JP3384758B2 - Component position recognition device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a component position recognizing apparatus for recognizing the position of a component such as a flat package IC having at least two projections such as electrodes. 2. Description of the Related Art In an automatic electronic component mounting apparatus for automatically mounting electronic components on a substrate, a target component is photographed by an image pickup device such as a camera, and the position of the target component is recognized using image processing. The target component is mounted at a predetermined position on the board based on the recognized position of the target component. In particular, in the case of a component whose electrode portion can be clearly identified, such as a flat package IC, it is possible to recognize the position of the entire component by detecting the electrode position. As a position recognition device for a flat package IC, there is one described in Japanese Patent Application Laid-Open No. 62-86789. In this position recognition device, an IC component that is roughly positioned in advance by mechanical means is imaged by an imaging device. A sample straight line is set for the image pattern of each electrode row of the IC component so as to be substantially orthogonal to the electrodes. The position of the electrode in each electrode row is determined by the change in brightness on the sample straight line, and the center position in the column direction of each electrode row is determined. Then, the position of the IC component is determined based on the determined center position in the column direction of each electrode row. [0004] By the way, a flat package, such as a quad flat package (QFP), in which electrodes protrude from four sides of a package body, SO
There are P (small out line package) in which electrodes protrude from two opposite sides of the package body. In the above-described conventional position recognition device, QFP (qua
d flat package) for ICs with electrodes protruding from the four sides of the package body, such as ICs
Since the position of the component is determined based on the center position in the column direction of the two electrode rows, highly accurate component position recognition can be performed. However, SOP (small out line pa)
In the case where an electrode protrudes from two opposing sides of the package body, such as an IC mounted with an integrated circuit (ckage), only the center of the two electrode rows in the column direction can be obtained. The component position with respect to the length direction of the electrode is accurately recognized, but the accuracy of the component position with respect to the length direction of the electrode is only guaranteed when the position is roughly determined in advance by mechanical means. Further, in the case of an IC having a protruding portion protruding from one side, the positioning accuracy is not further compensated. Further, since image data is processed, a long processing time is required. An object of the present invention is to provide a component position recognizing device capable of recognizing the position of a component having a projecting portion with high accuracy and high speed. A component position recognizing device according to the present invention is a component position recognizing device for recognizing the position of a component having a plurality of projecting portions. means and, in the grayscale image stored in the part image storage means, look including a tip portion for each projection of the component, and in a square shape having sides parallel to the projecting direction
From the image data of the set region, parallel to the projecting direction
Calculating means for calculating, for each region, a projection distribution in a different direction and a direction different from the direction, and connecting the projection distribution for each region calculated by the projection calculating means in each direction to form two types of primary
Projecting operation result connection storage means for storing as an original array , projection position calculating means for calculating the position of each projection from the one-dimensional array projection distribution stored in the projection operation result connection storing means, and the projection position Component position calculating means for calculating the position of the component based on the position of the protruding portion in each of the regions obtained by the calculating means. Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a schematic configuration of a component position recognition device. The component position recognition device includes an image memory 101,
Image data transfer means 102, horizontal projection calculation means 10
3. It comprises a vertical projection operation means 104, a projection operation result connection storage means 105, an electrode position detection means 106 and a position calculation means 107. The grayscale image data of the target component input from the imaging device (not shown) to the component position recognizing device is stored in the image memory 1.
