A three-dimensional space special-shaped surface rotation calibration physical model and a 3D camera rotation calibration method

Abstract

This invention proposes a three-dimensional spatial irregular curved surface rotation calibration solid model and a 3D camera rotation calibration method. The three-dimensional spatial irregular curved surface rotation calibration solid model includes an arc plate and multiple positioning and identification pillars. The arc plate has multiple equidistant first mounting holes, and the number of first mounting holes corresponds one-to-one with the number of positioning and identification pillars. The positioning and identification pillars are partially housed in the first mounting holes and fixedly connected to the arc plate. The three-dimensional spatial irregular curved surface rotation calibration solid model is easy to manufacture and can be completed by ordinary machine tools using turning, milling, and drilling, without complex processing procedures, resulting in low production costs. The 3D camera rotation calibration method only requires one laser 3D camera to complete the camera calibration, while ordinary planar calibration technology requires three or more laser 3D cameras, or three or more scans followed by image stitching. Compared with ordinary planar calibration technology, it is lower in cost and faster in efficiency.

Description

Technical Field

[0001] This invention relates to the field of camera calibration, mainly to the calibration of irregular curved surfaces in camera calibration space, and particularly to a three-dimensional irregular curved surface rotation calibration solid model and a 3D camera rotation calibration method. Background Technology

[0002] Currently, in the field of 3D vision, 3D calibration technology is generally planar calibration, that is, the camera scans the target object horizontally. However, the object exists in a three-dimensional space, so the camera cannot image all sides of the object at once. It may require multiple cameras to combine images and then stitch them together, or a single camera may need to scan multiple times and then stitch them together. This technology leads to long scanning time, high cost, loss of stitching accuracy, and will greatly increase the technical difficulties, development time and debugging time. Summary of the Invention

[0003] To address the aforementioned issues, this invention proposes a three-dimensional spatial irregular curved surface rotation calibration solid model and a 3D camera rotation calibration method that requires no image stitching and can complete camera calibration using only a laser 3D camera.

[0004] This invention is achieved through the following technical solution:

[0005] This invention proposes a three-dimensional spatial irregular curved surface rotation calibration solid model for camera calibration. The three-dimensional spatial irregular curved surface rotation calibration solid model includes an arc plate and multiple positioning and identification columns. The arc plate is provided with multiple equidistant first mounting holes, and the number of the first mounting holes corresponds one-to-one with the number of the positioning and identification columns. The positioning and identification columns are partially housed in the first mounting holes and fixedly connected to the arc plate.

[0006] Furthermore, the arc-shaped plate is provided with an arched surface and a positioning groove; the positioning groove is located in the center of the arched surface.

[0007] Furthermore, the positioning and identification column is provided with a positioning hole that penetrates through the positioning and identification column.

[0008] Furthermore, the three-dimensional spatial irregular curved surface rotation calibration solid model also includes a bottom mounting structure; the bottom mounting structure is fixedly connected to the arc plate, and the bottom mounting structure is also provided with a second mounting hole penetrating through the bottom mounting structure.

[0009] A 3D camera rotation calibration method includes a three-dimensional spatial irregular curved surface rotation calibration solid model as described in claims 1-4, the 3D camera rotation calibration method comprising the following steps:

[0010] S1: The three-dimensional irregular curved surface rotation calibration solid model is installed on the worktable by screws engaging with the second mounting hole;

[0011] S2: Align the laser line of the laser 3D camera with the cylindrical axis of the three-dimensional irregular curved surface rotation calibration solid model, and ensure that the field of view of the laser 3D camera encompasses the three-dimensional irregular curved surface rotation calibration solid model.

[0012] S3: The laser 3D camera rotates around the cylindrical axis of the solid model to scan and acquire an image.

[0013] S4: The software algorithm reconstructs the original image of the three-dimensional irregular curved surface rotation calibration solid model based on the imaging image, and calculates the original image of the solid model. Figure 3 The three-dimensional coordinates of the original image in D space are given. The original image is a point cloud. Each point in the point cloud has a unique (x, y, z) coordinate. The normal vector of each point is calculated from the shape of the original image, thereby calculating the (rx, ry, rz) coordinates of each point in the point cloud.

