CN110815201A - Method for correcting coordinates of robot arm - Google Patents

Abstract

The invention relates to a method for correcting coordinates of a robot arm, which extracts an image of a calibration square block, rapidly identifies four corners of the calibration square block through color difference, calculates coordinates of a central point through intersection points of diagonal lines, and enables a central axis of a camera to be aligned with the central point. And filtering laser by using the dark calibration square to obtain bright end points of two edges of the calibration square, forming a laser line, guiding the laser line to coincide with the central point of the calibration square, and obtaining a TCP coordinate so as to automatically correct the coordinate.

Description

The method of coordinate correction of robot arm

technical field

The invention relates to a robot arm, in particular to an industrial robot arm and a coordinate correction method for a tool center thereof.

Background technique

The robot arm has the characteristics of flexible movement, precise positioning and continuous operation, although it has become the best tool for manufacturing and assembly on the production line. However, through the multi-axis actuator drive, wear and component manufacturing and assembly errors, the robot arm forms a difference between the control and actual movement coordinates, resulting in the inability to accurately position the robot arm. Therefore, correcting the coordinates is an important issue for the robot arm.

When correcting the coordinates of the robot arm in the prior art, the tool center point (ToolCenterPoint, TCP for short) of the robot arm is usually driven to directly contact the reference point of known coordinates to correct the coordinates of the tool center point of the robot arm. However, when the movement error of the uncalibrated robot arm is too large, it often collides with the reference point, causing damage to the tool of the robot arm and the reference point instrument. Therefore, there is another prior art, which uses a precision measuring instrument such as a laser rangefinder to measure the coordinates of the center point of the tool. Although the coordinates can be corrected in a non-contact manner, the laser rangefinder is expensive and requires a large installation space. restrictions on use. Therefore, there is another prior art US published patent case US20110320039. In the position of the known coordinates relative to the robot arm, a black and white correction plate with a specific relationship is set, and the camera on the robot arm is used to extract the image of the black and white correction plate, and calculate the difference between the camera and the black and white correction plate. The relative relationship of the calibration plate with known coordinates to calibrate the robot arm.

However, although the aforementioned robotic arm in the prior art can automatically perform calibration without contact, there are many types of calibration plates from various manufacturers, and each type of calibration plate has its own unique installation structure and calculation method of relative relationship, which cannot be interacted or compatible. It is not easy to use, maintain and manage for general operators who use multiple types or brands of robotic arms. Therefore, there is still an urgent problem to be solved in the coordinate correction of the robot arm.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method for correcting coordinates of a robot arm, using the camera on the robot arm to extract the image of the calibration block, identifying the four corners of the calibration block by the difference between the calibration block and the environment, and obtaining the coordinates of the center point, so as to quickly align the camera the central axis.

Another object of the present invention is to provide a method for calibrating coordinates of a robotic arm, which uses a dark calibration square to filter laser light, obtains bright endpoints on the edge of the calibration square, forms a laser line, and then guides the laser line along the camera center axis to coincide with the calibration square. Center point to automatically correct the coordinates of the TCP.

Another object of the present invention is to provide a method for calibrating coordinates of a robot arm, setting calibration blocks with various orientations on the calibration target, and using different postures of the robot arm to correct the coordinates of the TCP on the calibration block of the calibration target, so as to automatically Correct the coordinates of the robot arm.

In order to achieve the purpose of the foregoing invention, in the method for calibrating coordinates of a robot arm of the present invention, the camera of the robot arm and the laser device have a fixed relative relationship, and the intersection of the central axis of the camera and the laser of the laser device is adjusted and set as the center point of the tool, and then the calibration is placed. Block, use the camera to extract the image of the calibration block, distinguish the four corners of the calibration block through image processing, obtain the coordinates of the center point of the calibration block by intersecting the diagonal lines, move the center axis of the camera to align the center point of the calibration block, filter and project to the calibration block The laser of the square, obtain the bright endpoints of the two edges of the calibration square, and establish the laser line. While maintaining the center axis of the camera to be aligned with the center point of the calibration square, move the camera to drive the laser line to coincide with the center point of the calibration square. The coordinates of the center point are set as the coordinates of the center point of the control tool.

