JPH04123497A - Detection of position of substrate and means of loading part on substrate - Google Patents

Abstract

PURPOSE:To accurately detect a mark position without being affected by the condition of the surface of an alignment mark prepared on a substrate by detecting the minimum one of configurations similar to a predetermined configuration at least partly in contact with the positioning mark obtained from a binary image signal and by using the position of the detected minimum configuration to detect the position of the positioning mark. CONSTITUTION:A pattern is detected (reference number 40), encoded into a binary value (reference number 41), and converted into a two dimensional binary digital image, followed by detecting the outermost border line of the pattern (reference number 42), further followed by obtaining summits of a totally convex polygon from a group of points composing the outermost border line (reference number 43), furthermore followed by obtaining the farmost point Voronoi-diagram for the summits of the totally convex polygon (reference number 44), again followed by obtaining a circle with the minimum radius from circles having the center at the Voronoi point and passing the three points corresponding to three Voronoi areas adjacent to the Voronoi points, that is, the minimum inclusive circle (reference number 45), and finally followed by letting the center of this minimum inclusive circle be the position of positioning mark. The minimum inclusive circle can be determined at a constant position despite a defect of the profile section, thereby being able to detect the position of the substrate accurately. With this, the position of a substrate can be accurately detected. Since the inclusive configurations have only to be circum-scribed about the positioning mark, they have only to be made similar to one another.

Description

[Detailed Description of the Invention] [Industrial Field of Application] The present invention provides eight methods for detecting the position of a board and its use, such as when automatically mounting electronic components on a board or inspecting the mounted components. The present invention relates to a device for mounting components onto a printed circuit board.

[Conventional technology]

As a board position detection method for automatic electronic component mounting machines, Tsuji et al.
i, review, vo 1.68. No, 10. p
p.

21-24 (1986-10),
A known method is to detect a brightness waveform along a measurement line of a pattern and detect its changing points. The same paper also describes that a method of determining the center of gravity of a positioning mark is also commonly used.

Other board position detection methods include, for example, Sugawara "Screen printing machine with automatic positioning function for large industrial boards" sensor technology, v o l, 8 r No 6 + p
p.

132-135 (1988-5), two or more circular alignment marks provided on the substrate are detected with a camera, and the center of gravity of the binary image is determined to determine the position of the substrate. method is known.

Furthermore, a method for detecting the center position of a cross-shaped alignment mark from its projection is described in Japanese Patent Application Laid-Open No. 61-165184, Vol. 96-6, Journal of the Institute of Electrical Engineers of Japan.

No. 1 (1976) "Group Control Transistor Assembly System Using Time-Division Pattern Recognition Technology"
Detecting the position by matching a stored specific pattern to template 1, JP-A-57-1548
It is also known that the position is detected from the one-dimensional detection waveform of the alignment mark described in Japanese Patent No. 87.

[Problem to be solved by the invention]

The above-described conventional position detection method is based on the premise that the alignment mark or specific pattern on the substrate is correctly image-detected in a predetermined shape. For this reason, if the mark cannot be detected in the correct shape, such as when the board is coated with solder, the error in detecting the position of the board increases, making it impossible to correctly mount and inspect components.

SUMMARY OF THE INVENTION In order to solve the above problems, an object of the present invention is to provide a substrate position detection method that accurately detects the mark position without being affected by the surface condition of an alignment mark provided on the substrate.

Another object of the present invention is to provide a device for mounting components on a board, which is capable of accurately detecting mark positions and mounting components on a board without being affected by the surface condition of alignment marks provided on the board. Our goal is to provide the following.

[Means to solve the problem]

In order to achieve the above object, the present invention arranges marks of a specific shape at multiple locations on a substrate, detects the shape of the mark from a specific position, converts it into an electrical signal, and images the converted electrical signal. is binarized so that the mark pattern is Pa1" and the others are 11011, and the minimum size figure that includes the binarized pattern (called the minimum inclusive figure) is detected, and based on its position. This is to determine the position of the substrate.

That is, the present invention arranges patterns of specific shapes as positioning marks at a plurality of locations on a substrate, optically detects the positioning marks to obtain an optical image, converts the optical image into an electrical signal, and converts the optical signal into an electrical signal. a predetermined shape that circumscribes the shape of at least a portion of the positioning mark obtained by the binary image signal; This method of detecting the position of a substrate is characterized by detecting the smallest figure among figures similar to the figure, and detecting the position of the positioning mark from the position of the detected smallest figure.