[0014] S5: Use the TCP tip of the positioning mechanism to detect the (x,y,z) coordinates of the positioning hole, and record the (x,y,z) coordinates of multiple positioning holes according to the requirements of the software algorithm;

[0015] S6: The software algorithm combines the (x,y,z) coordinates of the point cloud map with the (x,y,z) coordinates of the positioning hole detected by the TCP tip of the positioning mechanism. Through the software algorithm, the (x,y,z) coordinates of the point cloud map are transformed into the (x,y,z) coordinates of the TCP. The (rx,ry,rz) coordinates of the TCP are calculated entirely by the software algorithm. At this point, the laser 3D camera completes the calibration process.

[0016] Furthermore, after completing steps S1-S6, the three-dimensional irregular curved surface rotation calibration solid model can be removed from the workbench, the product to be inspected can be placed on the workbench, and the laser 3D camera can be used to scan the product to obtain the positioning coordinates or inspection data of the product to be inspected.

[0017] The beneficial effects of this invention are:

[0018] 1. The three-dimensional irregular curved surface rotation calibration solid model proposed in this invention is easy to process. It can be completed by ordinary machine tools with turning, milling and drilling. There is no need for complicated processing procedures and the production cost is low.

[0019] 2. The 3D camera rotation calibration method proposed in this invention only requires one laser 3D camera to complete the camera calibration, while ordinary planar calibration technology requires three or more laser 3D cameras, or three or more scans and then stitching the images together. Compared with ordinary planar calibration technology, it is lower in cost and faster in efficiency.

[0020] 3. The 3D camera rotation calibration method proposed in this invention is simple to operate. It only requires using a laser 3D camera to rotate around the cylindrical axis of the solid model once to calibrate the three-dimensional irregular curved surface.

[0021] 4. The 3D camera rotation calibration method proposed in this invention has higher data accuracy. Compared with the planar calibration technology of ordinary 3D laser 3D cameras, the 3D camera rotation calibration method proposed in this invention can obtain the three-dimensional image of the object without stitching images. Therefore, it will not lose accuracy due to stitching and ensures the accuracy of the data. Attached Figure Description

[0022] Figure 1 This is an exploded view of the three-dimensional spatial irregular curved surface rotation calibration solid model of the present invention;

[0023] Figure 2 This is a perspective view of the three-dimensional spatial irregular curved surface rotation calibration solid model of the present invention;

[0024] Figure 3 This is a point cloud diagram of the 3D camera rotation calibration method of the present invention;

[0025] Figure 4 This is a point cloud diagram of the 3D camera rotation calibration method of the present invention from another direction. Detailed Implementation

[0026] To more clearly and completely illustrate the technical solution of the present invention, the present invention will be further described below in conjunction with the accompanying drawings.

[0027] Please refer to Figures 1 to 4 This invention proposes a three-dimensional spatial irregular curved surface rotation calibration solid model for camera calibration. The three-dimensional spatial irregular curved surface rotation calibration solid model includes: an arc plate 10 and a plurality of positioning and identification columns 20. The arc plate 10 is provided with a plurality of equidistant first mounting holes 11, and the number of the first mounting holes 11 corresponds one-to-one with the number of the positioning and identification columns 20. The positioning and identification columns 20 are partially housed in the first mounting holes 11 and fixedly connected to the arc plate 10.

[0028] In this embodiment, the arc plate 10 is a semi-circular arc plate, and the number of positioning and identification pillars 20 is 21. The positioning and identification pillars 20 are installed equidistantly on the arc plate 10. The coordinates and directions are determined by the multiple positioning and identification pillars 20, thereby obtaining the directions of the X, Y, and Z axes of the coordinates. Then, based on the point cloud map scanned by the laser 3D camera, the normal vector on the corresponding surface can be obtained, thus obtaining the (x, y, z, rx, ry, rz) coordinates in three-dimensional space. The diameter of the positioning and identification pillars 20 matches the diameter of the first mounting hole 11, which facilitates the installation of the positioning and identification pillars 20. The three-dimensional irregular curved surface rotation calibration solid model is easy to process. It can be processed by ordinary machine tools by turning, milling, and drilling. There is no need for complicated processing procedures. Ordinary machine tools can meet the accuracy requirements of the three-dimensional irregular curved surface rotation calibration solid model, resulting in low production costs.

