JPH10117100A - Method for measuring position of component - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method, capable of measuring the position of a component, using a template matching technique. SOLUTION: A pictorial image of a component 20 in a shape point- symmetrical about its own center, is picked up, and an area 22 containing a part or the whole of the pictorial image of this component 20 is detected. A template 24 which is symmetric with respect to the area 22 about X-axis or Y-axis passing the center of the component 20 is formed, and its matching with this template 24 is performed. This matching is performed by rotating the pictorial image of the component 20 around the center. The inclination of the component 20 is detected by finding the amount of rotation θ, when matching is obtained. The area 22 is set by forming a template containing a part or the whole of the component 20, and by a template area capable of obtaining matching between the pictorial image and this template. Consequently, it is possible to increase the matching accuracy and to increase the accuracy of inclination detection.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a position of a component, and more particularly, to a position shift at the time of suction when a component such as an electronic component mounted on a chip mounter is sucked and conveyed to a mounting board. The present invention relates to a method for detecting the position of a component for detecting the position.

[0002]

2. Description of the Related Art Conventionally, a chip mounter is provided with a suction head having a suction nozzle for sucking an object, for example, an electronic component such as an IC chip component (hereinafter simply referred to as a component), and is supplied from a feeder. The components are sucked by the suction nozzle, transferred and mounted on the circuit board. Normally, a component is not necessarily picked up in a correct posture. Therefore, the picking up posture of the component is imaged by an image pickup device such as a CCD camera, and an image of the component is recognized. The suction nozzle of the component obtained based on the image recognition is obtained. After the deviation from the center position and the amount of inclination are corrected, the component is mounted on the circuit board.

In order to obtain or correct the positional deviation amount, a position detection algorithm according to the type of component has been developed. However, there is a problem that component data necessary for each algorithm must be described for each component. There is. As a general-purpose algorithm for solving this complexity, for example, template matching as described in Japanese Patent Publication No. Hei 8-12050 is known. The template matching is a method of creating a template representing a component, moving the template in an image including the component, and recognizing that the component is located at a template position where a large similarity is obtained between the image and the template.

In the above publication, a first template including a part or all of a part (reference mark) and a second template which is point-symmetrical to the first template are registered, and an image of the part to be inspected and the first and second parts are registered. Matching with the template is performed, and the center shift is determined with the midpoint of the two corresponding points of each template at the time of the matching as the center of the component.

The technique described in this publication can detect the center position of a component, but cannot detect the inclination of the component. In the above publication, instead of using a registered template, a window including a part or all of a part is set in advance, a point-symmetric template is created with this window, and matching is performed with the template to obtain the center position. Is also shown. In the case of this example, there is an advantage that the memory for storing the template can be omitted. However, a stable result can be obtained if the part is located at a substantially constant position in the image, but otherwise, the part does not enter the window and does not contain a valid feature, and the expected result is obtained. There is a problem that it cannot be obtained.

To cope with this, it is possible to cope to some extent by enlarging the window, but the template contains a lot of extra information, which not only lowers the matching accuracy but also increases the operation time. When the template is registered and executed in advance, the above does not occur. However, when the component has an inclination and there is a rotational deviation between the template and the template, the matching accuracy is reduced, and the center accuracy is also expected. You can't get a position.

[0007]

Since the template matching method can be implemented as a general-purpose position detection algorithm for irregularly shaped parts that fall outside the category of the conventional position deviation detection algorithm, the present invention provides a template matching method. It is an object of the present invention to provide a component position detection method capable of detecting a component position by a general-purpose, high-precision position detection algorithm using a method of the method.

[0008]

A part position detecting method according to the present invention captures an image of a part having a point-symmetrical shape with respect to its own center, detects an area including a part or all of the image of the part, A template symmetrical to the region with respect to the X-axis or the Y-axis passing through the center of the component is created, and by matching the template with the image of the component, the inclination of the component is detected.

Further, in the component position detecting method according to the present invention, an image of a component having a point-symmetrical shape with respect to its own center is taken, an area including a part or the whole of the image of the component is detected, and the area and the point are detected. Create a template that is symmetrical and replaced with the reference window itself or actual image data, and set the center point between the corresponding area and the corresponding two points of the template when the matching between the image and the template is obtained. And a template symmetrical to the region with respect to the X-axis or the Y-axis passing through the center of the component is created, and by matching the template with the image of the component, the inclination of the component is detected.

In such a configuration, the center and the inclination of the component can be accurately detected by template matching.

Preferably, the matching with the X-axis or Y-axis symmetric template is performed by rotating the image of the component around its own center, and the inclination of the component is determined based on the amount of rotation at the time of matching. Is detected.