01 is stored. The input data is digital data of 8 bits per pixel, and the image size is 512 × 512 pixels. Then, the entire image of the target component is included in one field of view. FIG. 2 shows an image of the target component imaged by the imaging device. Target part 1 is an 8-pin SOP
(small out line package), and four electrodes 2 protrude from two opposing sides of the package body. The image data transfer means 102 is provided in the projection calculation target area S (see FIG. 2) for the number of electrodes set so as to include the tip of the electrode 2 of the component to be recognized in the input image one by one. The image data is sequentially transferred to the horizontal projection operation unit 103 and the vertical projection operation unit 104. Each projection calculation target area S is set as follows, for example. After the target component 1 is positioned in advance by mechanical means, an approximate position of the target component 1 on the input image of each electrode 2 is specified by imaging the target component 1 with an imaging device. Then, based on the specified approximate position of each electrode 2, the projection calculation target area S
Is set in advance. Alternatively, the target part 1 is imaged by an image pickup device, and after storing the grayscale image data of the target part in the image memory 101, coarse position recognition is performed by appropriate image processing, and the position of each electrode 2 of the target part 1 is determined. Specify the approximate position on the input image. The projection calculation target area S is set based on the specified approximate position of each electrode 2. That is, coarse position data by mechanical coarse positioning or coarse position recognition result data by image processing;
Each projection calculation target area S is set based on the shape data of the target component such that the tip of each electrode 2 of the target component 1 is included in each projection calculation target area S. As shown in FIG. 2, when the target component 1 is imaged at an arbitrary angle with respect to the horizontal and vertical axes X and Y of the imaging screen, each projection calculation target region S is a parallelogram region. Is set as Two opposite sides of each projection calculation target area S are set so as to be parallel to the length direction of the electrode 2. The other opposing 2 of each projection calculation target area S
The side is set to be parallel to the horizontal axis X or the vertical axis Y of the imaging screen according to the tilt angle of the target component 1 with respect to the horizontal and vertical axes of the imaging screen. When the row direction of the electrode rows is closer to the horizontal direction than the vertical direction as shown in FIG. 2, one side of each projection calculation target area S is set to be parallel to the horizontal axis X of the imaging screen. . When the row direction of the electrode rows is closer to the vertical direction than the horizontal direction as shown in FIG. 3, one side of each projection calculation target area S is set so as to be parallel to the vertical axis Y of the image pickup screen. Is set. When the target component is imaged without an angle shift with respect to the horizontal and vertical axes X and Y of the imaging screen, each projection calculation target area S has two sides parallel to the horizontal direction and two sides parallel to the vertical direction. It has a rectangular shape with sides. FIG. 4 shows an example of the projection calculation target area S shown in FIG.
The figure shows a state in which image data is read out by the image data transfer means 102. As shown in FIG. 2, when the projection calculation target area S is a parallelogram having two sides parallel to the horizontal axis X of the imaging screen, the image data transfer direction is:
The direction is parallel to the horizontal axis X of the imaging screen as indicated by the arrow in FIG. Then, as shown in FIG. 5, the transfer start position Ps for each line is shifted in pixel units along a side parallel to the length direction of the electrode 2 in the projection calculation target area S. FIG. 6 shows an example of the projection calculation target area S shown in FIG.
The figure shows a state in which image data is read out by the image data transfer means 102. As shown in FIG. 3, when the projection calculation target area S is a parallelogram having two sides parallel to the vertical axis Y of the imaging screen, the image data transfer direction is:
The direction is parallel to the vertical axis Y of the imaging screen as indicated by the arrow in FIG. Then, the transfer start position for each line is shifted in pixel units along the side parallel to the length direction of the electrode 2 in the projection calculation target area S. FIG. 7 is an explanatory diagram of a projection calculation method by the horizontal projection calculation means 103 and the vertical projection calculation means 104. FIG. 7A shows a state in which image data is read from the projection calculation target area S in FIG. The image data in each projection calculation target area S is
Horizontal projection calculation means 103 and vertical projection calculation means 1
04 at the same time. In the horizontal projection calculating means 103, an arrow A
The pixel value is added in the projection direction indicated by. That is, assuming that a pixel value in the projection calculation target area S is f (x, y), as shown in Expression 1, for each horizontal transfer line,
The total sum H (y) of the density values of all the pixels included in each line is calculated. FIG. 7B shows a calculation result by the horizontal projection calculation means 103. As can be seen from FIG.
