JP2009257774A - Marker for vision sensor and method for making actuator recognize workspace - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for making an actuator recognize a workspace which enables measurement of three-dimensional coordinates of each marker and makes the actuator recognize the workspace, while making measuring equipment minimal, even when a work plane is somewhat inclined. <P>SOLUTION: A robot arm 1 driven by the actuator 2 includes a vision sensor 4. First to third spherical markers 5-7 disposed on the work plane 20a are photographed by the vision sensor 4 and thereby the actuator 2 is made to recognize the workspace. Accordingly, even when the work plane 20a is somewhat inclined, the three-dimensional coordinates of each of the spherical markers 5-7 can be measured and the actuator 2 can be made to recognize the workspace, while the measuring equipment is made minimal. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention includes a vision sensor marker that is photographed and recognized by a vision sensor and a driving body driven by an actuator, for example, a vision sensor at the tip of a robot arm, and three visions arranged on a work plane by the vision sensor. The present invention relates to a working space recognition method for photographing a sensor marker and causing an actuator to recognize the working space.

In general, an actuator that drives a robot arm or the like has a unique work space. Therefore, before performing work by the robot arm, it is necessary to convert the work space unique to the actuator into an actual work space.
As shown in FIG. 4, the conventional method for recognizing the working space for the actuator has a shape in which first to third markers 31 to 33 marked + are connected to the markers 31 to 33 on the outer peripheral portion of the work plane 20 a. Three points are arranged so as to draw a right triangle in plan view, and it is necessary for the actuator to recognize the three-dimensional coordinates (X, Y, Z) of the first to third markers 31-33. In this case, it is a condition that the posture of the vision sensor 4 provided at the tip of the robot arm is not changed when the three-dimensional coordinates of the markers 31 to 33 are recognized.
Specifically, the vision sensor 4 provided at the tip of the robot arm is moved above the first marker 31 by an actuator, and the lengths of the vertical lines and horizontal lines of the first marker 31 by the vision sensor 4 are measured. The posture and position of the vision sensor 4 whose values respectively match the reference value are calculated, and then the vision sensor 4 is sequentially moved by the actuator by the actuator without changing the posture of the vision sensor 4. 33, and the position of the vision sensor 4 where the measured values of the lengths of the vertical and horizontal lines of the second and third markers 32 and 33 by the vision sensor 4 coincide with the reference value is calculated. The three-dimensional coordinates of the first to third markers 31 to 33 based on the vision sensor 4 of the same posture are calculated and recognized by the actuator. There is a need to recognize the work space to the mediator.

However, the vision sensor 4 is moved by the actuator from above the first marker 31 to above the second marker 32 without changing its posture, and the vertical and horizontal lines of the + mark of the second marker 32 by the vision sensor 4 are moved. In order to make the measurement values of the lengths coincide with the reference values, it is necessary to arrange the vision sensor 4 in the vertical direction with respect to the second marker 32. However, the work plane 20a is slightly inclined with respect to the horizontal direction. In such a case, it is impossible to dispose the vision sensor 4 in the direction perpendicular to the second marker 32 without changing the posture of the vision sensor 4.
In short, in the conventional method of recognizing the working space for the actuator, the vision sensor 4 is placed above the first to third markers 31 to 33 marked + without changing the posture of the vision sensor 4. It was impossible to calculate the three-dimensional coordinates of the first to third markers 31 to 33 by arranging the markers 31 to 33 in the vertical direction.
Therefore, in the conventional working space recognition method, as shown in FIG. 4, the first to third markers 31 to 31 are photographed by photographing only the intersections of the + marks of the first to third markers 31 to 33 with the vision sensor 4. 33 two-dimensional coordinates are calculated, and the laser length measuring device 30 is arranged above the first to third markers 31 to 33, and the height from the first to third markers 31 to 33 to the laser length measuring device 30 is calculated. By measuring the length, the three-dimensional coordinates of the first to third markers 31 to 33 are calculated, and finally the actuator is made to recognize the working space.