The above-mentioned area is set as a template area in which a template including a part or the whole of a part is created and an image is matched with the template.
In this way, even if the position of the acquired component image is unstable, it is guaranteed that some or all of the component exists in the window.

Further, the rotation is performed at a predetermined step angle, and the inclination of the component is detected from the step angle of the maximum correlation value obtained by calculating the difference or differentiation of the correlation value of the matching. As a result, the same effect as increasing the resolution of the step angle can be obtained, and the accuracy of tilt detection can be improved.

Further, a farthest point from the center of the component is determined, and a region centered on this point is rotated based on the determined inclination of the component, and a region with a Y-axis symmetric or X-axis symmetric template in this region is determined. By performing matching, the inclination of the component is obtained. In this case, the inclination of the component obtained based on the rotation amount can be verified, and the inclination detection accuracy can be further improved.

[0015]

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

[Overall Configuration] FIG. 1 is a configuration diagram of a system for implementing a component position detecting method according to the present invention. In the following, the object to be processed is a part that is sucked by the suction nozzle of the chip mounter and that is point-symmetric with respect to its own center.

This system comprises a component imaging stage 1,
A position control device 2 for controlling the orientation of the component, a man-machine interface 3 for the operator to control the position control device 2 and the image processing device 12, a TV camera 4 for imaging the component, and a display for displaying the image. It comprises a display device 5 and an image processing device 12 which processes the image data and detects the center and inclination of the component. Further, the image processing device 12 includes a template memory 6 in which the stored template is expanded, an image memory 7 in which an image from the TV camera 4 is stored, a non-volatile memory 11 in which the created template is stored, and a template matching. An arithmetic unit 10 for performing calculations, an interface 8 with an external circuit, and a control unit 9 for controlling the flow and processing of each data.

FIG. 2 is a flowchart schematically showing the flow of the entire operation and processing. The position detection module drives based on recognition condition data created in advance so as to flexibly respond to various components and operating conditions (operation speed priority, accuracy priority, etc.). The operator can select a position detection process suitable for the system by creating a recognition condition module (data group) in consideration of the features and operation conditions of the components.

Hereinafter, these will be described.

First, in step S1, it is determined whether to use the reference window setting position function by template matching. This function is a function for assuring that a part or the whole of a component exists in the window even if the position of the component on the image is unstable.

If the position of the component is stable, the process proceeds to step S2 without using this function, and a reference window is determined. The reference window can be specified as a known coordinate value, and the window position can be changed according to the type of component. The operator sets the window area in the recognition condition module.

[Template Creation] On the other hand, if the position of the component is unstable, the process proceeds to step S3 to create a template. The template creation routine is shown in detail in FIG. First, an image including a component in a normal posture is input in advance (step T1). This imaging is shown in FIG.
An image including all the components set on the component imaging stage 1 is captured by the TV camera 4 and taken into the image memory 7. Then, this image is displayed on the display device 5. Since the component does not always have a normal posture on this image, the operator controls the position control device 2 via the man-machine interface 3 to move the component imaging stage 1 and correct the component posture (step T2, T3). During this operation, the image is fetched into the image memory 7 in real time and displayed on the display device 5 as described above. FIG. 7 schematically illustrates the state of the component posture correction, and the image of the component 20 that was initially at the position illustrated by the dotted line has the normal posture as illustrated by the solid line due to the above-described correction. .

Next, the operator, as shown in FIG.
Image processing device 12 via man-machine interface 3
A reference point R and an area for creating a template 21 in which a part or the whole of the component 20 is placed are designated (step T4). If the template is determined (step T5), the control unit 9 of the image processing apparatus receives the designation via the interface 8, obtains image data from the image memory 7, and registers the template in the nonvolatile memory 11. (Step T6). On the other hand, if the template to be registered has not been determined yet, the process returns to step T1 to repeat the above processing.

After the template is created and registered in this way, the process returns to step S4 in FIG. 2 and the template is set in the recognition condition module. Subsequently, when the inclination calculating function is used based on the relative positions of the corner points of the component described with reference to FIG. 14, the external dimensions of the component are set in the recognition condition module (step S6). This step S6 is skipped, a data group of the recognition condition is transmitted in step S7, and the position of the component is detected in step S8.

This position detection routine is shown in detail in FIG. First, as shown in step R1, a template or a window position is specified according to the recognition condition data set in step S2 or S4. If the component position on the image is stable, the window position is specified.
In, the reference window setting position is obtained. This is performed by the following processing.