In the horizontal projection calculation result, a portion where the projection distribution sharply changes appears corresponding to the edge position in the length direction of the electrode. ## EQU1 ## The vertical projection calculating means 104 adds pixel values in the projection direction indicated by arrow B. The projection direction B is parallel to a side parallel to the length direction of the electrode 2 in the projection calculation target area S. That is, as shown in Expression 2, by performing a cumulative addition operation of the pixel values corresponding to the direction of the arrow B of each transfer line for each pixel on the transfer line, the total sum of the density values in the projection direction B is obtained. V (x) is calculated. FIG. 7C shows the calculation result by the vertical projection calculation means 104.
As can be seen from FIG. 7C, in the vertical projection calculation result, two portions where the projection distribution sharply changes corresponding to the edge position in the width direction of the electrode appear. ## EQU2 ## The projection calculation in the projection direction B is performed, for example, as follows. As shown in FIG. 8, one adder 301 and a number of buffers 302 corresponding to the number of pixels of one transfer line are prepared, and each buffer 302 is in the middle of the addition operation of the corresponding pixel value between the transfer lines. Store the result. By repeating the addition operation of the corresponding pixels between the transfer lines with the result of the addition and the sequentially transferred pixel data, the projection operation in the projection direction B is completed at the end of the transfer of all the pixel data in the projection operation target area S. The result is obtained. Thus, the horizontal projection calculating means 1
03 and the vertical projection operation means 104 execute a projection operation on the image data in each projection operation target area S transferred from the image data transfer means 102. The projection calculation result data obtained by the horizontal projection calculation means 103 and the vertical projection calculation means 104 is:
It is sent to the projection calculation result connection storage means 105. As shown in FIG. 9, the projection calculation result link storage means 105 links the projection calculation results for the number of the projection calculation target areas S for each horizontal projection calculation result and each vertical projection calculation result. Are stored as two types of one-dimensional arrays. FIG. 9A schematically shows linked data of the result of the horizontal projection operation, and FIG. 9C schematically shows linked data of the result of the vertical projection operation. Actually, the connection data of the horizontal projection operation results and the connection data of the vertical projection operation results corresponding to the eight projection operation target areas S shown in FIG. However, FIG. 9 shows only a part of the data, that is, only four pieces of linked data of the horizontal projection calculation results and four pieces of linked data of the vertical projection calculation results. When the connection data of the horizontal projection operation result and the connection data of the vertical projection operation result are stored in the projection operation result connection storage means 105, the electrode position detection means 1
At 06, edge detection processing is performed on each data. FIG. 9B shows the result of the edge detection processing on the concatenated data of the result of the horizontal projection operation. FIG. 9D shows an edge detection processing result for the concatenated data of the vertical projection operation result. That is, in the electrode position detecting means 106, a process such as a one-dimensional differential filter is performed on each connected data, and a portion where the projection distribution changes sharply, that is, the edge position in the length direction of the electrode 2 (See FIG. 9B) and two edge positions in the width direction of the electrode 2 (see FIG. 9D), and from these edge positions, the center position of the tip of each electrode 2 is determined. Detected as position. In the detection of two edge positions in the width direction of the electrode 2, the projection direction B is parallel to the electrode direction, so that the gradient of the projection distribution at a portion corresponding to the edge position becomes steep. Therefore, two edge positions in the width direction of the electrode 2 can be accurately detected. On the other hand, in detecting the edge position in the length direction of the electrode 2, the projection direction A is the horizontal direction (in the case of FIG. 4) or the vertical direction (in the case of FIG. 6) regardless of the electrode direction. In general, the width of the electrode 2 is sufficiently smaller than its length, so that even if the target component 1 is inclined obliquely, the slope change of the projection distribution of the portion corresponding to the edge position in the length direction of the electrode 2 becomes less sharp. small. Therefore, the edge position in the length direction of the electrode 2 can be detected with high accuracy. FIG. 10 is an explanatory diagram for explaining a position calculating method by the position calculating means 107. The position calculating unit 107 calculates the position of the center of gravity of the target component 1 and the inclination angle of the target component 1 with respect to the coordinate axis based on the positions of all the electrodes 2 detected by the electrode position detecting unit 106.