In Patent Document 1, when teaching a mark reading work point, the photographed video input device is moved right above the mark to teach this position, and on the screen of the mark captured by the photographed video input device. A position correction method for a robot equipped with an automatic guided vehicle that stores a coordinate position together with a work operation is disclosed.
JP 63-120088 A

As described above, in the conventional method for recognizing the working space for the actuator, as shown in FIG. 4, the markers 31 to 33 marked + are used as vision sensor markers. In order to measure the three-dimensional coordinates, it is necessary to employ the laser length measuring machine 30, and the work such as fine adjustment of each measuring device becomes complicated, and the equipment cost increases, and there are many problems.
Moreover, in the position correction method of the robot equipped with the automatic guided vehicle according to the invention of Patent Document 1, only the two-dimensional coordinates of the marker are calculated, and the above-described problem cannot be solved.

  The present invention has been made in view of the above points, and provides a vision sensor marker in which a captured image has a similar shape no matter which direction is taken by a vision sensor, and the work plane is slightly inclined. Another object of the present invention is to provide a work space recognition method for an actuator that can measure the three-dimensional coordinates of each marker with a minimum number of measuring devices and cause the actuator to recognize the work space.

In order to solve the above problem, the vision sensor marker of the present invention is configured as a spherical marker.
Accordingly, it is preferable that the captured image becomes a circular similar shape, regardless of the direction from which the spherical marker is imaged by the vision sensor.
According to another aspect of the present invention, there is provided a method for recognizing a working space of an actuator, wherein a driving body driven by the actuator includes a vision sensor, and the vision sensor photographs and recognizes three markers arranged on a work plane. A method for recognizing a working space is characterized in that a spherical marker is used for each marker.
As a result, it is possible to measure the three-dimensional coordinates of the center of each spherical marker without changing the attitude of the vision sensor without employing a laser length measuring instrument as in the past, and to make the actuator recognize the work space. be able to.
Various aspects of the workspace recognition method for the vision sensor marker and actuator of the present invention and their functions will be described in detail in the section of the aspect of the invention below.

(Aspect of the Invention)
In the following, some aspects of the invention that can be claimed in the present application (hereinafter sometimes referred to as “claimable invention”) will be exemplified and described. In addition, each aspect is divided into a term like a claim, it attaches | subjects a number to each term, and is described in the format which quotes another term as needed. This is for the purpose of facilitating the understanding of the claimable invention, and is not intended to limit the combinations of the constituent elements constituting the claimable invention to those described in the following sections. In other words, the claimable invention should be construed in consideration of the description, embodiments, etc. accompanying each section, and as long as the interpretation is followed, there may be embodiments in which other constituent elements are added to the aspects of each section. In addition, an aspect in which the constituent elements are deleted from the aspect of each item can be an aspect of the claimable invention. In each of the following items, each of (1) to (3) corresponds to each of claims 1 to 3.

(1) A vision sensor marker comprising a marker that is photographed and recognized by a vision sensor as a spherical marker.
Therefore, in the vision sensor marker of item (1), the captured image has a circular similar shape regardless of the direction from which the spherical marker is imaged by the vision sensor.

(2) A method in which a driving body driven by an actuator is provided with a vision sensor, and the vision sensor captures and recognizes three markers arranged on the work plane, and causes the actuator to recognize the working space, A working space recognition method for an actuator, wherein a spherical marker is used for each of the markers.
Therefore, in the method for recognizing the work space for the actuator of item (2), a spherical marker is used for each marker, so that the captured image is circular regardless of the direction from which the spherical marker is captured. As a result, even if the work plane is slightly inclined, the vision sensor can be arranged so that the measurement radius of each sphere marker and the reference value coincide with each other without changing the attitude of the vision sensor. It is possible to measure the three-dimensional coordinates of the center of the actuator, and to make the actuator recognize the working space.

