Device for locating the drain outlet of large bag by using machine vision and its use method
The machine vision system calculates the central position of the drain outlet of the continuous casting large bag and corrects the robot's operating arm, which solves the problem that large bag positioning accuracy depends on human operation in continuous casting production, and achieves high-precision, safe and reliable robot operation.
Patent Information
- Application Number
- CN201811247514.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2018-10-15
- Filing Date
- 2018-10-25
- Publication Date
- 2024-12-17
- Estimated Expiration
- 2038-10-25
AI Technical Summary
At the continuous casting production site, the robot's positioning accuracy of large bags depends on the operator's habits and level, resulting in errors and position uncertainties, affecting the efficiency and safety of the robot's operation.
Using a machine vision system, the central position of the continuous casting large bag sewer is calculated through cameras and image processing computers, and the position of the robot operating arm is corrected through the position controller. The 2D or 3D positioning method is realized using a monocular or binocular vision system to ensure that The robot's precise operation of large bags.
The precise positioning and operation of the large-pack sewer port is achieved, the accuracy and safety of robot operation are improved, and human error is reduced.
Smart Images
Figure CN111036863B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of program-controlled manipulators, and in particular to a device for locating a drain outlet of a large bag by using machine vision and a use method thereof. Background Art
[0002] As robots are increasingly used, more and more robots are taking on heavy production work at continuous casting production sites. Some robots are also deployed around the ladle turret, such as robots that complete large ladle-related work at the ladle receiving position and casting position. The operations completed by robots in the large ladle area include installing the oil cylinder, installing the corresponding oil pipe and gas pipe signal line plugs, and installing the long nozzle.
[0003] In order to achieve continuous production, modern continuous casting machines mostly use rotary tables or transverse trolleys. Generally, large bags are transported to the rotary table or transverse trolley by crane. The positioning accuracy at this time is mainly determined by the operating habits and operating level of the on-site operators, and errors are inevitable. At the same time, the movement of large bags causes the position of the large bags to be different each time, which will cause many obstacles for the robot to operate the large bags, because the robot can no longer perform related operations on the large bags in a fixed position. For this reason, it is necessary to add a robot vision system to complete the corresponding operations. Summary of the invention
[0004] In order to overcome the defects of the prior art and provide a mechanical operating device with precise control, safety and reliability, the present invention discloses a device for locating a drain outlet of a large bag using machine vision and a method for using the device.
[0005] The present invention achieves the purpose of the invention through the following technical solutions:
[0006] A device for locating a water outlet of a ladle using machine vision comprises a continuous casting ladle and a robot. The bottom of the continuous casting ladle is provided with a water outlet, and the operating arm of the robot is arranged on one side of the continuous casting ladle. The device is characterized in that it also comprises a base, a camera, an image processing computer and a position controller.
[0007] The base is fixed on the operating arm of the robot. At least one camera is fixed on the base. The objective lenses of the cameras are all aimed at the water outlet of the continuous casting ladle. The signal output end of the camera is connected to the image processing computer through a signal line. The signal output end of the image processing computer is connected to the position controller through a signal line. The command output end of the position controller is connected to the robot body through a signal line.
[0008] The device for locating the water outlet of the ladle using machine vision is characterized in that the angle between the central axis of the camera objective lens and the central axis of the water outlet of the continuous casting ladle is not greater than 60°;
[0009] The position controller is a control cabinet or a programmable controller.
[0010] The method for using the device for locating the drain outlet of a large bag using machine vision is characterized by being implemented in sequence according to the following steps:
[0011] ① Shooting: Align the objective lenses of all cameras to the outlet of the continuous casting ladle, and shoot the position image of the outlet of the continuous casting ladle in real time. The camera converts the captured position image into an electrical signal and inputs it into the image processing computer through the signal line;
[0012] ② Calculation: The image processing computer calculates the center position of the continuous casting ladle outlet and outputs the center position information to the position controller through the signal line;
[0013] ③ Correction: The position controller corrects the operating position of the robot operating arm according to the measured central position information of the continuous casting ladle outlet, so that the robot operating arm is aligned with the continuous casting ladle outlet.
