JPH07105355A - Method and device for recognizing picture - Google Patents

Abstract

PURPOSE:To accurately decide a coordinate axis so as to acquire an outer shape of the object accurately and to prevent positional deviation of components at a component mount device of a next stage by transmitting at least one of a light passing between an objective lens of a video camera and an object or a light passing between a lighting means and the object through an optical filter. CONSTITUTION:An optical filter 13 is arranged under a lighting means 6 comprising a circular fluorescent lamp and an objective lens of a video camera 7, rays emitted from the lighting means 6 reache an upper side of a board 1 and a position confirming mark 3 via an optical filter 13 and the rays reflected in each upper side reache the objective lens of the video camera 7 again via the optical filter 13. The color of the optical filter is selected so as to increase the contrast between the board 1 and the position confirmation mark 3 when the video camera 7 picks up a picture and the optical filter 13 having an exposure factor to a degree at which a dark part of the circumferential ridge of the position confirming mark 3 is not remarkable is properly selected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to an image recognition method and apparatus,
Particularly, the present invention relates to an image recognition method and apparatus suitable for a system for automatically mounting components on a printed circuit board.

[0002]

2. Description of the Related Art In an automatic assembly system, a component mounting device mounts a component at a predetermined position based on information from an image recognition device that recognizes the mounting position of the component based on an image captured by a video camera. Therefore, in a chip mounting device in which fine parts are mounted with high precision, for example, a chip component is mounted on the surface of a printed circuit board, high recognition accuracy of the image recognition device is required. In particular,
When the chip part shape is extremely small, such as 1 × 0.5 × 0.5 mm (1005 standard), a slight 0.2 mm misalignment often causes solder defects, commonly referred to as Manhattan defects. It has been known.

Therefore, when forming a printed wiring pattern on a printed circuit board, at the same time, a position confirmation mark is printed at a predetermined position on the substrate, and the chip mount device uses the position confirmation mark as an image by an image recognition device. It is common to recognize the component mounting position of the printed wiring pattern as a coordinate by capturing. FIG. 3 shows an example of the position confirmation mark printed on the printed circuit board. The position confirmation marks 3 and 4 are printed on the surface of the substrate 1 along with the printed wiring pattern 2 so as to be positioned on the diagonal line of the substrate 1. The image recognition device of the chip mount device recognizes the outer shape of each of the position confirmation marks 3 and 4 and obtains the center point thereof, and then determines the coordinate axis of the substrate 1 based on this.

FIG. 4 is a diagram showing the structure of a general chip mount device, which includes a carrier device 5 for holding and carrying the substrate 1, an illuminating means 6 for irradiating the surface of the substrate 1 with light, and a video camera 7. And an image recognition device 9 including a signal processing means 8 for processing an image signal from the video camera 7, and a video monitor 10 for monitoring an image of the image recognition device 9.
And a component mounting device 11 for mounting chip components and a control device 12 for controlling them. Incidentally, the illumination means 6
In order to irradiate the substrate 1 with a uniform light beam, the video camera 7
A circular fluorescent lamp arranged so as to surround the vicinity of the lens tip is used as the video camera 7 of a monochrome type.

The image recognition device 9 operates as follows.
First, after the substrate 1 is transported by the transport device 5 so that the position confirmation mark 3 is located immediately below the video camera 7, the video camera 7 captures the position confirmation mark 3 as a video image. The signal processing means 8 uses the base material of the substrate 1 and the position confirmation mark 3
Digital signal processing is performed to judge the outer shape of the position confirmation mark 3 and calculate the center point thereof by determining the density of the image due to the difference in the material. Therefore,
The image recognition device 9 obtains the center point of the position confirmation mark 4 in the same procedure and sends the data of both to the control device 12, and the control device 12 determines the coordinate axis of the substrate 1 from the position data of the two center points. Then, the component mounting device 11 in the next stage performs positioning when mounting the chip component. That is, the image recognition device 9 creates basic data for positioning when a fine chip component is mounted on the substrate 1. As the chip component is miniaturized as described above, its recognition accuracy is extremely important. Has become.