(3) In the workspace recognition method for the actuator, the vision sensor is moved above the first sphere marker by the actuator, and the measurement radius of the first sphere marker by the vision sensor coincides with the reference radius. A first step of calculating a posture and a position of the vision sensor to be moved, and the vision sensor is moved above the second spherical marker by the actuator in the same posture as the first step, and the second sphere by the vision sensor From the calculation result of each step of the second step of calculating the position of the vision sensor where the measurement radius of the marker matches the reference radius, and the third step of repeating the second step for the third spherical marker, Center coordinates of the 1st to 3rd spherical markers based on vision sensors with the same posture It calculates and characterized in that to recognize the working space to the actuator (2) working space recognition method of the actuator according to claim.
Therefore, in the workspace recognition method for the actuator of item (3), the three-dimensional coordinates of the center of the first to third sphere markers based on the vision sensor of the same posture are calculated from the calculation results of the first to third steps. Thus, the actuator can recognize the work space.

  According to the present invention, it is possible to provide a vision sensor marker in which a captured image has a similar shape regardless of the direction taken by the vision sensor, and minimizes the measuring equipment even if the work plane is slightly inclined. Thus, it is possible to provide a method for recognizing a work space for an actuator that can measure the three-dimensional coordinates of each marker and cause the actuator to recognize the work space.

Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to FIGS.
As shown in FIG. 1, an actuator 2 is connected to a robot arm (driving body) 1, and the robot arm 1 performs a predetermined work on a work table 3 by driving the actuator 2.
As described above, since the actuator 2 has a unique work space, before the robot arm 1 performs the work, the work space unique to the actuator 2 is changed to the actual work space 20 (dotted line in FIG. 2). Range).

A vision sensor 4 is provided at the tip of the robot arm 1. A controller 8 is connected to the actuator 2 and the vision sensor 4.
The control device 8 includes an actuator controller, a vision sensor controller, a position storage unit, and a calculation unit.
The work table 3 has first to third spherical markers 5 to 7 (φ40 mm) according to the embodiment of the present invention such that the shape connecting the centers of the spherical markers 5 to 7 draws a right triangle in plan view. Each is arranged. Each of the spherical markers 5 to 7 is held by a box-like holding body 10. As shown in FIGS. 1 and 2, the range (300 mm × 600 mm) of the work table 3 parallel to the plane formed by the three central points of the first to third spherical markers 5 to 7 is the robot arm 1. In the work plane 20a of the (actuator 2), the height at which the robot arm 1 can be raised from the work plane 20a becomes the work space 20 (range surrounded by a dotted line in FIG. 2) of the robot arm 1 (actuator 2).

Then, a work space recognition method for the actuator 2 according to the embodiment of the present invention will be described.
First, the vision sensor 4 is moved above the work plane 20 a of the work table 3 by the actuator 2, and the first spherical marker 5 is found from the captured image of the vision sensor 4.
Next, the vision sensor 4 is moved above the first sphere marker 5 by the actuator 2, and the radius of the circular first sphere marker 5 photographed by the vision sensor 4 is set as shown in FIG. The measurement is performed by the controller, and the vision sensor 4 is brought close to the first spherical marker 5 so that the measured value matches the reference value. Subsequently, the position of the vision sensor 4 and the position of the vision sensor 4 when the measurement radius of the first sphere marker 5 by the vision sensor 4 matches the reference radius are stored in the position storage means of the control device 8 (first). 1 step).

  Next, as shown in FIG. 2, the vision sensor 4 is moved above the second spherical marker 6 by the actuator 2 while maintaining the posture of the vision sensor 4. Subsequently, as in the case of the first sphere marker 5, as shown in FIG. 3, the radius of the circular second sphere marker 6 photographed by the vision sensor 4 is measured by the vision sensor controller, and the measured value is the reference value. The vision sensor 4 is brought close to the second spherical marker 6 so as to match the value. Subsequently, the position of the vision sensor 4 when the measurement radius of the second sphere marker 6 by the vision sensor 4 coincides with the reference radius is stored in the position storage means of the control device 8 (second step).

