] The determining unit 14 is used to determine whether the intersection of the actual endpoint of the weld and the calculated fitted standard circle of the chassis edge and the fitted weld curve coincides.

[0087] Figure 11This is a schematic diagram of the structure of the computer device described in the third aspect of an embodiment of this application. Figure 11 As shown, the computer device 20 includes a processor 21 and a memory 22. The memory 22 stores program instructions. When the program instructions are executed by the processor 21, the memory 22 causes the processor 21 to perform the steps in the impeller weld start-end point detection method described above. The processor 21 has the ability to process signals and can be an integrated circuit chip, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array or other programmable logic device, a discrete gate or transistor logic device.

[0088] In a fourth aspect of this application, a storage medium is provided that stores program instructions capable of implementing the impeller weld start-end detection method described above. The storage medium may include random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (FLASH RAM).

[0089] Example 2

[0090] Example 2 also provides a method for detecting the outer endpoint of the weld formed by the blade and the chassis by acquiring three-dimensional point cloud data of the chassis plane and chassis edge data through a 3D surface structured light camera, replacing the method in Example 1 of this application that acquires three-dimensional point cloud data of the chassis plane and chassis edge data through a 3D line structured light camera. The technical solution of Example 2 is the same as that of Example 1 above.

[0091] Example 3

[0092] like Figure 12 As shown, the technical solution of Embodiment 3 is not limited to obtaining three-dimensional point cloud data of the chassis plane and three-dimensional point cloud data of the chassis edge using visual guidance. It can also be used for the method of obtaining three-dimensional point cloud data of the blade edge 1204. By performing curve and straight line fitting on the weld 1203 and the blade edge 1204, the intersection point is found, and finally the true endpoint of the weld 1203 is obtained.

[0093] Example 4

[0094] like Figure 12 As shown, the technical solution of Embodiment 4 is not limited to detecting the weld end point formed by the blade 1205 and the chassis 1202, but also provides a method for detecting the weld end point formed by the blade 1205 and the front disc cover 1201.

Claims

1. A method for detecting the starting point of an impeller weld, the method being based on initial workpiece information acquired through a pre-scanning mode, characterized in that, Includes the following steps: Step S1: Scan a section of the chassis edge in front of the weld to obtain three-dimensional point cloud data of the chassis plane and the chassis edge; Step S2: Based on the three-dimensional point cloud data of the chassis plane obtained in step S1, fit the chassis plane using the RANSAC algorithm. The fitted chassis plane corresponds to the actual chassis. Step S3: Based on the three-dimensional point cloud data of the chassis edge obtained in step S1, project the obtained three-dimensional points of the chassis edge onto the chassis plane fitted in step S2 to obtain the three-dimensional points of the chassis edge on the fitted chassis plane. Step S4: Based on the three-dimensional points of the chassis edge on the fitted chassis plane obtained in step S3, fit a standard circle of the chassis edge using the RANSAC algorithm. Step S5: Scan the entire weld seam according to the preset start and end coordinates obtained from the pre-scan to obtain the three-dimensional point cloud data of the weld seam; Step S6: Based on the weld three-dimensional point cloud data obtained in step S5, project the weld three-dimensional coordinate points onto the chassis plane fitted in step S2. Step S7: Use a curve fitting algorithm to fit the three-dimensional points of the weld obtained in step S6 to obtain the fitted weld curve. Step S8: Calculate the intersection point of the standard circle of the chassis edge fitted in step S4 and the weld curve fitted in step S7. If the actual endpoint of the weld coincides with the intersection point, the intersection point is the coordinate of the starting endpoint of the weld. If the actual endpoint of the weld does not coincide with the intersection point, the position coordinates of the actual endpoint can be calculated based on the known distance.