The calibration square of the method for calibrating coordinates of the robot arm of the present invention is an identifiable color or dark or black square, which is used to absorb the projected laser light and generate bright endpoints with obvious brightness difference between the two edges of the calibration box to obtain two bright endpoints. When coincident, the coordinates of the center point of the calibration block fall on the laser line, and the relative fixed relationship between the camera, the laser device and the center point of the calibration block is recorded when the coincidence is recorded. In addition, when placing the center point of the calibration block, the actual measurement coordinates are used as the actual tool center point coordinates, and the error is obtained by comparing the control tool center point coordinates with the actual tool center point coordinates to correct the coordinates of the robot arm tool center point.

Another embodiment of the present invention is a method for calibrating coordinates of a robot arm. The camera of the robot arm and the laser device have a fixed relative relationship, and then a calibration target with known coordinates and including a calibration block is set at a fixed position relative to the robot arm. The calibration target is The cube is set with a plurality of calibration squares on each plane, and the camera is used to drive the camera to extract the image of the calibration target, and a calibration square on the calibration target is automatically selected. The line intersects to obtain the coordinates of the center point of the calibration block, moves the center axis of the camera to align the center point of the calibration block, filters the laser projected to the calibration block, obtains the bright endpoints on both edges of the calibration block, and establishes laser lines. When the center axis is aligned with the center point of the calibration block, move the camera to drive the laser lines to coincide with the center point of the calibration block. Record the control coordinates of the center point of the calibration block when they overlap, and check that the number of poses of the robot arm does not reach the preset threshold. Change the posture of the robot arm to drive the camera to repeat the calibration steps, and check that when the number of postures of the robot arm reaches the preset threshold, the obtained control coordinates of the center point of the calibration block are calculated by the matrix sequence, and the calibration parameters are obtained to correct the coordinates of the robot arm.

Description of drawings

Fig. 1 is the schematic diagram of the coordinate correction of the robot arm of the present invention;

Fig. 2 is the calibration block image diagram extracted by the camera of the present invention;

Fig. 3 is the schematic diagram of the moving laser line of the present invention;

4 is an image diagram of the center point of the laser line coincident with the calibration square of the present invention;

Fig. 5 is the flow chart of the method for coordinate correction of the robot arm of the present invention;

6 is a schematic diagram of a robot arm calibrating coordinates according to another embodiment of the present invention;

FIG. 7 is a flowchart of a method for calibrating coordinates of a robot arm according to another embodiment of the present invention.

Symbol Description

1 Robot arm

2-axis arm

3 Actuators

4 base

5 terminator

6 cameras

7 Laser device

8 Work Shelves

9 Calibration block

10 Laser

11 Center axis

12 images

13 Four Corners

14 center point

15 Laser Lines

16 bright points

17 Calibration target

TCP Tool Center Point

Detailed ways

Regarding the technical means and effects adopted by the present invention in order to achieve the above-mentioned objects, preferred embodiments are now given and described below with reference to the accompanying drawings.

Please refer to FIG. 1 and FIG. 2 at the same time, FIG. 1 is a schematic diagram of a robot arm correcting TCP coordinates according to the present invention, and FIG. 2 is a calibration block image diagram extracted by a camera of the present invention. In FIG. 1 , the robot arm 1 of the present invention is formed by a plurality of shaft arms 2 and actuators 3 in series, one end is fixed on the base 4 to form the coordinate M of the robot arm, and the other end is connected to the end device 5 at the end of the shaft arm 2, and the end The controller 5 can calculate the coordinates at the coordinate M of the robot arm by recording the rotation of the actuator 3 . A camera 6 and a laser device 7 are arranged on the end device 5 , and the camera 5 and the end device 5 have a fixed relative relationship, so the coordinates at the coordinate M of the robot arm can also be calculated. The robot arm 1 drives and extracts the image of the calibration block 9 on the work frame 8 (see FIG. 2 ), and the laser device 7 is used for emitting a planar scanning laser 10 . The camera 6 and the laser device 7 of the present invention have a fixed relative relationship, and the intersection of the central axis 11 of the camera 6 and the laser device 7 scanning laser 10 is adjusted and set as the tool center point (TCP).

