JPH02201206A - Method and apparatus for detecting position - Google Patents

Abstract

PURPOSE:To enable the determining of a position of a part to be detected as a whole at a high accuracy by a method wherein a part to be detected is irradiated with a slit light to take an image thereof in a clear light contrast from other parts to determine an interface point between the at to be detected and the other parts at points and the interface point is used to obtain a position of the part to be detected. CONSTITUTION:A part M to be detected is irradiated with a slit light 6 at points thereof over the part to be detected and other parts and an image of the part M to be detected irradiated with the slit light 6 is taken by a camera means 3. Therefore, the part M to be detected is lighted by a bright slit light 6 at points thereof, which allows the taking of the image clearly at these points. By processing an image signal thus obtained by an image processing means 14, an interface point can be determined accurately between the part M to be detected and other parts at points thereof. The interface point is used to determine a position of the part M to be detected thereby enabling the determining of a position of the part M to be detected as a whole at high accuracy.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a position detection method and an apparatus thereof, and in particular to a method and apparatus for highly accurately determining the position of a detected part of an article to which parts are to be assembled. Regarding.

[Prior art]

2. Description of the Related Art Recently, technology has been widely adopted in automobile factories and the like to automatically assemble various dovetail accessories using assembly robots onto vehicle bodies being transported on transport lines.

In such automatic assembly, the vehicle body itself includes dimensional errors due to the assembly of the vehicle body, and errors also occur in the stopping position when the vehicle is stopped at a predetermined position on the conveyance line. Therefore, in order to assemble parts accurately and efficiently using an assembly robot or an automatic assembly device, it is necessary to detect the position of a detected part of a target workpiece such as a car body with high precision.

゛Therefore, conventionally, as described in JP-A No. 62-12483, when assembling car window glass with an assembly robot, the target workpiece is imaged with a television camera, and the target workpiece is identified from the image information. 2. Description of the Related Art An automatic automobile window glass mounting apparatus is known that detects the amount of positional deviation of a target workpiece, inputs the positional deviation amount to an assembly robot, and corrects control information of the assembly robot.

This automatic mounting device detects the position of a spacer provided on the wind frame of the vehicle body in order to determine the vertical position of the window glass.

[Problem to be solved by the invention]

In the automatic installation device for automobile window glass described in the above publication, the spacer provided on the wind frame is imaged with a television camera, and the position of the entire wind frame is estimated from the position of the spacer. Not only does the installation position of the wind frame include errors, but the overall shape of the wind frame also includes errors that occurred during the vehicle body assembly stage, so the position of the entire wind frame can be estimated using the spacer position as a reference. With this method, it is difficult to determine the position of the entire wind frame with high precision.

Therefore, a method of capturing an image of the entire wind frame with a television camera and determining the position of the entire wind frame from the image information may also be considered.

However, the wind frame and other parts (e.g.
There is a problem in that it is difficult to clearly image the outline of the wind frame using a relatively inexpensive industrial television camera because there is no significant contrast between the window frame and the floor panel (inside the vehicle, etc.).

[Purpose of the invention]

An object of the present invention is to provide a position detection method and apparatus that can reliably and accurately capture a detected part by irradiating a plurality of locations on the detected part with high-intensity slit light and taking images. .

[Means to solve the problem]

The position detection method according to the first claim irradiates a plurality of locations of a detected part with slit light that covers the detected part and other parts, and captures an image of the detected part irradiated with the slit light. By capturing an image with the imaging means and processing the image signal obtained by the imaging means with the image processing means, boundary points between the detected part and other parts at multiple locations of the detected part are determined, and these boundaries are The position of the detected part is determined using points.

A position detection device according to a second aspect includes a slit light irradiation means for irradiating a plurality of locations of a detected part with slit light that covers the detected part and other parts; an imaging means for capturing an image of the detected part; and an imaging means for receiving an image signal from the imaging means, determining boundary points between the detected part and other parts at a plurality of locations of the detected part, and using these boundary points to determine the boundary points between the detected part and other parts. and image processing means for determining the position of the detection section.

