JP2008014684A - Method of acquiring camera scaling in component placement machine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method of acquiring a scaling of a component recognition camera enabling to improve a component recognition precision in a component placement machine. <P>SOLUTION: Image of a tool formed with a plurality of patterns is picked up with a camera, a camera view 30 is divided into a plurality of view areas 30a-30i so as to pick up at least one of the imaged patterns within a view area. With processing a pattern image of each view area, a scaling value is calculated for each view area and stored. In this configuration, since the scaling value can be calculated with reference to each scaling value within the view area, the actual coordinate of the part can be measured at a high precision and the part recognition precision can be improved. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method for acquiring a scaling value of a component recognition camera in a component mounter that mounts electronic components on a printed circuit board.

  2. Description of the Related Art Conventionally, there has been known a component mounter that mounts an electronic component on a circuit board by sucking an electronic component supplied from a component supply device with a suction nozzle of the mounting head, moving the mounting head to the circuit board. If suction displacement occurs when parts are picked up, accurate mounting is not performed, so picked up electronic parts are picked up by a part recognition camera, and the mounting position is corrected based on the part center and picking angle required by the image processing. And component mounting.

  Component recognition for obtaining the component center and suction angle is performed by obtaining the scaling value of the component recognition camera (actual size per pixel: also called pixel rate), so accurate scaling values are obtained for accurate component mounting. There is a need to.

As a method of acquiring the camera scaling value in this mounting apparatus, the number of pixels between the cores of the two circles is obtained by imaging a jig whose distance between the cores of the two circles is grasped in advance by the camera. Or obtaining a scaling value from the recognition coordinates by moving the jig in four directions, up and down, left and right, as disclosed in Patent Document 1. ing.
Japanese Patent No. 352001

  However, in the conventional method for acquiring the scaling value in the electronic component mounting apparatus, the influence of tilt at the time of camera assembly, tilting of the camera CCD surface, and the like has not been taken into consideration. For this reason, since an error has occurred in the scaling value in the field of view, there is a problem that it is not possible to meet the demands for miniaturization, high functionality, and high density mounting.

  Accordingly, the present invention has been made to solve such problems, and it is an object of the present invention to provide a camera scaling acquisition method capable of improving component recognition accuracy in a component mounter.

The present invention for solving the above problems
A camera scaling acquisition method for a component mounter,
Take a picture of a jig with multiple patterns with a camera,
Dividing the camera field of view into a plurality of field areas such that at least one of the captured patterns is imaged within the field area;
A pattern image in each visual field area is processed, a scaling value is calculated for each visual field area, and stored.

  According to the present invention, since individual scaling values within the divided camera field of view can be referred to, the actual coordinates of the component electrodes can be measured with high accuracy, and the recognition accuracy of the components can be improved.

  Hereinafter, the present invention will be described in detail based on embodiments shown in the drawings.

  FIG. 1 is a schematic front view of a component mounter 1 in which the present invention is used, and FIG. 2 is a block diagram of its control system. In FIG. 1, the component mounting apparatus 1 is provided with a mounting head unit 2 equipped with a suction nozzle 3 that sucks an electronic component (not shown) supplied from a component supply unit. It is moved in the X-axis and Y-axis directions by the XY drive unit 4. A substrate recognition camera 5 constituted by a CCD camera or the like is fixed to the mounting head unit 2, and this substrate recognition camera 5 is a substrate (not shown) on which electronic components are mounted or, as described later, The jig 6 for camera scaling acquisition installed on the base frame of the component mounting machine 1 is imaged. In addition, a component recognition camera 8 constituted by an electronic component sucked by the suction nozzle 3 or a CCD camera for imaging the jig 6 is fixed to the base of the component mounting machine 1.

  As shown in FIG. 2, the XY drive unit 4 includes an X-axis motor 11 and a Y-axis motor 12 that are driven based on control of a controller 10 including a CPU, a ROM, a RAM, and the like. The mounted head unit 2 is driven by the X-axis motor 11 and the Y-axis motor 12 and moves in the XY directions. Further, the suction nozzle 3 attached to the mounting head unit 2 is moved up and down by a Z-axis motor 13 driven based on the control of the controller 10 and is rotated around the suction shaft by a θ-axis motor 14. .

  The image processing device 17 includes a CPU 17a, a memory 17b, an A / D converter 17c, and the like. The image processing device 17 processes a board image captured by the board recognition camera 5, detects the position of the board, and also recognizes a component recognition camera. The image of the electronic component imaged in step 8 is processed to recognize the component, and the center position and suction inclination of the component are acquired.

  The image processing device 17 processes the image of the jig 6 captured by the substrate recognition camera to detect the position of the jig 6 and also processes the image of the jig 6 sucked by the suction nozzle 3. The scaling value of the component recognition camera 8 is acquired.