[Reference Window] As shown in FIG.
The image of the component 20 to be inspected is fetched from the TV camera 4 into the image memory 7, and the control unit 9 stores the previously registered template 21 from the nonvolatile memory 11 into the template memory 6.
And leaves the calculation unit 10 to perform the matching calculation processing.
The matching process is performed, for example, by setting the template in the X direction and the Y direction.
The pixel is moved one pixel at a time in the direction, and each time it is correlated with the captured image of the corresponding area, the position with the highest degree of correlation is determined as the matching position, and the process ends. After confirming the completion of the matching operation by the operation unit 11, the control unit 9 can know the position of the template 21 on the component image to be inspected. This position is defined as the upper left, and a rectangular area having the template width and height as the width and height, respectively, is defined as the area 22 for acquiring the center position of the component, that is, the reference window.

[Detection of center position] Subsequently, step R2
The center position of the component is detected on the basis of the reference window position acquired in step (1) or the designated window position (step R3). The details are shown in the subroutine of FIG. The essence is template matching by the two point-symmetric templates shown in step R32. This is performed as shown in FIGS. On the image including the part, the window 22 (area A) in which a part or the whole of the part is to be inserted is set in the template 21 described above.
Set by template matching by. The real image of this window (or the reference window itself) is used as the first template, and a second template 23 that is point-symmetric (rotated by 180 degrees) is created with respect to this template (FIG. 10). As shown in FIG. 11, the area B, that is, the second template area 23 is detected at the stage where the image of the component and the second template 23 are matched. A center M of a straight line connecting an arbitrary point a in the area of the first template 22 and a point b in the area of the second template corresponding to the point is detected as the center point of the component.

When the real image of the window is used as the first template (branch to "yes" in step R30 in FIG. 5), the above-described matching using the point symmetric template is performed (step R35). However, if the reference window itself is used as the first template, the component may have an inclination and a rotational deviation from the template created in step S3 may occur. If the rotational deviation is large, the matching accuracy is lowered, and it is not always possible to set the optimum reference window region in step R2. Therefore, the process of Step R32 is repeated several times near the position obtained in the reference window (Step R31), and Step R33
Correct the reference window with. Among them, the combination having the highest correlation value is adopted to obtain the center position (step R34).

[Inclination Detection] Subsequently, returning to FIG. 4, in step R4, the inclination of the component is detected. This is performed according to the routine shown in FIG. In the case where the part is a figure that is point-symmetric with respect to its own center, in the XY coordinate system whose origin is set at the center of the part shown in FIG. An area 24 (window C) is set. This area is used as a template for tilt detection (step R4).
1). Assuming that the inclination of the component is θ, when the graphic in the window C is rotated by −2 * θ with the center of the component as the center of rotation, a graphic that is symmetric with respect to the graphic in the window A appears.

Using this relationship, the image is continuously changed in the window C at a predetermined angle of rotation. Matching is performed with the rotated image obtained at that time (step R44), and the step is updated at step R45 to rotate the figure (step R43), and this is repeated (step R42).

The correlation value of the matching when the rotation angle is updated becomes a distribution curve C as shown in FIG. 13, and the inclination θ of the component is determined by determining the rotation angle i at which the correlation value becomes maximum. (Step R46). Further, in order to improve the accuracy, the difference between the position i having the highest correlation value Y1 and the correlation values Y2 and Y3 at the step angles (i-1) and (i + 1) on both sides thereof is calculated, and the zero-cross of the difference value is calculated. By obtaining the point Z, the inclination θ can be obtained with a resolution equal to or greater than the step angle (R47). Further, as shown by a dotted line C ′ in FIG. 13, the zero-cross point of the distribution curve C may be obtained by differentiating the distribution curve C.

In the illustrated example, the Y-axis symmetric template 24 is created. However, it is also possible to determine the X-axis symmetric template and perform template matching in the same manner to determine the component inclination. .

In order to improve the tilt accuracy, it is necessary to make the notch angle smaller, but this requires a long calculation time. If the rotation shift falls within one pixel, the shift amount cannot be detected. Therefore, if it is desired to further improve the accuracy (branch to "Yes" in step R47 '), the center of the component and a certain degree of inclination are known in the processing up to this point. (Step R48), it is possible to set the best template window centered on the farthest point P of the component having the largest deviation amount (Steps R49, R50).
This is illustrated in FIG.

For example, since the external dimensions of the component are known, a window (D) 30 having the upper left corner P of the component substantially at the center can be set. When the image of the part is rotated by −2 * θ with the center of the window as the center of rotation, and a template 31 that is symmetric with respect to the Y axis is created,
The template is a template showing features near the upper right corner Q of the component. By moving this template in the X and Y directions and performing matching as shown in the lower part of FIG. 14, the position of the upper right corner Q of the part on the image can be obtained. Can be obtained (steps R51, R52).

