KR20080091891A - The automatic calibration method in robot based multi-laser vision system - Google Patents

Abstract

An automatic calibration method of a robot based multi-laser vision system is provided to reduce manufacturing time of products by performing rapid and precise calibration. A calibration jig mounted on an end effector of a square beam of a robot is moved to the initial position(S200). A trigger signal is transmitted from a controller of the robot to a multi-laser vision system(S210). The multi-laser vision system, receiving the trigger signal, captures an image(S220). A center line of laser stripe patterns is detected from the captured image(S230). A character point on an image coordinate system, corresponding to a corner point of a triangle pattern, is extracted from the center line(S240). Noise of the extracted character point is removed(S250). A position value of the end effector is transmitted from the controller to the multi-laser vision system(S260). Calibration is performed using a position value of the noise removed character point and the position value of the end effector(S270). By comparing a calibration error with an allowable error, whether to perform recalibration is determined(S280). A projection matrix is obtained through the performed calibration(S290).

Description

The automatic calibration method in robot based multi-laser vision system

1 is a flowchart illustrating an embodiment of a calibration method of a robot-based multi-laser vision system according to the present invention.

FIG. 2 is a conceptual diagram of a robot-based multi-laser vision system illustrated to describe the embodiment shown in FIG. 1.

3 is a conceptual diagram illustrating a jig for calibration used in the embodiment shown in FIG.

The present invention relates to a calibration method of a robot-based multi-laser vision system.

In general, all metrology systems are calibrated based on each metrology system coordinate system. At this time, the calibration finds an accurate feature point and requires as much data as possible over various areas based on the feature point.

A laser vision system (LVS) includes a light output unit (usually a 'laser diode unit' is used) for emitting light and a camera for capturing light emitted from the light output unit. In addition, a plurality of such laser vision systems are used for one purpose by interlocking with each other. This is called a multi-laser vision system (MLVS).

One robot-based Laser Vision System (LVS) is also commercially available and is being studied with a variety of 3D scanning technologies.

However, such a robot-based multi-laser vision system is difficult to calibrate before the actual measurement work, and it takes a lot of time to calibrate, and new countermeasures have emerged.

The present invention devised to solve the above problems is a robot-based multi-laser vision system for measuring the measurement object on the bottom by using a multi-laser vision system arranged in a line at equal intervals on the square beam of the front and rear movable robots The technical challenge is to provide a way to perform the calibration quickly and accurately.

In order to solve the above technical problem, the present invention provides a robot-based multi-laser vision for measuring a measurement object on the bottom by using a multi-laser vision system arranged in a line at equal intervals on a square beam of a front and rear movable robot. In a calibration method of a system, a calibration jig having a plurality of triangular patterns is mounted on an end effector on a square beam of the robot, and the calibration jig is moved to an initial position. Calibration preparation step to make; A trigger signal transmission step of transmitting a trigger signal for an image capture command from the controller of the robot to the multi-laser vision system; An image capturing step of capturing an image in the multi-laser vision system receiving the trigger signal; A laser strip center line detecting step of detecting a center line of the laser strip from the captured image; A feature point extraction step of extracting feature points on an image coordinate system corresponding to corner points of the triangular pattern from the detected laser strip center line; A feature point noise removing step of removing noise of the extracted feature points; An end effector position value transmission step of transmitting a position value on a Cartesian coordinate system of the end effector from the controller of the robot to the multi-laser vision system; A calibration calculation step of performing calibration based on a least square method using the position values of the feature points from which noise is removed and the position values of the end effector; A calibration error comparison step of determining whether to perform re-calibration by comparing the calculated calibration error with an appropriate tolerance; And a calibration-projection matrix acquiring step of acquiring a calibration-projection matrix through the calculated calibration.

In order to fully understand the present invention, the advantages of the operability of the present invention, and the objects achieved by the practice of the present invention, reference should be made to the accompanying drawings which illustrate preferred embodiments of the present invention and the contents described in the accompanying drawings.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Like reference numerals in the drawings denote like elements.