[Effect]

In the position detection method according to the first aspect, by moving one slit light beam while irradiating it or by keeping it stationary while irradiating a plurality of slit light beams, a plurality of locations of the detected part are exposed. A slit light is irradiated across the detection part and other parts, and an image of the detected part irradiated with this slit light is captured by the imaging means, so multiple parts of the detected part are illuminated with bright slit light. Therefore, images of these multiple locations can be clearly captured.

By processing the image signal obtained as described above by the image processing means, it is possible to reliably determine the boundary points between the detected part and other parts at a plurality of locations of the detected part. Since the position of the detected part is determined by image processing using the plurality of boundary points, the position of the entire detected part can be determined with high precision.

In the position detection device according to the second aspect, the slit light irradiation means irradiates the slit light to a plurality of locations on the detected portion and the other portions. However, the slit light irradiation means is configured to move while irradiating one slit light, or to remain stationary while irradiating a plurality of slit lights.

The imaging means takes an image of the detected part illuminated with the slit light, but as far as the plurality of locations illuminated with the slit light are concerned, the detected part is sufficiently brightly illuminated by the slit light, so a clear image can be obtained. It will be done.

The image processing means receives an image signal from the imaging means, processes the image signal, determines boundary points between the detected part and other parts at a plurality of locations of the detected part, and uses these boundary points to determine boundary points between the detected part and other parts. Find the position of the detected part.

Regarding the multiple locations irradiated with the slit light, the boundaries between the detected part and other parts are clear, so multiple boundary points can be reliably determined, and these multiple boundary points can be used to identify the detected part. Since the position is determined, the entire position of the detected part can be determined with high precision.

In addition, since images are captured by irradiating slit light onto a plurality of locations on the detection target, the number of image signals does not increase unnecessarily, the burden on the image processing means is reduced, and the processing time can be shortened. Moreover, the power consumption in the slit light irradiation means is also small.

〔Effect of the invention〕

According to the position detection method according to the first claim, the above [effect]
As explained in section 2, by irradiating slit light onto multiple parts of the detected part, an image is taken with a clear contrast between the detected part and other parts, and from the image signal, multiple parts of the detected part are imaged. It is possible to reliably find the boundary point between a part and other parts. Since the position of the detected part is determined using these plurality of boundary points, the position of the entire detected part can be determined with high precision.

According to the position detection device according to the second claim, the above [action]
As explained in the section, the same effects as in the first claim can be obtained.

In addition, since images are captured by irradiating multiple locations on the detected part with slit light, the number of image signals does not increase unnecessarily.
The burden on the image processing means is reduced and the processing time is also shortened.

Moreover, the power consumption in the slit light irradiation means is also small.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

This embodiment applies the present invention to a window opening position detection method and device for detecting the position of the rear window opening of a vehicle body when automatically assembling a rear window glass to a vehicle body using an assembly robot in an automobile factory. It is a matter of the case.

The window opening detection device and window opening detection method will be explained based on the drawings.

As shown in Fig. 1 and Fig. 2, the car body M is conveyed by a conveyor along an assembly line where various repair parts and equipment are assembled to the car body M in an automobile assembly factory, and the rear window glass is assembled. At the stage run (hereinafter referred to as a glass assembly station), the rear window glass 1 is assembled onto the vehicle body M.

The glass assembly station is provided with one or more limit switches 5 (see FIG. 5) for detecting when the vehicle body M has been conveyed to a substantially predetermined position and stopped.

Furthermore, television cameras 3 are installed at predetermined positions behind and above the vehicle body M that has stopped at approximately the predetermined positions, so as to face the corners 2a to 2d of the rear window opening 2 of the vehicle body M, respectively.
Four sets of projectors 4 are fixedly provided.