The present invention also provides a first means for detecting the shape of the positioning mark on the substrate, and a second means for detecting the position of the substrate based on information about the outer peripheral contour of the detected shape detected by the first means.
This is a device for mounting components on a board, characterized by comprising means for positioning and mounting the component on the board based on the position of the board detected by the second means.

[Effect]

When a mark used for position detection provided on a board is detected using, for example, a ring-shaped illumination light source, if the mark surface is coated with solder, the three-dimensional shape will cause a difference between ), a mark may be detected as a pattern with part of it missing.

However, since it is rare for more than half of the original outline of the mark to be missing, the smallest size figure (minimum hanger figure) that includes the detected pattern is However, the position of the substrate can be determined at a constant position, thereby making it possible to accurately detect the position of the substrate. Since the above-mentioned inclusive figures need only circumscribe one positioning mark, they only need to be similar to each other, and do not necessarily have to have the same shape as the positioning mark as shown in FIG.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 and 4 to 8.
This will be explained using figures. In this embodiment, a circular pattern is used as the mark used to detect the position of the substrate.

As shown in FIG.
After converting into a value digital image, the outermost contour of the pattern is detected (indicated by 42), then the white vertices are found from the school of fish forming the outermost contour (indicated by 43), and then the white vertices are determined. Find the farthest point Voronoi diagram of the points (indicated by 44),
Find the circle with the minimum radius, that is, the minimum enclosing circle (indicated by 45) among the circles centered on the Voronoi point and passing through the three points adjacent to the Voronoi point and corresponding to the three Voronoi regions (indicated by 45), and use this center as the positioning mark. position.

Next, the outermost contour line, white color, and farthest point Voronoi diagram used in this embodiment will be described.

To detect the outermost contour line, as shown in Fig. 5, the detected binary image is first searched from left to right and from top to bottom, and the value at each Y coordinate is changed from "it O" to "1".
11 List the coordinates on the image of the points that changed to P+, then search from right to left, bottom to top, and similarly from "0" to "
A series of coordinates is generated by enumerating the coordinates of points that have changed to 1''.

As is clear from the detection method, the outermost contour line is detected by detection 2.
This is the candidate point of the outermost contour of the value pattern, and the subsequent white detection result is u 1 +t of the detected binary pattern.
It matches when 6 capsules are directly detected using the coordinates of all points in the part. Since detection of the outermost contour line is simple image processing, this can reduce the amount of input data for white detection, which requires more complex processing, and, as a result, reduce the overall processing time. .

As shown in FIG. 6, white is a convex polygon that includes all points on its vertices, sides, or inside. The purpose of white detection is to extract only white vertices from the outermost contour, without losing information on the three points on the contour that determine the minimum enclosing circle, and to detect the farthest point, which requires more complex processing. The goal is to reduce the amount of input data for the Voronoi diagram generation process, and as a result, reduce the overall processing time. For example, the white detection method is described in "Computational Geometry and Geographic Information Processing J.
Bit Bessatsu-PP79 can be realized by using Graham's algorithm shown in Kyoritsu Shuppan (1986). In this case, the outermost contour line has already been detected in the order of left to right, top to bottom, right to left, and bottom to top as described above, so it is used as preprocessing in rGraham's algorithm. There is no need for sorting of declinations.

The farthest point Voronoi diagram is used, for example, when the minimum enclosing circle is efficiently and precisely determined, as described in the above-mentioned ``Computational Geometry and Geographical Information Processing JppHO''. The farthest point Voronoi diagram for points a-g in Figure 6 is shown. ) and their boundaries, the ``Voronoi edges'', and the intersections of the Voronoi edges, ``
It consists of "Voronoi points". All points in each Voronoi region, for example A, have the property that the distance to point a is greater than the distance to other points b-g. Therefore, as shown in FIG. 8, the Voronoi points that are the boundary points of three (or more than four) Voronoi regions are the points (C, D, E in FIG. 8) corresponding to the respective Voronoi regions (C, D, E in FIG. In the same figure, C2
The distances to points d and e) are equal, and the distances to any other point are smaller than the distances to points c, d, and e. As a result, the Voronoi point becomes the center of an inclusive circle (a circle that includes all points on or inside the circumference) (in the case of the figure, passing through points c, d, and e). In this way, the farthest point Voronoi diagram is generated,
If we find the Holonoi point whose radius is the minimum by determining the inclusive radius for all Poronoi points, we can determine the center coordinates of the minimum inclusive point. How to generate the farthest point Voronoi diagram.

