JPS629894A - Method of determining optimum position of holding in shape recognition device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Application The present invention relates to a shape recognition processing method that uses an intelligent robot having a visual sensor to automate transportation in a factory.

(b) Conventional technology In a conventional robot-based transport system in a factory, if the object to be transported changes, the position of the robot's fingers (holding section) must also be changed. In the current high-mix, low-volume production, robot adjustment and so-called set-up time not only deteriorates in production efficiency, but also automation cannot be implemented, resulting in an extremely low productivity.

Furthermore, since humans adjust the holding position of the robot's fingers based on intuition and experience, it is not certain whether the position is the optimal position for holding the object. For this reason, in a transport system that transports objects at high speed, the force generated from acceleration may cause misalignment of the object, or in the worst case, the object may fall off the robot's fingers.

(c) Problems to be Solved by the Invention The present invention takes into consideration the rotational moment generated from the center of gravity position and acceleration of the object, and includes a shape recognition processing device in order to prevent the object from falling and flying out in the transport direction. The holding unit can be automatically controlled by inputting the object's two-dimensional image data from the camera, calculating it on a computer, determining the optimal holding position of the object, and transmitting that position data to the robot body. As a result, transportation can be automated.

(d) Means for Solving Problems In the following, functions and signal flows in each circuit will be explained using a hardware configuration diagram of an apparatus implementing the present invention. Two-dimensional image data of an object is input from the camera 9 in FIG. 2 to the video amplification/synchronization separation circuit 10. In this circuit, a video signal and a synchronization signal included in a video signal are amplified and separated. The video signal is converted into a digital signal by the A/D conversion circuit 11 and stored in the image memory circuit 13.

On the other hand, the synchronization signal is input to the control circuit 12 and used as a timing signal for each circuit. Image memory circuit 13
The stored image data is converted from a digital signal to an analog signal by a D/A conversion circuit 14, and two-dimensional image data of the object is displayed on a CRT 15 of a monitor television.

The two-dimensional image data is input from the input/output interface circuit 16 to a memory circuit 18 inside the microcomputer and is temporarily stored therein.

Next, the CPU circuit of the microcomputer executes calculations based on the program for determining the optimum holding position.

The optimum holding position data obtained through the calculation is transmitted to the robot body 20 from the communication interface circuit 19. Further, each part from the video amplification synchronization separation circuit 10 to the D/A conversion circuit 14 is collectively referred to as a shape recognition device A, and the components from the input/output interface circuit 16 to the communication interface circuit 19 are referred to as a microcomputer B.

Incidentally, a program based on the optimal holding position determining method of the present invention is stored in the memory circuit 18 of the microcomputer B.

Next, the gravity center position determining means and the first, second, and third holding position determining means shown in FIG. 1, which functionally divide the processing of the present invention, will be explained respectively.

In FIG. 1, an object 1 is input as two-dimensional image data by an imaging device 2 and stored in an image memory circuit 3. The gravity center position determining means 4 uses the image data to determine the gravity center position of the object. Figure 3 shows the CR of monitor TV.
This is image data displayed on T15. The center of gravity position yw21 in the horizontal direction corresponds to a position that is half of the area 22 obtained by multiplying the image data of "1" in the horizontal direction from ymin to ymtrx by n. “1” image data refers to digital signals.

This is the part in the High signal state, and corresponds to the diagonally shaded area in FIG. The vertical center of gravity position xw23 is Xm1n
This corresponds to a position that is half of the area 24 obtained by adding up the image data of "1" in each vertical direction from Xmax to Xmax. The center of gravity position (xw, yW) 2 of the object is determined by the above center of gravity determining means.
5 is determined.

Next, a method for determining the optimal holding position of an object using image data and the center of gravity position (xw, yw) will be explained.

FIG. 4 illustrates the result of determining the optimal holding position for the gourd-shaped object.

Position 26 is the first holding position of the robot, and position 2
7 is the second holding position of the robot.