Since the intersection TCP of the virtual central axis 11 of the camera 6 and the scanning laser 10 of the laser device 7 cannot be visually determined, the TCP of the robot arm cannot be directly determined visually. In order to obtain the coordinates of the robot arm TCP in the present invention, a calibration block 9 is placed on the work frame 8, and the calibration block 9 can be a black square. Although the calibration block 9 in this embodiment is illustrated by a black square, it includes but is not limited to black squares. Squares, any squares of easily recognizable color or dark color and easy to determine the center point, all belong to the scope of the invention of the present invention.

The robot arm 1 of the present invention drives the camera 6 to extract the image 12 of the calibration block 9 on the work frame 8, such as the image 12 of the calibration block 9 in FIG. The camera 6 that knows the coordinates can calculate the coordinates of each point of the image 12 . Since the black calibration block 9 in the image 12 is significantly different from the surrounding light-colored environment, the calibration block 9 can be quickly distinguished through image processing, and the coordinates of the four corners 13 of the calibration block 9 are determined. For the center point 14, the coordinates of the center point 14 of the calibration block 9 are obtained. Next, the camera 6 is driven by the robot arm 1 so that the center axis 11 of the camera 6 is aligned with the center point 14 of the calibration block 9 .

In FIG. 2, the laser device 7 emits a planar scanning laser 10, and forms a laser line 15 across the calibration square 9 in the image 12. The laser line 15 of the laser is relatively bright, but the laser line 15 projected on the black calibration square 9, The black calibration block 9 absorbs the light, and the brightness is relatively dim, so the bright endpoints 16 with obvious brightness differences are produced on the two edges of the calibration block 9, in order to obtain the coordinates of the two bright endpoints 16, and the straight line formed by the two bright endpoints 16 can be used. Position the laser line 15.

Please refer to FIG. 3 and FIG. 4 at the same time, FIG. 3 is a schematic diagram of the moving laser line according to the present invention, and FIG. 4 is an image diagram of the laser line overlapping the center point of the calibration square according to the present invention. In FIG. 3, while keeping the center axis 11 of the camera 6 aligned with the center point 14 of the calibration block 9, the robot arm drives the camera 6 to move the laser device 7 along the center axis 11, so that the laser line 15 emitted by the laser device 7 moves toward the calibration block. 9 center point 14. In FIG. 4 , when the coordinates of the center point 14 fall on the laser line 15 , it can be confirmed that the laser line 15 coincides with the center point 14 . Then record the relative fixed relationship between the camera 6, the laser device 7 and the center point 14, and use the camera 6 to extract the focal length of the image 12, the center and the pixel of the image 12, calculate the coordinates of the center point 14, and set the coordinates of the center point 14 as TCP coordinate of.

The aforementioned coordinates of the TCP are the control TCP coordinates obtained from the control of the robot arm. If the robot arm has been calibrated, the control TCP coordinates are the actual TCP coordinates. Otherwise, when the calibration block 9 is placed on the work frame 8, the coordinates of the center point 14 of the calibration block 9 can be actually measured, and the coordinates of the center point 14 are used as the actual TCP coordinates, and the comparison between the TCP coordinates and the actual measured center point 14 can be used. Coordinates, obtain the error between the control TCP coordinates and the actual TCP coordinates, and further correct the coordinates of the robot arm TCP.