[Computational Geometry and Geographic Information Processing Jpp136-14]
As described in 1, this can be efficiently realized using the "sequential addition method", "recursive bisection method", etc.

- Once the positions of the individual positioning marks on the substrate are determined, the position of the substrate can be determined using a technique similar to the conventional substrate positioning method.

According to this embodiment, since a circular pattern is used as the positioning mark, the position of the minimum included figure can be determined with the simplest processing, and the position of the substrate can be determined in a short time.

In this embodiment, pattern detection can be performed using any known pattern detection method, such as a TV camera or CCD line sensor scanning as a detector, or ring-shaped illumination as an illumination method, as long as the positioning mark can be clearly detected. It's okay. Furthermore, if an increase in processing time is allowed, the detection of the outermost contour line or the detection of the outermost contour line and white color may be omitted. and the ratio of the diameters in each direction is P:
An ellipse such as q may be used as a positioning mark. In this case, after detecting white, the XY coordinates of the vertices are multiplied by 1 and -, respectively, the minimum inclusiveness is obtained in the same manner as P q , and the center coordinates thereof are multiplied by 2 and q. This method can also be applied when the sampling intervals of an image do not match in the X and Y directions during pattern detection, and as a result, a circular pattern becomes an ellipse on the image. Since the coordinate transformation described above is performed on a small number of points after white detection, the processing time can be significantly shortened compared to the case where all detected patterns are transformed into circles and processed. has exactly the same effect.

The processing device for carrying out this embodiment is any type of processing device that can binarize an image signal, sample it at regular time intervals, store it in a memory, and freely process the contents of the memory using a program. Any known processing device may be used.

Next, using FIG. 9 and FIG.
An example will be described below.

In this embodiment, as in the first embodiment, a circular pattern is used as a positioning mark, and in addition to the processing up to the minimum inclusive detection 23 described in the first example.

Extraction of circumference candidate points 51 and calculation of the most/IX squared approximation circle 5
Do step 2.

As shown in FIG. 10, the circumferential candidate points are extracted from among the white vertices 47 whose distance from the circumference of the minimum inclusive 46 is within a certain value α (for example, a scene corresponding to one pixel). This is done by extracting points. Therefore, 48 circumferential candidate points are removed.

The coordinates of the extracted point are (xi, yi)i=1+2,...
・, N, and the equation of the circle is x2+y"+2gx+2fy+c=Q-old-(1), the least squares approximation circle is ε 2 : Σ(x+"+y, "+2 gx, +2fy
++c)'' ----(2) If l is set and the three-dimensional linear simultaneous equations are solved for gyfyc, the center (-g+-f)l radius J?117° below- can be obtained.

In this embodiment, the center of the least squares approximation circle obtained in this manner is set as the position of the positioning mark, and the position of the substrate is determined.

According to this embodiment, since the mark position is finally detected as the center of the least squares approximation circle, it is possible to minimize the image sampling error and to perform highly accurate position detection in pixel units or less. effective.

In this embodiment, as in the first embodiment, any known pattern detection and processing device may be used. Further, oval positioning marks can also be handled in the same way. In the case of this embodiment, the xy coordinates of the white vertices are set as the factor and the center coordinates for the minimum 2 q power approximation are set as 2 times and q times.

In the embodiments described above, circular patterns and oval patterns were used as positioning marks, but as shown in FIG. Furthermore, if the pattern includes a known shape (the size may be unknown) on the outermost contour, such as a pattern that exists in a blank space or a pattern that consists of an arbitrary known curve, positioning can be performed according to the principles of the present invention. It is clear that the position of the mark can be detected accurately.

Next, an apparatus for mounting components on a board according to the present invention will be described with reference to FIG.

Above mentioned 1. According to the second embodiment, the position of the positioning mark on the board can be detected with high accuracy without error, regardless of its surface condition such as solder coating. First, the board 11
0 is transported to the component mounting device by the board transport mechanism, and roughly positioned mechanically by guide pins 111. The board 110 has two circular positioning marks, 112a,
112b is formed. Next, the overall control device 10
According to commands 1 to 9, ring lights 107a, 10
positioning mark 112a of the board illuminated by 7b,
112b is detected by the TV cameras 106a and 106b, and the image is captured into the image processing device 100.

The image processing bag [100 indicates the position detection power C or l'~' camera 106 shown in the first embodiment or the second embodiment.
This is executed for each of the images detected at 106a and 106b, and the positions of the marks 112a and 112b are output to the overall control device]01.