The coordinates (Gx, Gy) of the position of the center of gravity of the target component 1 are as follows. When n electrodes are completely symmetrically arranged like a component on which SOP or QFP is mounted, the total number of the electrodes 2 is n and each electrode Assuming that the position coordinates of (2) are (xk, yk) (1.ltoreq.k.ltoreq.n), they can be obtained based on the following equation (3). ## EQU3 ## That is, the average value of the x coordinate of each electrode position is x
The coordinates become Gx, and the average value of the y coordinates of each electrode position becomes the y coordinate Gy of the position of the center of gravity. In addition, the inclination angle of the target component 1 with respect to the coordinate axis can be obtained by assuming a straight line L connecting the positions of all the electrodes 2 projecting from one side of the target component 1 in one direction, and using the inclination of the straight line L. Regarding a part whose electrodes are not arranged symmetrically, it is possible to calculate the inclination angle of the part with respect to the coordinate axis of the center of gravity of the part and the coordinate axis of the part by using data of the electrode position required according to the shape of the part. . In the above embodiment, since the positions of all the electrodes 2 of the target component are detected, it is also possible to check the bending of each electrode 2 in a two-dimensional plane based on the position data of these electrodes 2. It is possible. In the above embodiment, the position of the target component is recognized by obtaining the positions of all the electrodes of the target component 1. However, the position of the target component is recognized by obtaining the positions of some of the electrodes. It is also possible. In this way, the number of projection calculation target areas can be reduced, and higher-speed processing can be performed. Taking the target part 1 of FIG. 2 as an example, the electrodes at both ends of the electrode row are selected from each of the two electrode rows, and the positions of the selected four electrodes are detected. , Coordinates of the position of the center of gravity of the target part 1 and the target part 1
Can be calculated with respect to the coordinate axis. Further, the position of any two electrodes of the target component 1 is detected, and based on the detected two electrode positions and the shape data of the target component, the coordinates of the center of gravity of the target component 1 and the position of the target component 1 are determined.
Can also be calculated with respect to the coordinate axis. If the accuracy of detecting the rough position of the target component for setting the projection calculation target region S is not sufficiently obtained,
In the following manner, the projection calculation target area S can be set as a suitable area by increasing the detection accuracy of the coarse position of the target component. That is, first, the positions of several electrodes of the target component are detected by the same method as in the above-described embodiment, and the rough position of the target component is obtained based on the detection result and the shape data of the target component. Thereafter, the projection calculation target area S is updated based on the obtained coarse position. Then, using the updated projection calculation target area S, the position of the target component is obtained by the same method as in the above embodiment. Such a process may be repeated a plurality of times to update the projection calculation target area S, thereby improving the position recognition accuracy of the target component. First, the positions of all the electrodes protruding in any one direction of the target component are detected by the same method as in the above embodiment, and other electrode positions are detected based on the detection result and the shape data of the target component. The positions of all the electrodes protruding in the direction are detected. The projection calculation target area S is updated based on the positions of all the electrodes thus obtained. Then, using the updated projection calculation target area S, the position of the target component is obtained by the same method as in the above embodiment. Such a process may be repeated a plurality of times to update the projection calculation target area S, thereby improving the position recognition accuracy of the target component. FIG. 11 shows a schematic configuration of another position recognition device. This component position recognizing device comprises an image memory 20
1, an image data transfer unit 202, a horizontal projection operation unit 203, a vertical projection operation unit 204, a projection operation result storage unit 205, an electrode position detection unit 206, and a position calculation unit 207. The grayscale image data of the target component input from the imaging device (not shown) to the component position recognizing device is stored in the image memory 2.