[0014] The method for using the device for locating the drain outlet of a large ladle using machine vision is characterized in that: in step ②, if the number of cameras is one, that is, monocular vision is used, a 2D positioning method is used to identify the contour of the drain outlet of the continuous casting ladle and calculate the center position of the drain outlet. The specific steps of the 2D positioning method are: the drain outlet of the continuous casting ladle is circular, and when the camera photographs the drain outlet at the bottom of the continuous casting ladle, since there is an angle of no more than 30° between the camera and the vertical direction of the bottom of the continuous casting ladle, the image of the drain outlet photographed is elliptical. According to image recognition theory, the center point of the elliptical image of the drain outlet is calculated to be the center position of the drain outlet;
[0015] If there are more than one cameras, that is, binocular vision or multi-eye vision is used, the center position of the water outlet is calculated by the 3D positioning method. The specific steps of the 3D positioning method are: positioning the water outlet by identifying multiple feature points on the continuous casting ladle, and the number of the feature points is not less than three, identifying the position and posture of the water outlet, and calculating the parameters of the water outlet in the Cartesian coordinate system, that is, (x, y, z, a, b, c).
[0016] The present invention adopts a monocular or binocular vision system to realize the positioning of the ladle system through an image recognition algorithm, and sends the change of this position to the robot, thereby realizing the hand-eye coordination of the robot.
[0017] The present invention can be used to detect the position of the drain outlet of a ladle. When the robot needs to operate the drain outlet, the drain outlet position is detected and calculated through monocular, binocular or multi-eye vision and then sent to the robot, thereby completing the hand-eye coordination of the robot.
[0018] The beneficial effects of the present invention are: precise control, safety and reliability. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 It is a structural schematic diagram of the present invention. DETAILED DESCRIPTION
[0020] The present invention is further described below by means of specific examples.
[0021] Example 1
[0022] A device for positioning a ladle outlet using machine vision comprises a continuous casting ladle 1, a robot 2, a base 3, a camera 4, an image processing computer 5 and a position controller 6, such as Figure 1 As shown, the specific structure is:
[0023] A water outlet 11 is provided at the bottom of the continuous casting ladle 1, and an operating arm of the robot 2 is provided at one side of the continuous casting ladle 1;
[0024] The base 3 is fixed on the operating arm of the robot 2. At least one camera 4 is fixed on the base 3. The objective lenses of the cameras 4 are all aimed at the water outlet of the continuous casting ladle 1. The signal output end of the camera 4 is connected to the image processing computer 5 through a signal line. The signal output end of the image processing computer 5 is connected to the position controller 6 through a signal line. The command output end of the position controller 6 is connected to the robot 2 body through a signal line.
[0025] In this embodiment: the angle between the central axis of the objective lens of the camera 4 and the central axis of the drain port 11 of the continuous casting ladle 1 is not greater than 60°;
[0026] The position controller 6 uses a control cabinet or a programmable controller;
[0027] The camera 4 may use laser detection.
[0028] When using this embodiment, follow the following steps in sequence:
[0029] ① Shooting: aim the objective lenses of all cameras 4 at the water outlet 11 of the continuous casting ladle 1, and shoot the position image of the water outlet 11 of the continuous casting ladle 1 in real time. The cameras 4 convert the captured position image into an electrical signal and input it into the image processing computer 5 through a signal line;
[0030] ② Calculation: The image processing computer 5 calculates the center position of the water outlet 11 of the continuous casting ladle 1 and outputs the center position information to the position controller 6 through a signal line;
[0031] If the number of cameras 4 is one, that is, monocular vision is adopted, the 2D positioning method is used to identify the outline of the drain port 11 of the continuous casting ladle 1 and calculate the center position of the drain port 11. The specific steps of the 2D positioning method are: the drain port 11 of the continuous casting ladle 1 is circular. When the camera 4 photographs the drain port 11 at the bottom of the continuous casting ladle 1, since there is an angle of no more than 30° between the camera 4 and the vertical direction of the bottom of the continuous casting ladle 1, the image of the drain port 11 photographed is elliptical. According to the image recognition theory, the center point of the elliptical image of the drain port 11 is calculated as the center position of the drain port 11;
[0032] If the number of cameras 4 is more than one, that is, binocular vision or multi-eye vision is used, the center position of the water outlet 11 is calculated by a 3D positioning method. The specific steps of the 3D positioning method are: positioning the water outlet 11 by identifying multiple feature points on the water outlet 11 of the continuous casting ladle 1, the number of the feature points is not less than three, identifying the position and posture of the water outlet 11, and calculating the parameters of the water outlet 11 in the Cartesian coordinate system, that is, (x, y, z, a, b, c);
[0033] The Cartesian coordinate system is used to describe and determine the position and posture of the robot 2 operating arm. The origin of the Cartesian coordinate system is set at the center of the first joint of the robot 2 operating arm. First, determine the x-axis and y-axis of the plane rectangular coordinate system, and then determine the positive direction of the z-axis according to the right-hand rule. The parameters x, y, and z in the Cartesian coordinate system refer to the distance between the projection of the object on the x-axis, y-axis, and z-axis of the Cartesian coordinate system and the origin, respectively. The parameters a, b, and c in the Cartesian coordinate system refer to the rotation angle of the object around the x-axis, y-axis, and z-axis, respectively. Usually, a is called the deflection angle, b is called the pitch angle, and c is called the rotation angle;
[0034] ③ Correction: The position controller 6 corrects the operating position of the operating arm of the robot 2 according to the measured central position information of the water outlet 11 of the continuous casting ladle 1, so that the operating arm of the robot 2 is aligned with the water outlet 11 of the continuous casting ladle 1.