As shown in FIG. 3, each projector 4 is designed to simultaneously irradiate about 5 to 10 powerful slit lights 6 in parallel and spaced apart from each other by a predetermined distance toward the corresponding corner portions 2a to 2d. Each set of slit light beams 6 is directed in a substantially diagonal direction of the rear window opening 2. Therefore, each set of slit light beams 6 reliably intersects the outer peripheral contour line 7 (see FIG. 4) of the rear window opening 2. At the peripheral edge of the rear window opening 2, as shown in FIG.
A mounting wall part 10 of a predetermined width for fixing the outer peripheral part of the rear window glass l extends from the lower peripheral edge in parallel to the rear window glass 1, and its inner peripheral surface forms an inner peripheral contour line 8.

The glass assembly station is equipped with an articulated robot 11 that automatically assembles the rear window glass l into the rear window opening 2, and the rear window glass 1 is placed in a predetermined position near the robot 11 in a predetermined posture. The rear window glass 1 is attached to the rear window opening 2 by sucking the outer surface of the rear window glass 1 using four suction cups 12 of the robot 11's hand.

The robot controller 13 of the robot 11 stores information such as the position information of the rear window opening 2, the supplied position information of the rear window glass l in a predetermined posture, and the movement path for assembling the rear window glass l. As with the robot, the robot controller 13 receives position correction information for correcting the position of the rear window opening 2 from the image processing control unit 14, which will be described later, for each vehicle body M. The robot 11 is controlled to correct the positional information of the opening 2 and automatically assemble the rear window glass 1.

The limit switch 5 for detecting the vehicle body M, the four sets of television cameras 3 and floodlights 4, and the robot controller 13 are connected to an image processing control unit 14.

Next, the overall configuration of the control system of the window opening position detection device will be explained based on FIG. 5.

The image processing control unit 14 includes a CPU 1
5 (central processing unit) 16, ROM 16 (read-only memory), and RAM 17 (random access memory), and various buses 18 (control bus, address bus, data bus) to the computer. The connected I10 port 19 (input/output interface), the drive circuit 4a connected to the I10 port 19 and connected to each of the four sets of floodlights 4,
An operation panel 23 for operating the computer is connected to the I10 baud) 19, and the computer is connected to the I10 port 19.
It is connected to the robot controller 13 via a port 19.

When each drive circuit 4a receives a control signal from the computer and outputs a drive current to the corresponding light projector 4, the light projector 4 emits slit light 6.

The television camera 3 is equipped with a solid-state imaging device in a matrix of, for example, 480 rows by 512 columns, similar to a normal industrial mono-color television camera, and each television camera 3 has a corresponding one shown in FIG. The regions 2A to 2D are imaged.

Each of the image signal processing and storage units 20 includes an A/D converter that A/D converts one screen worth of analog image signals captured by the corresponding television camera 3 in time series, and a digital signal from the A/D converter. An interface that receives image data and converts it into binary digital image data of "l" and "0" by dividing it at a predetermined threshold and supplies it to the image data memory, and one screen supplied from the interface. Each image signal processing storage unit 20 is connected to the I10 port 19 by a bidirectional signal line 21, and has an 8-bit memory for outputting image data to a computer. The data bus 22 connects the I10 port 19 to the I10 port 19. The image signal processing storage unit 20 is a well-known existing unit configured to binarize the image signal obtained by the television camera 3 into a digital signal and store it, and supply the stored image data to a computer. It is similar to that.

The above ROM 16 contains four sets of floodlights 4 and a television camera 3.
Then, the image signal processing and storage unit 20 is operated to capture images of the four corners 2a to 2d of the rear window opening 2, and the position of the rear window opening 2 is calculated using the image data. A window opening detection control program is input and stored in advance for determining the position correction amount of the opening 2 with respect to the reference position and outputting a position correction amount signal to the robot controller 13, and this control program includes the following subroutines. ing.