The image processing device 100 is capable of sampling an input image at a constant clock and importing it into memory as a two-dimensional digital image, and processing the imported image using dedicated hardware or software such as a microprocessor. Any known type of equipment can be used as long as it can be processed freely by the following methods. Next, the overall control device issues a command to the xyz orthogonal robot 103 and supplies the component 1 to be loaded from the component supply mechanism
09 using the component suction mechanism 105 and positioning marks 112a and 112b from the image processing device 100.
Mount the component at the corrected position on the board.

Note that correction of one position is not limited to correction of the position of the XYZ orthogonal robot 103;

You may also move it to Y. Thereafter, after all the parts have been taken out and mounted on the designated parts, the board 110 is transported again using the board transport mechanism 108.

Note that the parts may be taken out/loaded using any device that can correctly mechanically take out/load parts, such as an XYZ orthogonal robot and a parts suction mechanism, or a so-called scalar robot.

In addition to the combination of a ring illumination and a TV camera, the image may be detected by any known method as long as the positioning mark can be clearly detected as a two-dimensional image.

According to this embodiment, it is possible to accurately mount components regardless of the surface condition of the board positioning mark.

In addition, in the above embodiment, mounting of parts was taken as an example, but 1
It goes without saying that the method of the present invention can be applied to any device that needs to detect the position of a board, such as inspection of a board on which components are mounted, and similar effects can be obtained.

(Effects of the Invention) According to the present invention, the position of the positioning mark can be accurately detected regardless of the surface condition of the mark, so it is possible to mount components on the board or inspect the board with high precision and without errors. There is an effect.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the flow of the automatic position detection method according to the present invention, FIG. 2 is a diagram showing an example of detecting positioning marks on a board, FIG. 3 is a diagram explaining the minimum included figure, and FIG. 4 is a diagram according to the present invention. FIG. 3 is a diagram showing the flow of the first embodiment. Figure 5 is a diagram showing a method for detecting the outermost contour line, Figure 6 is a diagram explaining 6 capsules, Figure 7 is a diagram explaining the farthest point Voronoi diagram, and Figure 8 is a diagram explaining the furthest point Voronoi diagram. Diagram explaining properties, No. 9
The figure is a diagram showing the flow of the second embodiment according to the present invention.
The figure shows a method for extracting candidate points on the circumference according to the second embodiment.
FIG. 11 is a diagram showing another example of a positioning mark on a board that can be used in the position detection method according to the present invention, and FIG. 12 is a diagram showing the configuration of an embodiment of a component mounting apparatus on a board according to the present invention. DESCRIPTION OF SYMBOLS 100... Image processing device, 101... Overall control device, 103...
T'V camera, 107a. b...Ring illumination, 108...Substrate transport mechanism, 109
...Mounted parts, 110-board, ILI...guide bin, 112a, b positioning marks. No. 1 Suji Snake (cL) <b> (C) (d-) 躬蔀S Hidai Daidai Shirohama, Cd, eUJh3'2+Yendaikuchidaiku (α) (b) <C> ( cL) Second Flash

Claims (4)

[Claims] 1. A pattern with a specific shape is placed as a positioning mark at multiple locations on the substrate, the positioning mark is optically detected to obtain an optical image, the optical image is converted into an electrical signal, and the electrical signal is used as the positioning mark of the positioning mark. At least one part and the other part are binarized and converted into a binarized image signal, and a figure similar to a predetermined shape circumscribing the shape of at least a part of the positioning mark obtained by the binarized image signal is created. A method for detecting the position of a substrate, comprising: detecting the smallest figure among them, and detecting the position of the positioning mark from the position of the detected smallest figure. 2. 2. The method for detecting the position of a substrate according to claim 1, wherein the outermost circumferential outline and the smallest circumscribed figure of the positioning mark are circles. 3. Extract the points constituting the shape of the binarized positioning mark within a certain distance from the detected minimum circumscribed circle, apply the least squares approximation to the extracted points to fit a circle, and position using the center position of the circle. 3. The method for detecting a position of a substrate according to claim 2, wherein the position is a mark position. 4. a first means for detecting the shape of the positioning mark on the substrate; a second means for detecting the position of the substrate based on information on the outer peripheral contour of the detected shape detected by the first means; and the second means. 1. A device for mounting components on a board, comprising means for aligning and mounting components on a board based on the position of the board detected by the method.