As shown in FIG. 5 , it is a flowchart of a method for correcting coordinates of a robot arm according to the present invention. The detailed steps of the method for correcting the coordinates of the robotic arm of the present invention are described as follows: in step S1, when the TCP is started to be corrected, the camera of the present invention and the laser device have a fixed relative relationship, and the center axis of the camera is adjusted and set. The intersection point of the laser device scanning laser Be TCP; Step S2, place the calibration block; Step S3, utilize the camera to extract the image of the calibration block; Step S4, distinguish the calibration block through image processing, obtain the four corners of the calibration block, and utilize the diagonal intersection to obtain the center point coordinates of the calibration block; Then step S5, move the center axis of the camera to align the center point of the calibration square; Step S6, filter the laser light emitted by the laser device on the calibration square, obtain bright endpoints on both edges of the calibration square, and establish laser lines; Step S7, maintain the camera When the center axis of the calibration block is aligned with the center point of the calibration block, move the camera to drive the laser line, so that the laser line coincides with the center point of the calibration block; step S8, record the relative fixed relationship between the camera, the laser device and the center point when they overlap, and the center point The coordinates are set as the control TCP coordinates; in step S9, the control TCP coordinates are compared with the known actual coordinates of the center point of the calibration block, and correction and compensation are performed; then in step S10, the TCP coordinate correction is ended.

Therefore, the method of the present invention for calibrating the coordinates of the robot arm can use the camera on the robot arm to extract the image of the calibration block, and quickly identify the coordinates of the four corners of the calibration block through the obvious difference between the color of the calibration block and the environment. The intersection point calculates the coordinates of the center point, so that the robot arm automatically moves the center axis of the camera to quickly align the center point. Then use the dark calibration square to filter the laser to obtain the bright endpoints of the two edges of the calibration square to form a laser line, and then let the camera guide the laser line along the central axis to coincide with the center point of the calibration square. The TCP coordinates are compared with the actual measured center point coordinates to further achieve the purpose of automatically correcting the TCP coordinates.

As shown in FIG. 6 , it is a schematic diagram of a method for calibrating coordinates of a robot arm according to another embodiment of the present invention. This embodiment basically utilizes the technique of correcting the TCP coordinates in the previous embodiment to correct the coordinates of the robot arm. In order to simplify the description, the same part numbers are used for the same components as in the previous embodiment, which will be described together first. In this embodiment, a plurality of calibration blocks 9 are respectively set on each azimuth of the calibration target 17 , and then the calibration target 17 is set at a fixed position relative to the base 4 of the robot arm 1 and has known robot arm coordinates. Although the calibration target 17 in this embodiment is illustrated by a cube, it includes and is not limited to a cube, such as a polyhedron.

Next, the robot arm 1 is calibrated, the camera 6 is driven by the robot arm 1 to extract the image of the calibration target 17, a calibration block 9 on the calibration target 17 is automatically selected, and the TCP calibration step of the previous embodiment is performed, that is, the selected calibration block 9 is extracted. Image, identify the four corners of the calibration block 9, determine the center point of the calibration block 9 by the diagonal line, automatically move the central axis 11 of the camera 6 to align the center point of the calibration block 9, and then filter the laser device 7 to project the laser to form a laser line. The camera 6 moves along the central axis 11, and drives the laser line to coincide with the center point of the calibration block 9, which is regarded as the coincidence of the TCP and the center point, so as to obtain the controlled TCP coordinates as the control coordinates of the center point of the selected calibration block 9. . Then change the posture of the robot arm 1, select another calibration block 9, repeat the aforementioned TCP correction steps, until the number of postures of the changed robot arm 1 reaches the preset threshold, and then use the control coordinates of the center point of the calibration block 9 obtained above, For example, the matrix sequence calculation in the prior art such as Jacobian Matrix is performed to obtain the parameters of coordinate correction, so as to complete the correction of the coordinates of the robot arm.

As shown in FIG. 7 , it is a flowchart of a method for calibrating coordinates of a robot arm according to another embodiment of the present invention. The detailed steps of the method for correcting the coordinates of the robotic arm of the present embodiment are described as follows: at step T1, start to correct the coordinates of the robotic arm; step T2, the known coordinates and the calibration target comprising the calibration block are set at a fixed position relative to the robotic arm; step T3, automatically select a calibration block on the calibration target to perform TCP calibration; step T4, record the control coordinates of the center point of the selected calibration block; then step T5, check whether the number of postures of the robot arm has reached the preset threshold? If the number of postures does not reach the preset threshold, then go to step T6, change the posture of the robotic arm, then go back to step T3 to repeat the steps, if the number of postures reaches the preset threshold, then go to step T7, the control of the center point of the obtained calibration block Coordinates, perform matrix sequence calculation, and obtain correction parameters to correct the coordinates of the robot arm; then in step T8, end the coordinate correction.