(1) An imaging control subroutine that operates four sets of floodlights 4 based on signals from limit switches 5, and outputs control signals to four sets of television cameras 3 to operate imaging for a predetermined time; (2) 4 It outputs a control signal to the image signal processing storage unit 20 in synchronization with the activation of the set of television cameras 3 to operate it, and reads the image data in the image data memory to the CPU 15 via the I10 port 19. Image data control subroutine, (3) Boundary point for reading image data sequentially from each of the four sets of image signal processing storage units 20 and calculating the boundary point where the slit light 6 intersects with the outer peripheral contour line 7 in FIG. Arithmetic subroutine, (4) 4
Based on the data of the boundary points of the set of corner parts 2a to 2d, the robot calculates the position correction amount of the rear window opening 2 to be detected with respect to the rear window opening 2 at the reference position (position when the error is zero). Correction control subroutine output to controller 13 The RAM 17 is provided with a memory, a counter, a pointer, a flag, etc. necessary for performing window opening detection control.

Next, the above-mentioned rear window opening detection control routine will be explained with reference to the schematic flowchart shown in FIG.

However, SI (1=1, 2.3...) in the figure indicates each step.

Control is started by turning on the start switch on the operation panel 23, and when the vehicle body M is carried in and stopped at approximately a predetermined position and the limit switch 5 is turned on (Sl), the process moves to S2 and processes four sets of image signals. Initial settings such as clearing the image data memory of the storage unit 20 and various memories of the RAM 17 are performed, and then the four sets of projectors 4 and television camera 3 are activated to perform imaging (
S3).

In parallel with this imaging, each image signal processing and storage unit 20 digitizes and binarizes the image signal input from the television camera 3 and stores it in the image data memory. When the storage of image data for one screen is completed (
S4), the four sets of projectors 4 and television cameras 3 are stopped (35).

Next, the soft counter i provided in the RAM 17 is set to i=1 (S6), and then the image data memory of the image signal processing storage unit 20 corresponding to the first corner portion 2a is The image data is read in a predetermined order (S7), and then the boundary point of the first corner portion 2a is calculated from the image data of the first corner portion 2a.

In this case, for example, xy including four sets of corner portions 2a to 2d
Using the coordinate plane as a reference plane, as shown in FIG. 4, regarding the corner portion 2a, the first slit beam 6a is excluded from the image data, and line segments 6bo to 6bo representing the centers of the second to fifth slit beams 6b to 6e are drawn. 6e.

The lower end points of the four line segments 6bo to 6eo are determined as boundary points P2 to P. The xy coordinate values of the four boundary points P2 to P obtained as described above are stored in the memory of the RAM 17.

Next, the counter i is incremented (S9), and then it is determined whether i=5 or not. If no, S7 and 38 are repeated to find the boundary point of the second corner portion 2b and store it in the memory, and the same goes on. Repeat from the first corner part 2a to the fourth corner part 2a.
When the boundary points of the corner part 2d are calculated and all their xy coordinate values are stored, i=5, so the process moves to Sll, and the xy coordinates of the boundary points of the first corner part 2a to the fourth corner part 2d are Based on the coordinate values and the data of the outer peripheral contour of the rear window opening 2 at the reference position on the xy coordinate plane stored in advance in the program, the rear window to be detected is detected for the rear window opening 1 2 at the reference position. The amount of positional deviation of the opening 2, that is, the amount of positional correction consisting of the amount of X-axis correction, the amount of X-axis correction, and the amount of rotational correction around the two axes is calculated (311), and this positional correction amount is sent to the robot controller 1.
3 (312), and then the limit switch 5 turns O.
It is determined whether or not it is FF (S13), and if it is OFF, the process moves to 31, and control regarding the next vehicle body M is executed in the same manner as above.

Next, it is also possible to partially change the window opening detection device and its detection method as follows.

[1] As shown in FIG. 7, instead of the four sets of floodlights 4, there is a floodlight that irradiates one slit light 6A that extends over the entire width of the rear window opening 2 in the vertical direction. Using a projector that moves at a predetermined speed, images are taken by four sets of television cameras 3 at predetermined minute intervals, and the image data is updated in the image data memory, while the rlJ data corresponding to the slit light 6A is not updated. By storing the image data without starting, it is possible to finally store image data similar to that in the above embodiment in the image data memory.