However, for example, when a substrate having a copper thin film pattern formed as a position confirmation mark on a white ceramic plate is recognized by the image recognition apparatus as described above, the outer shape of the position confirmation mark is accurately determined. There is a problem that the phenomenon that the coordinate axis cannot be grasped and the displacement of the coordinate axis becomes large frequently occurs, and in the worst case, a soldering defect occurs due to a mounting displacement of components.

[0007]

SUMMARY OF THE INVENTION The present invention has been made in order to eliminate the drawbacks of the conventional image recognition method and apparatus as described above, and accurately grasps the outer shape of a subject and uses the component mounting apparatus in the next stage. It is an object of the present invention to provide an image recognition method and device capable of accurately determining coordinate axes so that no positional deviation occurs.

[0008]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention illuminates an object with illumination means to photograph the object with a video camera and at the same time outputs a predetermined information from an output signal of the video camera with a signal processing means. In the image recognition method and apparatus for extracting the light, at least one of the light passing between the objective lens of the video camera and the subject or the light passing between the illumination means and the subject is transmitted through the optical filter. A characteristic image recognition method and apparatus.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Prior to the description of the present invention, a conventional image recognition apparatus will be considered briefly in order to help understanding of the present invention. For example, the base material of the substrate 1 is a white ceramic plate, the position confirmation mark 3 is a square with a side of 2 mm, and the thickness is 40 μm.
5A showing the image of the video monitor 10 and FIG. 5B showing the distribution of the luminance signal at AA ′ in the figure, the surface of the substrate 1 There is a band-shaped dark portion (a shaded portion in FIG. 6A) surrounding the periphery of the position confirmation mark 3, and the luminance signals of the position confirmation mark 3 and the substrate 1 are almost equal in level. There is. The width w of the band-shaped dark part is approximately 0.5 mm, and it becomes difficult for the image recognition device to accurately recognize at which position of the band-shaped dark part the peripheral edge of the position confirmation mark exists. Is thought to occur.

That is, in order to improve the recognition accuracy of the image recognition apparatus, it is important to eliminate the band-shaped dark portion and to make the difference in the luminance signal level between the position confirmation mark 3 and the substrate 1 clear (large). is expected. As a result of conducting research based on this prediction, it was found that the problem disappears when an optical filter is arranged between the objective lens of the video camera and the substrate which is the subject, or at least one of the illumination means and the subject. Of.

The present invention will be described below in detail with reference to the drawings showing an embodiment. FIG. 1 is an embodiment showing the configuration of an image recognition apparatus according to the present invention, in which an optical filter 13 is arranged below the illumination means 6 composed of a circular fluorescent lamp and the objective lens of the video camera 7 in the figure, The light rays emitted from the illumination means 6 reach the upper surfaces of the substrate 1 and the position confirmation mark 3 via the optical filter 13, and the light rays reflected by the respective surfaces are again passed through the optical filter 13 to the objective lens of the video camera 7. It is configured to reach.

FIG. 2 shows an optical filter 1 in the apparatus of FIG.
3 (a) and (b) are yellow filters (Y) for the camera.
When (2) is used, (c) and (d) respectively show the state of the monitor screen and the distribution of the luminance signal when the red filter (R1) for the camera is used. Yellow filter (Y
When 2) is used, the difference in the luminance signal level between the region of the position confirmation mark 3 and the region of the substrate base material becomes large, as is clear from FIG. The frequency of erroneously recognizing the outer shape of the confirmation mark 3 was reduced. Further, since the red filter (R1) has a large exposure multiple indicating the amount of transmitted light of the filter, which is 8, the monitor image becomes dark as a whole as shown in FIG. 6C, and FIG. Although the brightness signal level becomes low as a whole, the difference in the brightness signal level between the area of the position confirmation mark 3 and the area of the substrate becomes clear, and there is no drop in the brightness signal level at the periphery of the position confirmation mark. The frequency of erroneously recognizing the outer shape of No. 3 was greatly reduced, and it became possible to eliminate the problem almost completely. Incidentally, FIG.
In (c), the darker the image, the narrower the width of the diagonal line.