01 is stored. FIG. 12 illustrates an image of the target component captured by the imaging device. The target component 1 is an IC on which an 8-pin small out line package (SOP) is mounted, and four electrodes 2 protrude from two opposing sides of the package body. The image data transfer means 102 transfers the image data of the projection calculation target area S (see FIG. 12) set so as to include only one electrode of the target component in the input image to the horizontal projection calculation means 203. And to the vertical projection calculation means 204. The method for setting the projection calculation target region S is the same as the method for setting the projection calculation target region in the component position recognition device of FIG. The shape of the projection calculation target area S is shown in FIG.
Is a parallelogram having two sides parallel to the length direction of the electrode 2 and two sides parallel to the horizontal axis X of the imaging screen. Therefore, the reading direction of the image data in the projection calculation target area S by the image data transfer unit 202 is a direction parallel to the horizontal axis X of the imaging screen as shown in FIG. Then, as shown in FIG. 5, the transfer start position Ps for each line is shifted in pixel units along a side parallel to the length direction of the electrode 2 in the projection calculation target area S. When image data in one projection calculation target area S is sent to the horizontal projection calculation means 203 and the vertical projection calculation means 204, each calculation means 203
Perform a projection operation. FIG. 13A shows how image data in the projection calculation target area S is read out, as shown in FIG.
FIG. 1 shows the projection result obtained by the horizontal projection operation means 203.
3 (c) shows a projection result by the vertical projection calculation means 204. The projection calculation method by each of the calculation means 203 and 204 is the same as the calculation method described with reference to FIG. The horizontal projection calculation result and the vertical projection calculation result for the image data in one projection calculation target area S obtained by each calculation means 203 are stored in the projection calculation result storage means 205. Thereafter, the electrode position detecting means 20
6, edge detection processing is performed on each data. FIG. 13D shows an edge detection processing result with respect to the horizontal projection calculation result. FIG. 13E shows an edge detection processing result with respect to the vertical projection calculation result. That is, in the electrode position detecting means 206, processing such as a one-dimensional differential filter is performed on the horizontal projection calculation result and the vertical projection calculation result, and the portion where the projection distribution changes sharply, that is, the electrode 2 An edge position in the length direction (see FIG. 13D) and two edge positions in the width direction of the electrode 2 (see FIG. 13E) are detected, and from these edge positions, the center position of the tip of the electrode 2 is detected. Are detected as the position of the electrode 2. Such an electrode position detection process is repeatedly executed for the number of electrodes necessary for detecting the position of the target component 1. Here, it is assumed that the electrode position detection processing has been executed for the electrodes 2 at both ends of each electrode row of the target component 1 in FIG. FIG. 14 is an explanatory diagram for explaining a position calculating method by the position calculating means 207. The position calculating means 207 calculates the center of gravity position G of the target component 1 and the inclination angle of the target component 1 with respect to the coordinate axis based on the positions P1, P2, P3, P4 of the four electrodes 2 detected by the electrode position detecting means 206. Is done. That is, the electrodes 2 facing the target component 1
Based on the position coordinates P1 and P3, and P2 and P4, coordinates C1 and C2 of a position exactly intermediate between the electrodes facing the target component 1 are obtained. The inclination of the straight line S connecting the coordinates C1 and C2 of the two points is obtained as the inclination angle with respect to the coordinate axis of the target component 1. In addition, the position coordinates just intermediate between the coordinates C1 and C2 of the two points are obtained as the center of gravity position G of the target component 1. In the above embodiment, the position of the center of gravity of the target component 1 and the inclination angle of the target component 1 with respect to the coordinate axis are detected by detecting the positions of the four electrodes 2 among the eight electrodes 2 of the target component 1. Although the calculation is performed, the positions of all the electrodes may be detected, and the position of the center of gravity of the target component 1 and the inclination angle of the target component 1 with respect to the coordinate axis may be calculated based on the positions of all the electrodes. The position of any two electrodes of the target component 1 is detected, and based on the detected two electrode positions and the shape data of the target component, the coordinates of the center of gravity of the target component 1 and the target component 1 are determined. Can also be calculated with respect to the coordinate axis. Further, by the same method as in the above embodiment,