[0035] When using the 3D positioning method, since it is impossible to directly locate the water outlet 11, some feature points are required. These feature points are generally of the equipment itself, such as the round hole of the continuous casting ladle 1, etc. However, if the equipment itself does not have them, feature points can be artificially added at fixed positions of the continuous casting ladle 1, such as adding some objects of special shapes such as circles and rectangles as marks. The number of the feature points is not less than three, and the relative positions of these special-shaped marks and the water outlet 11 are fixed. In 3D positioning, the position of the water outlet 11 can be indirectly calculated by identifying these marks as feature points.
Claims
1. A method for using a device for locating a drain outlet of a ladle using machine vision, the device for locating a drain outlet of a ladle using machine vision comprising a continuous casting ladle (1), a robot (2), a base (3), a camera (4), an image processing computer (5) and a position controller (6), A water outlet (11) is provided at the bottom of the continuous casting ladle (1), and an operating arm of the robot (2) is provided on one side of the continuous casting ladle (1); The base (3) is fixed on the operating arm of the robot (2), at least one camera (4) is fixed on the base (3), the objective lens of the camera (4) is aimed at the water outlet of the continuous casting ladle (1), the signal output end of the camera (4) is connected to the image processing computer (5) through a signal line, the signal output end of the image processing computer (5) is connected to the position controller (6) through a signal line, and the command output end of the position controller (6) is connected to the robot (2) body through a signal line; The angle between the central axis of the objective lens of the camera (4) and the central axis of the water outlet (11) of the continuous casting ladle (1) is not greater than 60°; Position controller (6) uses a programmable controller; Its characteristics are: Follow the steps below to implement: ① Shooting: aiming the objective lenses of all cameras (4) at the water outlet (11) of the continuous casting ladle (1), and shooting the position image of the water outlet (11) of the continuous casting ladle (1) in real time; the cameras (4) convert the captured position image into an electrical signal and input it into an image processing computer (5) through a signal line; ② Calculation: The image processing computer (5) calculates the center position of the water outlet (11) of the continuous casting ladle (1) and outputs the center position information to the position controller (6) through a signal line; When the number of cameras (4) is more than one, the center position of the water outlet (11) is calculated by a 3D positioning method, and the specific steps of the 3D positioning method are: positioning the water outlet (11) by identifying a plurality of feature points on the water outlet (11) of the continuous casting ladle (1), wherein the number of the feature points is not less than three, identifying the position and posture of the water outlet (11), and calculating the parameters of the water outlet (11) in a Cartesian coordinate system, namely (x, y, z, a, b, c); A Cartesian coordinate system is used to describe and determine the position and posture of the operating arm of the robot (2), wherein the origin of the Cartesian coordinate system is set at the center of the first joint of the operating arm of the robot (2); ③ Calibration: The position controller (6) calibrates the operating position of the operating arm of the robot (2) according to the measured central position information of the water outlet (11) of the continuous casting ladle (1), so that the operating arm of the robot (2) is aligned with the water outlet (11) of the continuous casting ladle (1); Feature points are artificially added as marks at fixed positions of the continuous casting ladle (1), and the number of the feature points is not less than three. The relative positions of these marks of special shapes and the water outlet (11) are fixed. During 3D positioning, the position of the water outlet (11) is indirectly calculated by identifying these marks as feature points.