If the window 30 used here is located farther from the center than the window 22 in FIG. 12, it is possible to expect higher accuracy than that obtained in FIGS. 12 and 13 by this method. At the time of this matching, the position can be obtained with sub-pixel accuracy by sub-pixel operation from the position having the highest correlation value and the nine correlation values in the vicinity thereof, and PQ is the combination having the largest X component difference. Therefore, the influence of the matching error is reduced, and better accuracy can be expected. In this way, the inclination obtained based on the amount of rotation can be verified and corrected to obtain a highly accurate inclination.

As described above, when the detection of the position of the component is completed, the flow returns to the flow shown in FIG. 2, and the detection result is received in step S9 following step S8. In step S10, the center position and / or inclination is determined based on the detection result. The correction is performed to correct the suction posture of the component.

[0037]

As described above, according to the present invention,
Using the template matching method, the center and inclination of the component can be accurately detected. Therefore, it can be implemented as a more general-purpose algorithm in a chip mounter or the like, and the electronic components can be more easily and accurately mounted.

When the inclination is obtained by rotating the image,
Since the matching area and the template size become equal, only one correlation operation is required for each inspection angle, and since the template is created from the image, there is no need for normalization. Therefore, the difference can be used as an evaluation function for matching,
Calculation time is reduced. In addition, since the matching area has already been determined, when a rotated image is acquired, the rotation conversion operation of only the matching area may be performed, and the calculation time can be shortened in this respect as well.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an overall configuration of a system in which the present invention is used.

FIG. 2 is a flowchart showing the entire flow.

FIG. 3 is a flowchart illustrating a routine for creating and registering a template.

FIG. 4 is a flowchart illustrating a routine for performing position detection.

FIG. 5 is a flowchart illustrating a routine for detecting a center position of a component.

FIG. 6 is a flowchart illustrating a routine for detecting the inclination of a component.

FIG. 7 is an explanatory diagram showing a state in which a component posture is corrected when a template is created.

FIG. 8 is an explanatory diagram showing creation of a template.

FIG. 9 is an explanatory diagram illustrating a state in which a reference window position is obtained.

FIG. 10 is an explanatory diagram showing a state in which a point-symmetric template is formed from a reference window.

FIG. 11 is an explanatory diagram showing matching with a point-symmetric template.

FIG. 12 is an explanatory diagram showing a state in which the inclination of a component is detected.

FIG. 13 is a diagram showing a distribution of correlation values by matching at the time of detecting a tilt.

FIG. 14 is an explanatory diagram showing another example of tilt detection.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 Component imaging stage 2 Position control device 3 Man-machine interface 4 TV camera 5 Display device 12 Image processing device 20 Components 21, 23, 31 Template 22, 24, 30 Window

Claims (6)

[Claims] 1. An image of a part having a point-symmetrical shape with respect to its own center is picked up, an area including a part or the whole of the image of the part is detected, and the area is determined with respect to an X axis or a Y axis passing through the center of the part. A component position detection method comprising: creating a template symmetrical to the above; and detecting the inclination of the component by matching the template with the image of the component. 2. An image of a part having a point-symmetric shape with respect to its own center is taken, an area including a part or the whole of the image of the part is detected, and a reference window itself or a real image which is point-symmetric with the area and is detected. A template replaced with data is created, and a midpoint connecting the corresponding area and the corresponding two points of the template when the matching between the image and the template is obtained is determined as the center of the component. A component position detection method comprising: creating a template symmetrical to the region with respect to the axis or the Y axis; and detecting the inclination of the component by matching the template with the image of the component. 3. Matching an image obtained by rotating the image of the part around its own center with the template symmetrical with the X-axis or Y-axis, and based on the amount of rotation at the time of the matching, The method according to claim 1, wherein the inclination is detected. 4. The template according to claim 1, wherein a template including a part or the whole of the part is created, and a template region in which an image can be matched with the template is set as the region. The method described in. 5. The method according to claim 1, wherein the rotation is performed at a predetermined step angle, and the inclination of the component is detected from the step angle of the maximum correlation value obtained by obtaining a difference or a derivative of the correlation value of the matching. The method according to 3 or 4. 6. A point farthest from the center of the component is determined, and a region centered on this point is rotated based on the determined inclination of the component, and the region is compared with a Y-axis symmetric or X-axis symmetric template of this region. Find the inclination of the part by matching,
The method according to any one of claims 3 to 5, wherein a part inclination obtained based on the rotation amount is verified.