The positions of the two electrodes of the target component are detected, and the rough positions of the other two electrodes are determined based on the detection result and the shape data of the target component. The projection calculation target area S may be set, and the other two electrode positions may be obtained using the set projection calculation target area S. In the above description, the position recognition of the component on which the flat package is mounted has been described. However, if the component has two or more protruding portions, the position of the component can be recognized by the position recognition device according to the present invention. According to the present invention, the position of a component having at least two projecting portions can be recognized with high accuracy and high speed.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is an electric block diagram showing a schematic configuration of a component position recognition device. FIG. 2 is a schematic diagram showing a target component image and a projection calculation target area when the column direction of the electrode row of the target component image is closer to the horizontal direction than the vertical direction. FIG. 3 is a schematic diagram showing a target component image and a projection calculation target region when the column direction of the electrode row of the target component image is closer to the vertical direction than the horizontal direction. FIG. 4 is a schematic diagram showing a transfer direction of image data in a projection calculation target area shown in FIG. 2; FIG. 5 is a schematic diagram showing a transfer start position of each transfer line when transferring image data in a projection calculation target area shown in FIG. 2; FIG. 6 is a schematic diagram showing a transfer direction of image data in a projection calculation target area shown in FIG. 3; FIG. 7 is a schematic diagram showing the results of projection calculation by horizontal projection calculation means and vertical projection calculation means. FIG. 8 is an electric block diagram showing a specific configuration example of a vertical projection calculation unit. FIG. 9 is a schematic diagram showing connection data of the projection operation result stored in the projection operation result connection storage means and an edge processing result for the connection data. FIG. 10 is a schematic diagram for explaining a method of calculating the coordinates of the position of the center of gravity of the target component and the inclination angle of the target component with respect to the coordinate axis. FIG. 11 is an electric block diagram illustrating a schematic configuration of another component position recognition device. FIG. 12 is a schematic diagram showing a target component image and a projection calculation target region. FIG. 13 is a schematic diagram showing the results of the projection calculation by the horizontal projection calculation means and the vertical projection calculation means, and the results of the edge processing performed on the projection calculation results. FIG. 14 is a schematic diagram for explaining a method of calculating the coordinates of the position of the center of gravity of the target component and the inclination angle of the target component with respect to the coordinate axis. [Description of Signs] 1 Target part 2 Electrode 101, 201 Image memory 102, 202 Image data transfer means 103, 203 Horizontal projection operation means 104, 204 Vertical projection operation means 105 Projection operation result connection storage means 205 Projection operation result storage Means 106, 206 Electrode position detecting means 107, 207 Position calculating means

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-5-196424 (JP, A) JP-A-6-18243 (JP, A) JP-A-2-91509 (JP, A) 28309 (JP, U) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06T 1/00 305 G01B 11/00 H05K 13/04 H05K 13/08

Claims (1)

(57) [Claim 1] In a component position recognizing device for recognizing the position of a component having a plurality of protrusions, a component image storage means for storing a gray image of the component, and the component image storage in grayscale image stored in the unit, seen including a tip portion for each projection of the component, and a square shape having sides parallel to the projecting direction
From the set region image data to, and the projecting direction Rights
A projection calculation means for calculating, for each region, a projection distribution in a line direction and a direction different from the direction, and connecting the projection distribution for each region calculated by the projection calculation means for each direction, and
Projection operation result connection storage means for storing as a dimensional array ,
A projection position calculating means for calculating the position of each projection from the one-dimensional array of projection distributions stored in the projection calculation result connection storage means; and a projection within each of the regions determined by the projection position calculation means And a component position calculating means for calculating the position of the component based on the position of the component.