2. The method for detecting the starting point of an impeller weld according to claim 1, characterized in that, Step S2 specifically involves: Step S201: Based on the chassis planar 3D point cloud data obtained in step S1, select any three non-collinear points and calculate the equation of the plane they form using Formula 1: Ax+By+Cz+D=0 Formula 1 Step S202: Calculate the distance d from all three-dimensional points on the chassis plane obtained in step S1 to the plane obtained in step S201. i Set the threshold distance d between the three-dimensional points on the chassis plane and the plane obtained in step S201. th If d i ≤d th If the value is true, then the point is an in-model sample point; otherwise, it is an out-of-model sample point. Record the current number of in-model points, and the distance d... i The calculation formula is: d i =|Ax i +By i +Cz i +D| Official 2 Step S203: Repeat the above steps and select the model parameters corresponding to the plane with the most interior point data as the best fitting parameters. Step S204: After the iteration is completed, the selected optimal model parameters are the final plane equation parameters. Let n = (n x ,n y ,n z ) is the unit normal vector of the plane.

3. The method for detecting the starting point of an impeller weld according to claim 1, characterized in that, Step S4 specifically involves: Step S401: Based on the three-dimensional points of the chassis edge on the fitted chassis plane obtained in step S3, select three non-collinear points P1(x1,y1,z1), P2(x2,y2,z2), and P3(x3,y3,z3). According to formulas 3 and 4, calculate the vector of line segment P1P2. Line segment P2P3 vector Given the normal vector n of the fitted chassis plane, the vectors V1 and V2 corresponding to the perpendicular bisector L1 of line segment P1P2 and the perpendicular bisector L2 of line segment P2P3 are obtained respectively: Step S402: Let the intersection point P0 = (x0, y0, z0) of lines L1 and L2 be the center of the standard circle of the chassis edge to be fitted, and let the radius of the standard circle of the chassis edge to be fitted be R, where the radius R is the distance from point P0 to P1. Step S403: Calculate the parametric equation of the fitted standard circle of the chassis edge based on the center and radius obtained in step S402. Let the vector of line segment P0P1 be... After normalization, set it as According to Formula 5, the vector u is cross-producted with the fitted chassis plane normal vector n to obtain vector V: After normalizing the vector V, set it as The parametric equation of the fitted standard circle of the chassis edge is shown in Formula 6, where the value of θ ranges from [0, 2π]: Step S404: Let d be the distance from all points on the fitted chassis edge plane to the center P0 = (x0, y0, z0). l Set a threshold distance d thresh When the distance d from the edge point of the chassis on the fitted chassis plane to the center P0 is... l The absolute value of the difference between the radius R of the fitted standard circle of the chassis edge and the radius |d l -R|≤d thresh When the point is on the fitted standard circle of the chassis edge, the number of chassis edge points on the fitted standard circle of the chassis edge is recorded, and the distance d is... l The calculation formula is: Step S405: Repeat the above steps to calculate the parametric equation of the chassis edge standard circle when the number of chassis edge points on the fitted chassis edge standard circle is the maximum.

4. The method for detecting the starting point of an impeller weld according to claim 1, characterized in that: Suitable for impeller front discs.

5. A device for detecting the starting point of an impeller weld, characterized in that, include: The scanning unit is used to scan the impeller and weld seam to obtain three-dimensional point cloud data of the chassis plane, chassis edge and weld seam; The fitting unit is used to fit the chassis plane, the standard circle of the chassis edge, and the weld curve based on the three-dimensional point cloud data obtained by the scanning unit. The fitted chassis plane corresponds to the actual chassis. The projection unit is used to project the coordinates of the acquired three-dimensional points on the chassis edge and the three-dimensional points on the weld onto the fitted chassis plane. The determining unit is used to determine whether the intersection of the actual endpoint of the weld and the calculated fitted standard circle of the chassis edge and the fitted weld curve coincides.

6. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores program instructions, which, when executed by the processor, cause the processor to perform the steps of the impeller weld start-end point detection method as described in any one of claims 1-3.

7. A storage medium, characterized in that, The system stores program instructions capable of implementing the impeller weld start-end detection method as described in any one of claims 1-3.

Images