1 is a flowchart illustrating an embodiment of a calibration method of a robot-based multi-laser vision system according to the present invention. FIG. 2 is a conceptual diagram of a robot-based multi-laser vision system illustrated to describe the embodiment shown in FIG. 1. 3 is a conceptual diagram illustrating a jig for calibration used in the embodiment shown in FIG.

A preferred embodiment of a calibration method of a robot-based multi-laser vision system according to the spirit of the present invention will be described with reference to FIGS. 1 to 3. First, an end effector below a square beam of a robot. And a calibration preparation step (S200) for mounting a calibration jig having a plurality of triangular patterns on the) and moving the calibration jig to an initial position.

In order to understand the calibration method of the robot-based multi-laser vision system of the present invention, first, a device concept related to the robot-based multi-laser vision system to which the present invention can be applied should be described, but will be briefly described with reference to FIG. 2.

2 is a conceptual diagram of a robot-based multi-laser vision system. In the robot-based multi-laser vision system 1, rails 7 are formed at both sides of the bottom surface on which the measurement object is seated, and the rails 7 move forward and backward on the rails 7 above the rails 7. The gantry-shaped robot 3 which can be formed is comprised.

The multi-laser vision system 5 is disposed on a lower surface of the square beam of the robot 3, and a plurality of laser vision systems are arranged in a line at equal intervals. The robot-based multi-laser vision system 1 quickly detects the object to be placed on the bottom surface between the rails 7 by the multi-laser vision system 5 while moving in the rail direction. It is a device that can measure three-dimensional shape.

An end effector 9 is installed on one side of a square beam of the robot-based multi-laser vision system 1, and a calibration jig (not shown) may be mounted at a lower end thereof.

Here, looking at the calibration jig with reference to Figure 3, a jig bracket 52 for fixing the calibration jig 50 and the calibration jig 50 in a two triangular pattern on the side, and It is configured to include a circular socket 54 is coupled to the upper end of the jig bracket 52 is formed to be detachable to the end effector 9 as described in FIG.

In the calibration preparation step (S200) illustrated in FIG. 1, the calibration jig 50 of FIG. 3 is mounted on the robot-based multi-laser vision system 1 of FIG. 2, and the calibration jig 50 is pre-designated. This is the step of moving to calibration positions.

Next, a trigger signal transmission step (S210) of transmitting a trigger signal for an image capture command from the controller of the robot to the multi-laser vision system. The control unit of the robot has a function of generating and transmitting a trigger signal for commanding image capture of a measurement object of a multi-laser vision system in hardware.

Next, an image capture step (S220) of capturing an image in the multi-laser vision system receiving the trigger signal. The multi-laser vision system receiving the trigger signal from the controller of the robot captures the captured image.

Next, a laser strip center line detecting step (S230) of detecting a center line of the laser strip from the captured image. In this case, in the laser strip center line detecting step (S230), since the center line of the laser strip may be detected from the captured image by using various conventional image processing methods, a detailed description thereof may be omitted.

Next, a feature point extraction step (S240) of extracting a feature point on the image coordinate system corresponding to the corner point of the triangular pattern from the detected laser strip center line. Since the corner point of the triangular pattern is related to the shape of the calibration jig mounted on the end effector under the square beam of the robot, in the present embodiment as shown in FIG. It is possible to extract the feature points on the image coordinate system that exist and correspond to the corner points thereof. However, in the case of using a plurality of triangular pattern calibration jig different from the present embodiment in another embodiment, the image coordinate system corresponding to the number of corner points accordingly. The feature points of the image must be extracted.

By the feature point extracting step S240, the feature point position values U and V on the two-dimensional image coordinate system can be calculated.

Next, a feature point noise removing step (S250) of removing the extracted noise of the feature point is included. Since the noise removal method that can be used in this step (S250) can use a conventional noise removal method as appropriate, a detailed description thereof can be omitted. Through this feature point noise removal step S250, the error of the extracted feature point can be corrected correctly.