[0029] Furthermore, the arc-shaped plate 10 is provided with an arched surface 12 and a positioning groove 13; the positioning groove 13 is located in the center of the arched surface 12.

[0030] In this embodiment, the positioning groove 13 is located in the center of the arched surface 12. The positioning groove 13 is used for positioning. The positioning groove 13 can accurately machine a plurality of equidistant first mounting holes 11 on the arc plate 10. In addition, when it is necessary to install the three-dimensional irregular curved surface rotation calibration solid model onto the external worktable, the precise positioning of the positioning groove 13 can accurately install the three-dimensional irregular curved surface rotation calibration solid model onto the external worktable.

[0031] Furthermore, the positioning and identification column 20 is provided with a positioning hole 21, which penetrates the positioning and identification column 20.

[0032] In this embodiment, the positioning and identification column 20 is provided with positioning holes 21. Each positioning hole 21 determines a mechanical coordinate, forming a single direction. This single direction plays an important role in imaging in three-dimensional space. When the three-dimensional irregular curved surface rotation calibration solid model is installed on an external worktable, the position of the positioning hole 21 is determined and unique. When using the three-dimensional irregular curved surface rotation calibration solid model for camera calibration, the TCP tip of the positioning mechanism is used to detect the (x, y, z) coordinates of the positioning hole 21 to meet the requirements of camera calibration.

[0033] Furthermore, the three-dimensional spatial irregular curved surface rotation calibration solid model also includes a bottom mounting structure 30; the bottom mounting structure 30 is fixedly connected to the arc plate 10, and the bottom mounting structure 30 is also provided with a second mounting hole 31 penetrating the bottom mounting structure 30.

[0034] In this embodiment, the bottom mounting structure 30 can stably place the three-dimensional irregular curved surface rotation calibration solid model on the external worktable, and the external screws can be used to cooperate with the second mounting hole 31 to firmly install the three-dimensional irregular curved surface rotation calibration solid model onto the external worktable.

[0035] Please refer to Figure 3 and Figure 4 A 3D camera rotation calibration method includes a three-dimensional spatial irregular curved surface rotation calibration solid model as described in claims 1-4, wherein the 3D camera rotation calibration method includes the following steps:

[0036] S1: The three-dimensional irregular curved surface rotation calibration solid model is installed on the worktable by screws engaging with the second mounting hole;

[0037] S2: Align the laser line of the laser 3D camera with the cylindrical axis of the three-dimensional irregular curved surface rotation calibration solid model, and ensure that the field of view of the laser 3D camera encompasses the three-dimensional irregular curved surface rotation calibration solid model.

[0038] S3: The laser 3D camera rotates around the cylindrical axis of the solid model to scan and acquire an image.

[0039] S4: The software algorithm reconstructs the original image of the three-dimensional irregular curved surface rotation calibration solid model based on the imaging image, and calculates the original image of the solid model. Figure 3 The three-dimensional coordinates of the original image in D space are given. The original image is a point cloud. Each point in the point cloud has a unique (x, y, z) coordinate. The normal vector of each point is calculated from the shape of the original image, thereby calculating the (rx, ry, rz) coordinates of each point in the point cloud.

[0040] S5: Use the TCP tip of the positioning mechanism to detect the (x,y,z) coordinates of the positioning hole, and record the (x,y,z) coordinates of multiple positioning holes according to the requirements of the software algorithm;

[0041] S6: The software algorithm combines the (x,y,z) coordinates of the point cloud map with the (x,y,z) coordinates of the positioning hole detected by the TCP tip of the positioning mechanism. Through the software algorithm, the (x,y,z) coordinates of the point cloud map are transformed into the (x,y,z) coordinates of the TCP. The (rx,ry,rz) coordinates of the TCP are calculated entirely by the software algorithm. At this point, the laser 3D camera completes the calibration process.

[0042] In this embodiment, the operation of step S3 is simple, requiring only the use of a laser 3D camera to scan the cylindrical axis of the solid model once around the three-dimensional irregular curved surface.

[0043] Figure 3 and Figure 4 The laser 3D camera scans the point cloud map of the three-dimensional spatial irregular curved surface rotation calibration solid model. The imaging surface of the point cloud map is a plane, and the rx and ry of the plane are very small. After the first calibration, a standard calibration process is formed, and the second calibration can be performed automatically.

[0044] The 3D camera rotation calibration method only requires one laser 3D camera to complete the camera calibration, while ordinary planar calibration technology requires three or more laser 3D cameras, or three or more scans and then stitching the images together. Compared with ordinary planar calibration technology, it is lower in cost.

[0045] The 3D camera rotation calibration method has higher data accuracy. Compared with the planar calibration technology of ordinary 3D laser 3D cameras, the 3D camera rotation calibration method proposed in this invention can obtain the three-dimensional image of the object without stitching images. Therefore, it will not lose accuracy due to stitching and ensures the accuracy of the data.

[0046] Furthermore, after completing steps S1-S6, the three-dimensional irregular curved surface rotation calibration solid model can be removed from the workbench, the product to be inspected can be placed on the workbench, and the laser 3D camera can be used to scan the product to obtain the positioning coordinates or inspection data of the product to be inspected.

[0047] In this embodiment, the outline of the product to be inspected must not exceed the scanning range of the laser 3D camera.

[0048] Of course, the present invention may have many other embodiments. Based on this embodiment, other embodiments obtained by those skilled in the art without any creative effort are all within the scope of protection of the present invention.

Claims

1. A three-dimensional irregular curved surface rotation calibration solid model for camera calibration, characterized in that, The three-dimensional spatial irregular curved surface rotation calibration solid model includes: an arc plate and multiple positioning and identification columns. The arc plate is provided with multiple equidistant first mounting holes, and the number of the first mounting holes corresponds one-to-one with the number of the positioning and identification columns. The positioning and identification columns are partially housed in the first mounting holes and fixedly connected to the arc plate. The arc-shaped plate has an arched surface and a positioning groove; the positioning groove is located in the center of the arched surface; The positioning and identification column is provided with a positioning hole, which penetrates the positioning and identification column; The three-dimensional spatial irregular curved surface rotation calibration solid model also includes a bottom mounting structure; the bottom mounting structure is fixedly connected to the arc plate, and the bottom mounting structure is also provided with a second mounting hole penetrating the bottom mounting structure.

2. A 3D camera rotation calibration method, comprising a three-dimensional spatial irregular curved surface rotation calibration solid model as described in claim 1, characterized in that, The 3D camera rotation calibration method includes the following steps: S1: The three-dimensional irregular curved surface rotation calibration solid model is installed on the worktable by screws engaging with the second mounting hole; S2: Align the laser line of the laser 3D camera with the cylindrical axis of the three-dimensional irregular curved surface rotation calibration solid model, and ensure that the field of view of the laser 3D camera encompasses the three-dimensional irregular curved surface rotation calibration solid model. S3: The laser 3D camera rotates around the cylindrical axis of the solid model to scan and acquire an image. S4: The software algorithm reconstructs the original image of the three-dimensional irregular curved surface rotation calibration entity model based on the imaging image, and calculates the three-dimensional coordinates of the original image in the 3D space. The original image is a point cloud image, and each point in the point cloud image has a unique corresponding (x, y, z) coordinate. Then, the normal vector of each point is extracted through the shape of the original image, thereby calculating the (rx, ry, rz) coordinates of each point in the point cloud image. S5: Use the TCP tip of the positioning mechanism to detect the (x,y,z) coordinates of the positioning hole, and record the (x,y,z) coordinates of multiple positioning holes according to the requirements of the software algorithm; S6: The software algorithm combines the (x,y,z) coordinates of the point cloud map with the (x,y,z) coordinates of the positioning hole detected by the TCP tip of the positioning mechanism. The software algorithm then transforms the (x,y,z) coordinates of the point cloud map into the (x,y,z) coordinates of the TCP. The (rx,ry,rz) coordinates of the TCP are calculated entirely by the software algorithm. At this point, the laser 3D camera completes the calibration process. After completing steps S1-S6, the three-dimensional irregular curved surface rotation calibration solid model can be removed from the workbench. The product to be inspected can be placed on the workbench, and the laser 3D camera can be used to scan the product to obtain the positioning coordinates or inspection data of the product to be inspected.

Images