  In addition, the component mounter 1 is provided with input devices such as a keyboard 15 and a mouse 16 for inputting component data, and these input data are stored in a storage device 18 including a hard disk, a flash memory, and the like. Stored. In addition, the component mounter 1 is provided with a monitor 19, and on the screen of the monitor 19, the board and jig imaged by the board recognition camera 5 and the electronic component and jig imaged by the component recognition camera 8 are displayed. It is possible to display the images.

  In such a component mounting machine 1, the board image picked up by the board recognition camera 5 is processed to perform board recognition. The electronic component supplied from the component supply unit is sucked by the suction nozzle 3 and then moved onto the component recognition camera 8 and picked up by the component recognition camera 8. The image is processed by the image processing device 17. Part recognition is performed. The controller 10 corrects the component mounting position stored in the storage device 18 based on the substrate recognition and the component recognition result, and mounts the electronic component at the correct position on the substrate.

  In component mounting using such a component mounting machine, the component recognition camera 8 is not mounted horizontally on the XY plane, and an error occurs in the scaling value (actual size per pixel) within the field of view due to the tilt angle. May occur. In such a case, since high-accuracy component recognition is not guaranteed, in the present invention, the scaling value of the component recognition camera is acquired using the jig 6 to perform component recognition.

  Below, the method to acquire the scaling value of a component recognition camera is demonstrated.

  FIG. 3 shows the jig 6 for obtaining the scaling value in detail. The jig 6 is formed of, for example, a rectangular transparent thin glass substrate, on which two positioning marks 6a and 6b, a number of horizontal lines 6c, and vertical lines 6d are gridd at equal intervals. A plurality of square patterns 6e are formed, which are drawn as a vertical pattern, and are formed by vertical and horizontal lines. The positions of the marks 6a and 6b, the line positions of the lines 6c and 6d, the line widths, and the like are accurately measured in advance. The shapes of the marks 6a and 6b may be circular or cross-shaped, and the marks 6a and 6b and the lines 6c and 6d are used in order to recognize the mark and the grid pattern satisfactorily. Is drawn in silver.

  Acquisition of the scaling value is performed according to the flow of FIG. First, the scaling acquisition jig 6 is placed at a predetermined position (step S1), and the substrate recognition camera 5 attached to the mounting head unit 2 is driven by driving the X-axis motor 11 and the Y-axis motor 12 to mark 6a on the jig 6. Move upward (step S2).

  Next, the mark 6a is imaged by the substrate recognition camera 5, and the image of the mark 6a is processed by the image processing device 17 to calculate the center of gravity. Subsequently, the mark 6b is imaged in the same manner to determine its center of gravity (steps S4 and S5), and the center position and inclination of the jig 6 are calculated from the center of gravity of each mark 6a and 6b (step S6).

  Subsequently, the suction nozzle 3 is mounted on the mounting head unit 2, the suction nozzle 3 is moved to the calculated center position of the jig 6, and the Z-axis motor 13 is driven to lower the jig 6 (step S7). ). Then, the θ-axis motor 14 is driven to correct the inclination of the jig according to the calculated angle, and the jig 6 is moved onto the component recognition camera 8 (step S8).

  Subsequently, an illumination light source (not shown) is turned on to illuminate the jig 6, and the jig 6 is imaged by the component recognition camera 8 (step S9). An image of the jig 6 is shown in FIG. 4, and an image 6 ′ of the jig 6 is shown in the field of view 30 of the component recognition camera 8. 6a 'and 6b' are images of the marks 6a and 6b, 6c 'and 6d' are images of the lines 6c and 6d, and the camera field of view 30 is divided into nine 3x3 field areas 30a to 30i. The visual field 30 is divided into a plurality of visual field areas such that at least one pattern 6e is captured as a pattern image 6e 'in each divided visual field area. Therefore, a division pattern other than 3 × 3 can be used. What is important is that the camera field of view is divided into a predetermined number of field areas so that accurate scaling values can be obtained in the individual field areas divided.

  As shown in FIG. 6, for example, in the visual field area 30e obtained by the division, a pattern composed of an image 6c ′ of two horizontal lines 6c and an image 6d ′ of two vertical lines 6d is a pattern image 6e. It has been imaged as'. The X direction edges E1 and E2 of this pattern image are detected using a known method, and the average value of each detected edge is obtained to determine the number of pixels between the edges E1 and E2 from the X coordinates X1 and X2 of the edges E1 and E2. Ask for. Since the actual distance between the edges E1 and E2 is known in advance, the pixel rate, ie, (distance between edges) / (number of pixels between edges) corresponding to the scaling value is calculated. Similarly, also in the Y direction, the edges E3 and E4 of the pattern image 6e 'are detected, the Y coordinates Y1 and Y2 of the edges E3 and E4 are obtained, and the pixel rate in the Y direction is obtained.

  The above-described processing is performed for each of the visual field areas 30a to 30d and 30f to 30i, and the scaling values in the X direction and Y direction obtained for each of the visual field areas 30a to 30i are stored in the memory 17b or the storage device 18.

  The above process is the process of step S10.