  Next, as shown in FIG. 2, the vision sensor 4 is moved above the third sphere marker 7 by the actuator 2 while maintaining the posture of the vision sensor 4. Subsequently, as in the case of the first and second spherical markers 5 and 6, as shown in FIG. 3, the radius of the circular third spherical marker 7 photographed by the vision sensor 4 is measured by the vision sensor controller, The vision sensor 4 is brought close to the third sphere marker 7 so that the measured value matches the reference value. Subsequently, the position of the vision sensor 4 when the measurement radius of the third sphere marker 7 by the vision sensor 4 coincides with the reference radius is stored in the position storage means of the control device 8 (third step).

  Then, the first to third sphere markers based on the vision sensor 4 of the same posture are calculated from the position of the vision sensor 4 when the control device 8 is positioned above the first to third sphere markers 5 to 7. The three-dimensional coordinates of the centers of 5 to 7 are calculated and made to be recognized by the actuator 2, and the actuator 2 is made to recognize the work space 20 where the work is actually performed.

In the embodiment of the present invention described above, when the work space 20 is recognized by the actuator 2, the first to third sphere markers 5 to 7 are employed as the vision sensor markers. No matter which direction is used to photograph .about.7, the photographed image has a similar circular shape.
Therefore, even if the work plane 20a is inclined, the vision sensor 4 can be measured by using the vision sensor 4 with the measurement radius and the reference value of the first to third spherical markers 5 to 7 without changing the posture of the vision sensor 4. Can be arranged at the positions where they match.
As a result, the position of the vision sensor 4 where the measurement radius of the first to third sphere markers 5 to 7 matches the reference value by the vision sensor 4 above the first to third sphere markers 5 to 7 is determined as the vision sensor. 4 can be measured without changing the posture, and by extension, the three-dimensional coordinates of the centers of the first to third spherical markers 5 to 7 based on the vision sensor 4 of the same posture can be calculated and the actuator 2 can be operated. The space 20 can be recognized.

FIG. 1 is a diagram illustrating an apparatus that embodies a workspace recognition method for an actuator according to an embodiment of the present invention. FIG. 2 is a diagram for explaining a working space recognition method for the actuator according to the embodiment of the present invention. FIG. 3 is a view of a photographed image of the sphere marker by the vision sensor. FIG. 4 is a diagram for explaining a conventional workspace recognition method for an actuator. FIG. 5 is a work plan view of an actuator, and is a diagram showing each marker used in a conventional work space recognition method for an actuator.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Robot arm (driving body), 2 Actuator, 3 Work bench, 4 Vision sensor, 5 1st sphere marker, 6 2nd sphere marker, 7 3rd sphere marker, 8 Control device

Claims (3)

  A marker for a vision sensor, characterized in that a marker recognized and photographed by the vision sensor is configured as a spherical marker. A method in which a driving body driven by an actuator is provided with a vision sensor, and the vision sensor captures and recognizes three markers arranged on a work plane, and causes the actuator to recognize an operation space,
A working space recognition method for an actuator, wherein a spherical marker is used for each of the markers. In the method of recognizing the working space for the actuator, the vision sensor is moved above the first sphere marker by the actuator, and the vision radius by which the measurement radius of the first sphere marker by the vision sensor coincides with the reference radius. A first step of calculating the posture and position of
The vision sensor is moved above the second sphere marker by the actuator in the same posture as in the first step, and the position of the vision sensor where the measurement radius of the second sphere marker by the vision sensor coincides with the reference radius A second step of calculating
And a third step of repeating the second step for the third sphere marker;
3. The actuator according to claim 2, wherein a center coordinate of the first to third sphere markers based on a vision sensor having the same posture is calculated from the calculation result of each step, and the operating space is recognized by the actuator. Work space recognition method.

Images