The first and second holding positions are determined by the first holding position determining means 5 and the second holding position determining means 8 in FIG. 1, and the third Φ holding position 28 is as follows. This was determined based on the optimum holding position determination method. As shown in FIG. 4, consider a two-dimensional coordinate system on the paper with X in the horizontal direction and y in the vertical direction. Let the coordinates of the robot's first, second, and third holding positions be (x, y
+), (x2, Y2), (x3, y3) and Sur. Fourth
Figure (7) f, , r2, f3 c shows the magnitude and direction of the force generated in each finger of the robot when the object is being transported in the X direction and the y direction, respectively. In addition, fw indicates the magnitude and direction of the force applied to the center of gravity during transportation,
The coordinates (xw, yw) are the center of gravity position of the object.

Here, the relational expression between the mechanical moment and the resultant force generated in the object while being transported in the X direction is as follows.

Similarly, the relational expression when being conveyed in the y direction is as follows.

fl, f2, and f3 in equations (2) and (4) are nothing but the frictional forces that act on the object and each finger while the robot is transporting the object. Frictional force is determined by the friction coefficient between the object and each finger and the normal force acting on each finger. Therefore, if we take into account the condition that the force of each finger to lift the object is strong enough, fI=f
The condition 2=f3 is established, and the equations (2) and (4) become as shown below.

fW=3 f (°, -f, =f2=f3=f
') - (5) Substituting equation (5) into equation (1), VI0M2+ y3= 3yW (
6), and by transforming equation (6), we get the following equation.

Y3=3Yw (y1+yz) (7
Similar to 1, substituting equation (5) into equation (3) yields the following equation.

x3= 3XW (XI+X2)
(8) Therefore, the third holding position (x3, y3) is
It is determined by equations (7) and (8), and is determined by subtracting the value of the sum of the first and second holding positions from the value three times the center of gravity position.

The optimal holding position for the conveyed object means determining each holding position so as to minimize the mechanical moment in the X and Y directions, and is determined from the first and second holding positions and the center of gravity position. .

Here, the first holding position determining means, the second holding position determining means 8 shown in FIG. 1, and the third holding position determining means 7 using the optimum holding position determining method will be explained using the flowchart shown in FIG. .

Since a robot's fingers can usually only move within a limited area, determining the optimal holding position for an object must also be determined within that limited area.

Furthermore, since the size of the robot's fingers differs depending on the robot, the area of the fingers must also be determined in advance. In process 29 in FIG. 5, a limited area of the robot's finger (hereinafter referred to as a window) and a holding area of the robot's finger are set. In process 30, the outer circumference of the object is y
X on the image data from the position where the maximum value is taken in the direction
Detect a position that has a holding area from the minimum value to the maximum value. In process 31, it is determined whether the holding area can be detected for the y coordinate, and if it can be detected, process 3
3, set the first holding position detectable flag,
The position data is stored in the memory and the process goes to step 35. However, if the holding area position cannot be detected, the process goes to step 32, where it is compared with the y-coordinate of the center of gravity, and if the y-coordinate currently being detected is larger than the y-coordinate of the center of gravity, 1 is subtracted from the y-coordinate. , go to process 30 and repeat the same process. However, if the y-coordinate of the currently detected y-coordinate is smaller than the y-coordinate of the center of gravity, it is assumed that the first holding position cannot be detected.
Go to process 35. Processes 30 to 34 are the first holding position determining means. The second holding position determining means is also substantially similar to the first holding position determining means. In process 35, positions having a holding area are sequentially detected from the minimum value of X to the maximum value of X on the image data from the position where the outer circumferential line of the object takes the minimum value in the y direction. In process 36, it is determined whether the holding area has been detected for the y coordinate currently being detected. If it has been detected, the process goes to process 38, where a second holding position detectable flag is set, and the position is detected. The data is stored in memory and the process goes to step 40. However, if the holding area position cannot be detected, the process goes to step 37, where it is compared with the y-coordinate value of the center of gravity. If the currently detected y-coordinate is smaller than the y-coordinate of the center of gravity, 1 is added to the current y-coordinate, and the process goes to step 35 to repeat the same process.

However, if the currently detected y-coordinate becomes larger than the y-coordinate of the center of gravity, it is assumed that the second holding position cannot be detected, and the process goes to step 40. Processes 35 to 39 are the second holding position determining means.

Next, the third holding position determining means based on the optimum holding position determining method will be explained. In process 40, the first and second holding position detectable flags are checked to determine how many holding positions have been detected. When both the first and second holding positions cannot be detected, the process goes to step 41, and a message indicating that the object cannot be held is sent to the CRT of the monitor television.
15, and the process ends.