Next, an end effector position value transmission step (S260) of transmitting a position value on the Cartesian coordinate system of the end effector from the controller of the robot to the multi-laser vision system.

Calibration of the robot-based multi-laser vision system using captured point image values (U, V) on the two-dimensional image coordinate system and the Cartesian coordinate system position of the end effector equipped with the calibration jig In order to calculate, the end effector position value transmission step (S260) is passed. This end effector position value transmission step (S260) may be performed using ordinary LAN or serial communication.

Next, a calibration calculation step (S270) of performing calibration based on a least square method using the position value of the feature point from which the noise is removed and the position value of the end effector is performed.

In this calibration calculation step (S270), calibration is performed based on Least-Square of Equation 1 below.

[Equation 1]

Where s is the magnification (scalefactor) A is STE rinsik calibration parameters (intrinsic calibration parameter), [R , t] is the extreme trim expression calibration parameters ^{(extrinsic calibration parameter), m -} = [uv 1] T, M - = [XYZ 1] means ^T.

Also,

The projection matrix P is

를 의미한다.

Means.

After the calibration is performed by the calibration calculation step (S270), a calibration error comparison step (S280) that compares the next calculated calibration error with an appropriate tolerance to determine whether to perform the recalibration (Scalibration) is included. . In general, since the error of the calibration is large, a step (S280) of comparing the error within an appropriate range, that is, the tolerance E _a with the calibration error E _c for the actually calculated calibration is passed.

At this time, if the calibration error E _c falls within the tolerance E _a , the optimal projection matrix is obtained. In contrast, the calibration error E _c is the tolerance E _a . If out of the, re-calibration (Recalibration) is performed to return to the calibration preparation step (S200), which is the first step in order to perform the calibration again using the remaining data except the calibration points out of the error.

The calibration-projection matrix acquiring step (S290) is finally obtained, after the previous step is repeated at least once, and the calibration-projection matrix is acquired.

Through all the above steps, one optimal 3by4 projection matrix for one laser vision system among the multi-laser vision systems can be obtained. Thus, this embodiment using four laser vision systems as shown in FIG. In the case of the example, all the above-described steps may be performed at least four times to derive a total of four 3by4 projection matrices related to the calibration.

Although the present invention has been described with reference to one embodiment shown in the drawings, this is merely exemplary, and those skilled in the art will understand that various modifications and equivalent other embodiments are possible therefrom. . Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

As described above, according to the calibration method of the robot-based multi-laser vision system of the present invention, a measurement object on the bottom is measured by using a multi-laser vision system arranged in a line at equal intervals on a square beam of a robot that can move back and forth. For robot-based multi-laser vision systems, calibration can be performed quickly and accurately, which has the beneficial effect of reducing product production time and improving product accuracy.

Claims (1)

In a calibration method of a robot-based multi-laser vision system for measuring the measurement object of the bottom using a multi-laser vision system arranged at equal intervals on a square beam of the robot that can move back and forth, A calibration preparation step of mounting a calibration jig having a plurality of triangular patterns on an end effector below the square beam of the robot and moving the calibration jig to an initial position; A trigger signal transmission step of transmitting a trigger signal for an image capture command from the controller of the robot to the multi-laser vision system; An image capturing step of capturing an image in the multi-laser vision system receiving the trigger signal; A laser strip center line detecting step of detecting a center line of the laser strip from the captured image; A feature point extraction step of extracting feature points on an image coordinate system corresponding to corner points of the triangular pattern from the detected laser strip center line; A feature point noise removing step of removing noise of the extracted feature points; An end effector position value transmission step of transmitting a position value on a Cartesian coordinate system of the end effector from the controller of the robot to the multi-laser vision system; A calibration calculation step of performing calibration based on a least square method using the position value of the feature point from which the noise is removed and the position value of the end effector; A calibration error comparison step of determining whether to perform recalibration by comparing the calculated calibration error with an appropriate tolerance; And And a projection matrix obtaining step of obtaining a projection matrix through the calculated calibration. Calibration method for robot-based multi-laser vision systems.

Images