  Actually, when an electronic component is mounted on a substrate, component recognition is performed using the stored scaling value for each field of view. The mounting head unit 2 is moved to the component supply unit by the XY drive unit 4. Then, the electronic component supplied from the component supply unit is sucked by the suction nozzle 3 and then moved onto the component recognition camera 8. An electronic component is imaged by the component recognition camera 8 on the component recognition camera 8.

  For example, FIG. 8 shows a state in which the electronic component 40 including the electrodes 40a, 40b, and 40c is imaged in the field of view 30 by the component recognition camera 8. In order to obtain the actual coordinates of the electrodes, the pixel rate of the pixels is required.

  FIG. 7 illustrates a method for obtaining the pixel rate of an arbitrary pixel P. In FIG. 7, it is assumed that the pixel P exists in the visual field area 30a, for example. When determining the pixel rate of the pixel P, the pixel rates of the visual field area 30a and the three visual field areas 30b, 30d, and 30e in the vicinity thereof are referred to. The pixels A, B, C, and D are center pixels of the visual field areas 30a, 30b, 30d, and 30e, and the pixel rate (scaling value) of each stored visual field area is used as the pixel rate.

  Lines AB, AC, and AD connecting the central pixels are drawn, and perpendiculars are drawn from the pixel P to the lines AB, AC, and AD to obtain intersections B ′, intersections C ′, and intersections D ′. The pixel rate is calculated for each intersection from the distance between the two intersections and the pixel A and the ratio of the distances between the pixel A and the pixels B ′, C ′, and D ′.

For example, the pixel rate of the intersection B ′ is
(Pixel rate of pixel A) + ((pixel rate of pixel B) − (pixel rate of pixel A)) × (length between line segments AB ′ / length between line segments AB)
It calculates according to the formula.

  Similarly, the pixel rate is calculated based on the pixels A and C for the intersection C ′, the pixel rate is calculated based on the pixels A and D for the intersection D ′, and the intersections B ′, C ′, and D are calculated. Let the average value of the three pixel rates obtained from 'be the pixel rate of the pixel P. The pixel rate of the pixel P is obtained for each of the X and Y directions.

  In this way, since the pixel rate of an arbitrary pixel can be calculated, as shown in FIG. 8, for example, the pixel rate of the pixel P in the vicinity of the electrode 40a is calculated by the method shown in FIG. The pixel Q is a pixel at the center of the field of view of the camera, and the actual coordinates are calculated with this pixel as a reference. As the pixel rate of the pixel Q, the pixel rate of the visual field area 30e stored in the memory 17 is used. Then, an average value of the pixel rates of the pixel P and the pixel Q is obtained, and the X direction real coordinates of the electrode 40a are calculated by multiplying the average value by the number of pixels in the X direction between the pixels PQ. The actual Y-direction coordinates of the electrode 40a are calculated by multiplying the number of Y-direction pixels between PQs. Here, the real coordinates are coordinates that can be handled by a machine, such as a unit of mm.

  Similarly, for the other electrodes 40b and 40c, the actual coordinates in the X and Y directions are obtained, and the component center and inclination are obtained from the obtained electrode coordinates.

  The pattern formed on the jig 6 is not a grid-like pattern as shown in FIG. 3, but can be any pattern as long as it is a plurality of patterns whose sizes and distances between patterns are known. . For example, a jig 60 on which a plurality of patterns 60b including a plurality of marks 60a as shown in FIG. 9 is drawn can be used. In this case, the camera field of view is divided so that an image is captured within the field of view where at least one pattern 60b is divided. The center of gravity of each mark 60a is acquired, and the scaling value in each visual field area is acquired from the number of pixels between the centers of gravity of the marks.

It is a front view which shows schematic structure of a component mounting machine. It is a block diagram which shows the structure of the control system of a component mounting machine. It is a top view which shows the structure of the jig for scaling acquisition. It is explanatory drawing which showed the state which divides | segments a camera visual field. It is the flowchart which showed the method of acquiring the scaling value of a camera. It is explanatory drawing for demonstrating acquisition of the scaling value of a visual field area. It is explanatory drawing for demonstrating acquisition of the scaling value of arbitrary pixels. It is explanatory drawing explaining the state which calculates the real coordinate of the electrode of components. It is the partially broken top view which shows the other structure of the jig for scaling acquisition.

Explanation of symbols

2 mounted head unit 3 suction nozzle 4 XY drive unit 5 substrate recognition camera 6 jig 8 component recognition camera 17 image processing device 30 camera field of view

Claims (2)

A camera scaling acquisition method for a component mounter,
Take a picture of a jig with multiple patterns with a camera,
Dividing the camera field of view into a plurality of field areas such that at least one of the captured patterns is imaged within the field area;
A method for obtaining camera scaling of a component mounting machine, wherein a pattern image in each visual field area is processed to calculate and store a scaling value for each visual field area.   2. The component mounting according to claim 1, wherein the scaling value of the pixel is calculated by using the scaling value in the visual field area where the pixel exists and the scaling value in the visual field area located in the vicinity of the visual field area. Camera scaling acquisition method.

Images