Further, when either the first or second holding position can be detected, the process goes to step 42, where one piece of holding position data is transferred to the robot body 20, and the process ends. Further, when both the first and second holding positions can be detected, the process goes to step 43, and the third holding position is determined based on the optimal holding position determination method that minimizes the mechanical moments in the X direction and the y direction. Find the holding position by calculation. In process 44, the third
The holding position may be a very large value, or the first and second
It is determined whether the third holding position is an appropriate position, such as when the third holding position overlaps with the third holding position or the position of a hole in the object. If the third position is an inappropriate position, the process goes to step 45, where the first and second holding position data are transferred to the robot body 20, and the process ends. If the calculated third position is an appropriate position, the process goes to step 46. In this process, if a window is set within the holding area of the object, it is determined whether the third holding position is within the window.

If the third holding position is outside the window, the process goes to step 47 and the position is corrected so that it falls within the window. If the calculated third holding position is within the window,
The process goes to step 48, where the first, second, and third holding position data are transferred to the robot body 20, and the process ends. Processing 40
to 48 are the third holding position determining means.

(e) Effects of the invention By making it possible to automatically determine the holding position of an object according to the present invention, robots can replace the transport process that traditionally relied on human eyes in factories. This has the effect of making it possible to implement complete automation. In addition, the position determined by the optimal holding position determination method is such that the mechanical moment is balanced in each direction, so it is possible for the object to shift due to the force generated by acceleration during transportation, or for the object to fall off. This has the effect of making it possible to carry out transportation with high positioning accuracy.

As a specific application example of the present invention, it is fully applicable to automatically detecting the suction/holding position of conveyed materials by a large crane, and to automatically convey products with complex shapes such as boxes.

[Brief explanation of the drawing]

1 and 2 are a block diagram and a device configuration diagram in which the present invention is divided into functional units, FIG. 3 is a diagram illustrating how to determine the center of gravity position of two-dimensional image data, and FIG. FIG. 5 is a flowchart illustrating the optimum holding position determining method from the first holding position determining means to the third holding position determining means. 9... Camera 10... Video amplification synchronization separation circuit 11... Width conversion circuit 12... Control circuit 13... Image memory circuit 14... D/A conversion circuit 15... CRT 16 ...Input/output interface 17...CPU circuit 18...
・Memory circuit 19...Communication interface 20...
・Robot A...Shape recognition device B...Microcomputer

Claims (1)

[Scope of Claims] 1) An imaging device that obtains two-dimensional image data of an object, an image memory circuit that stores the image data, and an image sensor that corresponds to the image data stored in the image memory device, y
A center of gravity determining means for determining the center of gravity of an object from the position of the cumulative area curve along the axis and the bisecting line of the area surrounded by the x and y axes, and from the position where the outer circumference of the object takes a maximum value in the y direction. a first holding position determining means for finding a holdable position on the object according to the image data along the x direction up to the center of gravity; a second holding position determining means for finding a holdable position on the object corresponding to the image data along the x direction; and a holding position determined by the center of gravity position and the first and second holding position determining means. A third holding position determining means for determining a holding position that minimizes the rotational moment acting on the object being transported based on the position, and an output device that outputs each holding position data to the outside. Holding positioning device. 2) Claim 1, wherein the output device includes a data transmission device for an object transport robot.
The apparatus for determining the optimal holding position of a conveyed object as described in 2. 3) Obtain two-dimensional image data of the object using an imaging device,
This is stored in the image memory circuit, and corresponding to the image data, the position of the bisecting line of the area surrounded by the x-axis and y-axis of the object is set as the center of gravity position (x_w, y_w) of the object, and the outer circumference of the object is From the position where x takes its maximum value in the y direction to the center of gravity, x
The first position on the object that can be held in accordance with the image data along the direction is the first holding position (x_1, y_1), and from the position where the outer circumference of the object takes the minimum value in the y direction to the center of gravity, Corresponding to the image data along the x direction, the first position in the holdable position on the object is changed to the second hold position (
, y_2) and the position calculated by
A method for determining an optimal holding position for a transported object, characterized in that: