Welding spot positioning device based on multi-view vision and detection method

Abstract

The invention discloses a welding spot positioning device based on multi-view vision and a detection method.The welding spot positioning device comprises a robot, the robot comprises a fixing base, a rotating disc and a mechanical arm installed on the rotating disc are arranged on the fixing base, a mounting frame is installed at the end, away from the rotating disc, of the mechanical arm, and a welding spot detection device is installed on the mounting frame; the robot and the welding spot detection device installed on the mechanical arm of the robot are adopted, the problem that the welding quality is low due to the fact that a worker holds a sensor to conduct positioning traditionally is solved, the stability of the welding spot quality and the welding spot detection efficiency are improved, the production cost is reduced, and the productivity is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of position and posture estimation of welding spots in the automobile manufacturing process, and in particular to a welding spot positioning device and a detection method based on multi-eye vision. Background Art

[0002] At present, automobile factories generally use ultrasonic testing, eddy current testing and metallographic testing to test the quality of welding points. Welding point testing is usually carried out by workers holding special sensors for testing, and is often carried out in the form of random inspections. There are problems such as inability to achieve comprehensive testing of the welding points of the car body, low testing efficiency, and inability to meet the production rhythm of automobile manufacturing. Summary of the invention

[0003] The technical problem solved by the present invention is to provide a method for locating welds using multi-eye vision positioning, which replaces the traditional positioning by workers holding sensors. The structure adopts a multi-eye + eddy current sensor method to replace the traditional workers holding detection devices, and the movement of the eddy current sensor is controlled by a linear guide rail to realize the multi-eye vision-based weld positioning device detection.

[0004] The technical solution adopted by the present invention to solve its technical problem is: a welding point positioning device based on multi-eye vision, including a robot, the robot including a fixed seat, a rotating disk and a mechanical arm installed on the rotating disk, a mounting frame is installed at one end of the mechanical arm away from the rotating disk, and a welding point detection device is installed on the mounting frame.

[0005] Furthermore, the welding spot detection device includes a mounting plate installed on the mounting frame and a linear guide rail arranged on the other side of the mounting plate, four cameras arranged to be tilted inwardly are evenly installed on the mounting plate, a light source is arranged in the middle of the four cameras arranged to be tilted inwardly, and also includes an eddy current sensor arranged on the linear guide rail and a driving motor installed on the linear guide rail and driving the eddy current sensor to slide on the linear guide rail.

[0006] Further: S1: Establish Algorithm 1:

[0007] Assume that the coordinates of the calibration object are (X c， Y c， Z c ), the camera coordinates are (X l， Y l， Z l ), the right camera coordinates are (X r， Y r， Z r );

[0008]

[0009]

[0010] It can be concluded that: It can be concluded that: Take multiple groups

[0011] The values ​​of A, B, and C can be obtained. The value of A, B, and C multiplied by the machine is the rotation and translation matrix between the two cameras;

[0012] S2: Establish Algorithm 2:

[0013] Assuming that the camera coordinate system of the left camera is the world coordinate system, the relative position relationship between the world coordinate system and the right camera coordinate system can be expressed as:

[0014]

[0015] The three-dimensional coordinates of the target point P in space can be obtained as follows:

[0016]

[0017] In summary, the focal lengths f of the two cameras are known. l, f r And the image coordinates of the target point in the left and right cameras. We only need to find the rotation and translation matrix between the two cameras to find the three-dimensional coordinates of the target point P in space.

[0018] The beneficial effects of the present invention are: using a robot and a welding spot detection device installed on the robot's mechanical arm to replace the traditional method of positioning by workers holding sensors with the problem of low welding quality, thereby improving the stability of welding spot quality and the efficiency of welding spot detection, reducing production costs and improving productivity. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is an overall schematic diagram of the solder joint positioning device based on multi-vision.

[0020] Figure 2 It is a schematic diagram of the specific structure of the solder joint detection device.

[0021] Marked in the figure are: robot 1, fixed seat 11, rotating disk 12, mechanical arm 13, mounting frame 2, welding point detection device 3, mounting plate 31, camera 311, light source 312, linear guide rail 32, eddy current sensor 321, and driving motor 322. DETAILED DESCRIPTION

[0022] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0023] like Figure 1As shown, a welding spot positioning device detection based on multi-eye vision includes a robot 1, the robot 1 includes a fixed seat 11, a rotating disk 12 and a mechanical arm 13 installed on the rotating disk 12 are provided on the fixed seat 11, a mounting frame 2 is installed on the end of the mechanical arm 13 away from the rotating disk 12, and a welding spot detection device 3 is installed on the mounting frame 2; the robot 1 and the welding spot detection device 3 installed on the mechanical arm 13 of the robot 1 are used to replace the traditional positioning by workers holding sensors with the problem of low welding quality, thereby improving the stability of welding spot quality and the efficiency of welding spot detection, reducing production costs and improving productivity.

[0024] On this basis, if Figure 2 As shown, the solder joint detection device 3 includes a mounting plate 31 mounted on the mounting frame 2 and a linear guide rail 32 arranged on the other side of the mounting plate 31, four cameras 311 tilted inwardly are evenly mounted on the mounting plate 31, a light source 312 is arranged in the middle of the four cameras 311 tilted inwardly, and also includes an eddy current sensor 321 arranged on the linear guide rail 32 and a drive motor 322 installed on the linear guide rail 32 and driving the eddy current sensor 321 to slide on the linear guide rail 32; the camera 311 is a monocular 3 camera, and the three-dimensional coordinates of the object can be calculated by using four monocular cameras, which saves hardware costs. When the four cameras 311 collect image information, the light source 312 provides light source to improve the image effect, and the drive motor 322 drives the eddy current sensor 321 to slide on the linear guide rail 32, replacing the traditional worker's handheld detection device, improving the stability of solder joint quality and the efficiency of solder joint detection, reducing production costs, and improving productivity.

[0025] On this basis, S1: establish Algorithm 1:

[0026] Assume that the coordinates of the calibration object are (X c， Y c， Z c ), the camera coordinates are (X l， Y l， Z l ), the right camera coordinates are (X r， Y r， Z r );

[0027]

[0028]

[0029] It can be concluded that: It can be concluded that: Take multiple groups

[0030] The values ​​of A, B, and C can be obtained. The value of A, B, and C multiplied by the machine is the rotation and translation matrix between the two cameras;

[0031] S2: Establish Algorithm 2:

[0032] Assuming that the camera coordinate system of the left camera is the world coordinate system, the relative position relationship between the world coordinate system and the right camera coordinate system can be expressed as:

[0033]

[0034] The three-dimensional coordinates of the target point P in space can be obtained as follows:

[0035]

[0036] In summary, the focal lengths f of the two cameras are known. l, f r And the image coordinates of the target point in the left and right cameras. We only need to find the rotation and translation matrix between the two cameras to find the three-dimensional coordinates of the target point P in space.

[0037] In summary, the focal lengths f of the two cameras are known. l, f r And the image coordinates of the target point in the left and right cameras. We only need to find the rotation and translation matrix between the two cameras to find the three-dimensional coordinates of the target point P in space.

[0038] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A welding spot positioning detection device based on multi-eye vision, comprising a robot (1), characterized in that: The robot (1) comprises a fixed seat (11), a rotating disk (12) and a mechanical arm (13) mounted on the rotating disk (12) are arranged on the fixed seat (11), a mounting frame (2) is mounted on one end of the mechanical arm (13) away from the rotating disk (12), and a welding point detection device (3) is mounted on the mounting frame (2).

2. The solder joint positioning detection device based on multi-eye vision as claimed in claim 1, characterized in that: The solder joint detection device (3) comprises a mounting plate (31) mounted on a mounting frame (2) and a linear guide rail (32) arranged on a side of the mounting plate (31) away from the mounting frame (2); four cameras (311) arranged inwardly tilted are evenly mounted on the mounting plate (31); a light source (312) is arranged in the middle of the four cameras (311) arranged inwardly tilted, and an eddy current sensor (321) is arranged on the linear guide rail (32); and a driving motor (322) is mounted on the linear guide rail (32) and drives the eddy current sensor (321) to slide on the linear guide rail (32).

3. The method for detecting solder joint positioning based on multi-eye vision as claimed in claim 1, characterized in that: S1: Establish Algorithm 1: Assume that the coordinates of the calibration object are (X c， Y c， Z c ), the camera coordinates are (X l， Y l， Z l ), the right camera coordinates are (X r， Y r， Z r ); It can be concluded that: It can be concluded that: Take multiple groups The values ​​of A, B, and C can be obtained. The value of A, B, and C multiplied by the machine is the rotation and translation matrix between the two cameras; S2: Establish Algorithm 2: Assuming that the camera coordinate system of the left camera is the world coordinate system, the relative position relationship between the world coordinate system and the right camera coordinate system can be expressed as: The three-dimensional coordinates of the target point P in space can be obtained as follows: In summary, the focal lengths f of the two cameras are known. l, f r And the image coordinates of the target point in the left and right cameras. We only need to find the rotation and translation matrix between the two cameras to find the three-dimensional coordinates of the target point P in space.

Images