Therefore, according to the method for calibrating coordinates of the robot arm of the present invention, calibration blocks with various orientations can be set on the calibration target, and the calibration blocks of the calibration target can be calibrated by using different postures of the robot arm to correct the coordinates of the TCP, and then the obtained calibration block The coordinates of the center point are calculated by using the matrix sequence to achieve the purpose of automatically correcting the coordinates of the robot arm without contact.

The above are only for the convenience of describing the preferred embodiments of the present invention, the scope of the present invention is not limited to these preferred embodiments, any changes made according to the present invention, without departing from the spirit of the present invention, are It belongs to the protection scope of the present invention.

Claims (10)

1. A method for correcting coordinates of a robot arm is provided, which enables a camera of the robot arm and a laser device to have a fixed relative relationship, and adjusts and sets an intersection point of a central axis of the camera and a laser of the laser device as a tool central point, and the method comprises the following steps:

placing a calibration block;

extracting an image of the calibration square by using a camera;

distinguishing four corners of the calibration square block through image processing, and obtaining the coordinates of the central point of the calibration square block by utilizing diagonal intersection;

the central axis of the mobile camera is aligned with the central point of the calibration square;

filtering the laser projected to the calibration square block, obtaining bright end points of two edges of the calibration square block, and establishing a laser line;

moving the camera to drive the laser line to coincide with the central point of the calibration square under the condition that the central axis of the camera is aligned with the central point of the calibration square;

and setting the coordinate of the central point of the calibration block when the calibration block is superposed as the coordinate of the central point of the control tool.

2. The method of claim 1, wherein the calibration square is a black square.

3. The method of claim 1, wherein the calibration block absorbs the projected laser light to generate bright points with distinct brightness difference at two edges of the calibration block, so as to obtain two bright points.

4. The method of claim 1, wherein the coordinates of the center point of the calibration square that coincide with each other fall on a laser line.

5. The method of claim 4, wherein the center points of the camera, the laser device and the calibration block are fixed relative to each other when they are overlapped.

6. The method as claimed in claim 1, wherein the coordinates of the center point of the calibration block are measured as the coordinates of the center point of the actual tool, and the coordinates of the center point of the control tool are compared with the coordinates of the center point of the actual tool to obtain an error, so as to calibrate the coordinates of the center point of the tool of the robot.

7. A method for calibrating coordinates of a robot arm, wherein a camera of the robot arm and a laser device have a fixed relative relationship, the method comprising the steps of:

setting a calibration target having known coordinates and containing a calibration square in a fixed position relative to the robot arm;

driving a camera to extract an image of a calibration target by utilizing the posture of the robot arm, and automatically selecting a calibration square block on the calibration target;

distinguishing four corners of the calibration square block through image processing, and obtaining the coordinates of the central point of the calibration square block by utilizing diagonal intersection;

the central axis of the mobile camera is aligned with the central point of the calibration square;

filtering the laser projected to the calibration square block, obtaining bright end points of two edges of the calibration square block, and establishing a laser line;

moving the camera to drive the laser line to coincide with the central point of the calibration square under the condition of keeping the central axis of the camera aligned with the central point of the calibration square;

recording the control coordinate of the central point of the calibration square block when the calibration square block is superposed;

and when the gesture number of the robot arm reaches a preset threshold value, performing matrix array calculation on the acquired control coordinate of the central point of the calibration block to acquire a correction parameter so as to correct the coordinate of the robot arm.

8. The method for calibrating coordinates of a robot arm as claimed in claim 7, wherein the camera is driven to repeat the calibrating step by changing the pose of the robot arm when the number of poses of the robot arm is not less than the predetermined threshold.

9. The method of claim 7, wherein the calibration target is a cube.

10. The method of claim 9, wherein the calibration target has a plurality of calibration blocks on each orientation plane.

Images