As described above, the color of the optical filter is set so that the contrast between the substrate and the position confirmation mark becomes high when the image is captured by the video camera, and the exposure multiple is such that the dark portion around the periphery of the position confirmation mark becomes inconspicuous. By appropriately selecting and using the optical filters having the above, it is possible to configure an image recognition apparatus that can accurately recognize the outer shape of the position confirmation mark.

Although the present invention has been described by taking the case where the same optical filter is mounted on both the video camera and the illumination means as an example, the present invention is not limited to this and only one of them is used. It may be provided with an optical filter, such as a polarizing filter as long as it eliminates the band-shaped dark portion around the position confirmation mark and makes clear the difference in the luminance signal level between the position confirmation mark and the substrate base material. Any optical filter may be used, and each of the video camera and the lighting means may be equipped with an optical filter having different properties such as color and exposure multiple, which are appropriately selected and provided.

Further, it is possible to configure the video camera so that it has a function similar to that of an optical filter by color-corresponding, analyzing the image signal and then performing electrical filtering. Also, by slightly defocusing the video camera, it is possible to mitigate the sudden drop in the luminance signal in the band-shaped dark area around the position confirmation mark.
It was confirmed that the same effect as when an optical filter was provided was provided. In order to accurately recognize the outer shape of the position confirmation mark, the focus is usually matched with the surface of the position confirmation mark. For the same reason, for example, in FIG. It is desirable to defocus to the upper side. Furthermore, the image recognition device according to the present invention is applied not only to a chip mount device, but also to other devices that require accurate position detection, such as a cream solder printing device that applies cream solder to a suitable position of a printed wiring pattern of a printed circuit board. Alternatively, the function of the signal processing means constituting the image recognition device may be provided in the control device for controlling the entire system.

[0016]

Since the present invention is configured as described above, it is possible to accurately grasp the outer shape of a subject and mount a component in the next stage by using an extremely simple configuration in which an optical filter is added. This is extremely effective in enabling accurate coordinate axis determination so that no positional deviation occurs.

[0017]

[Brief description of drawings]

FIG. 1 is a diagram showing an embodiment showing a configuration of an image recognition apparatus according to the present invention.

2 (a) and (b) are (c) and (c) when a yellow filter (Y2) for a camera is used in the apparatus of FIG.
FIG. 6D is a diagram showing a state of a monitor screen and a distribution of luminance signals when a red filter (R1) for a camera is used.

FIG. 3 is a diagram showing an example of a position confirmation mark printed on a substrate.

FIG. 4 is a diagram showing a configuration of a chip mount device.

5A and 5B are diagrams showing distributions of images and luminance signals on a video monitor.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Board 2 ... Printed wiring pattern 3, 4 ... Position confirmation mark 6 ... Illumination means 7 ... Video camera 8 ... Signal processing means 9 ... Image recognition device 12 ...・ Control device 13 ... Optical filter

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location H05K 13/08 U 8315-4E // G06T 7/00

Claims (4)

[Claims] 1. An image recognition method for illuminating a subject with light from an illuminating device, photographing the subject with a video camera, and extracting predetermined information from an output signal of the video camera with a signal processing device. An image recognition method characterized in that at least one of light passing between an objective lens of a camera and a subject or light passing between an illumination means and a subject is transmitted through an optical filter. 2. An image recognition apparatus comprising an illuminating means for irradiating a subject with light, a video camera for photographing the subject, and a signal processing means for extracting predetermined information from an output signal of the video camera. An image recognition device characterized in that at least one of an objective lens and an illumination means of a video camera is provided with an optical filter. 3. An image recognition apparatus comprising an illuminating means for irradiating a subject with light, a video camera for photographing the subject, and a signal processing means for extracting predetermined information from an output signal of the video camera. An image recognition apparatus characterized in that a video camera is made color-compatible and its output signal is electrically filtered by a signal processing means. 4. An image recognition apparatus comprising an illuminating means for irradiating a subject with light, a video camera for photographing the subject, and a signal processing means for extracting predetermined information from an output signal of the video camera. An image recognition apparatus, characterized in that the video camera is slightly out of focus from the subject.