[2] As shown in FIG. 8, four sets of floodlights 4 are provided in the same manner as in the previous embodiment, and one television camera 3 on the upper stage is moved in the vehicle width direction to illuminate the corners 2a and 2b of the upper stage. By making it possible to sequentially image the areas 2A and 2B, and by moving one television camera 3 in the lower tier in the vehicle width direction, it is configured to be able to sequentially image the areas 2C and 2D in the lower corner portions 2C and 2d.

[3] As shown in FIG. 9, a fixed type camera is used to image the area 2U at the center of the upper edge, the area 2S at the center of the lower edge, and the areas 2L and 2R at the center of the left and right side edges of the rear window opening 2. Images are taken by providing four sets of television cameras and a total of four sets of projectors that irradiate the regions 2U to 2R with slit lights 6B and 6C in the directions shown in the figure, respectively.

[4] As shown in FIG. 10, a floodlight is provided that simultaneously illuminates the entire rear window opening 2 with a plurality of slit lights 6D, and one mobile television camera on the upper stage is used to illuminate the upper three-part area 2A, for example. 2U and 2B are sequentially imaged, and one mobile TV camera in the lower row is used to capture images of, for example, three divided areas 2C and 2B in the lower row.
Image data of the entire rear window opening 2 is obtained by sequentially capturing images of 2S and 2D.

[5] As shown in Figure 11, use the projector to illuminate areas 2A and 2C.
A slit light 6E moving at a predetermined speed is irradiated onto the top and bottom 2.
A mobile TV camera is used to capture two people in an area at predetermined minute intervals.
2C, update the image data in the same way as in [1] above, and store the data of "1" corresponding to the slit light 6E without updating, thereby creating the area 2A.
The image data of 2C is stored in the image data memory, and the imaging and image data storage of areas 2U and 2S and areas 2B and 2D is similarly performed.

Although the above embodiment describes the case where the present invention is applied to a detection device and a detection method for detecting the position of the rear window opening 2, it goes without saying that the present invention can be similarly applied to the front window opening. Therefore, the present invention can be applied not only to openings but also to various position detection devices and methods that detect the contours of various objects or parts of objects to be detected and detect their positions. I can do it.

[Brief explanation of the drawing]

The drawings relate to an embodiment of the present invention, and FIG. 1 is a perspective view of the rear body of an automobile, FIG. 2 is a side view of the window opening detection device, and FIG. 3 is a slit light and imaging for the rear window opening. An explanatory diagram showing the area, FIG. 4 is an enlarged view of the first corner part in the slit light irradiation state, FIG. 5 is a configuration diagram of the control system of the window opening detection device, and FIG. 6 is the window opening detection control routine. Schematic flowchart of Figures 7 to 11
The figures are explanatory diagrams showing slit lights and imaging areas according to modified examples, respectively. M...Vehicle body, 2...Rear window opening, 3...TV camera, 4...Floodlight, 6, 6A to 6E...Slit light, 14...Control unit for image processing.

Claims (2)

[Claims] (1) irradiating a plurality of parts of the detected part with slit light that covers the detected part and other parts, and capturing an image of the detected part irradiated with the slit light with an imaging means; By processing the image signal obtained by using an image processing means, boundary points between the detected part and other parts at multiple locations of the detected part are determined, and
A position detection method characterized by determining the position of a detected part using these boundary points. (2) A slit light irradiation means for irradiating a plurality of parts of the detected part with slit light that covers the detected part and other parts; and an imaging means for capturing an image of the detected part irradiated with the slit light. and image processing that receives an image signal from the imaging means, determines boundary points between the detected part and other parts at multiple locations of the detected part, and uses these boundary points to determine the position of the detected part. A position detection device comprising: means.