JP4491630B2 - Electronic component lead direction discrimination method and apparatus - Google Patents

Description

  The present invention relates to an electronic component lead direction determination method and apparatus for determining whether the lead extraction direction of the electronic component is two directions or four directions based on image data of the electronic component.

  In recent years, in an electronic component mounting machine, when an electronic component sucked by a suction nozzle is mounted on a circuit board, the electronic component is imaged by a camera, and the electronic component is recognized by image processing technology. . At that time, as for electronic components with leads, as shown in Patent Document 1 (Japanese Patent No. 2941617), components such as lead length, lead pitch, lead width, number of leads, etc. are previously provided for each vertical and horizontal lead row. Data representing the characteristics is created as electronic component image recognition component data and stored in memory. During operation of the electronic component mounter, the electronic component image data is processed and read for each vertical and horizontal lead row. Some devices detect the length, lead pitch, lead width, number of leads, etc., and compare the detected data with the component data stored in the memory to recognize the suction posture of the electronic component.

This Patent Document 1 describes a method for automatically creating component data for electronic component image recognition. In this component data automatic creation method, image data of an electronic component is processed, and data such as a lead length, a lead pitch, a lead width, and the number of leads is calculated as component data for each vertical and horizontal lead row.
Japanese Patent No. 2941617 (page 4, etc.)

  By the way, among electronic components with leads, there are a type in which leads are taken out from the package body in two directions and a type in which leads are taken out in four directions. In addition, the shape of the lead of the electronic component is not a straight rectangle but a bent three-dimensional shape, so that the reflected state of the illumination light when taking an image with the camera is not uniform. As shown in FIG. 5A, the bent portion A of the lead may appear dark and one lead may be divided into two leads and recognized. In such a case, in the conventional image recognition technology, the leads at both ends of each lead row are placed in the other direction as shown in FIG. There is a problem in that it is erroneously determined as a lead, erroneously determined as a type in which the lead is taken out in four directions, and erroneous component data is automatically created.

  Such a problem can be solved if the operator manually specifies the lead extraction direction. However, in the manual operation specification by the operator, the lead extraction direction is incorrectly specified due to an operator error. In addition to eliminating the possibility of doing so, there is also a problem that it takes time and effort to create part data, contrary to the requirement for automation of part data creation.

  The present invention has been made in view of such circumstances. Therefore, the object of the present invention is to automatically determine whether the lead extraction direction of the electronic component is two or four based on the image data of the electronic component. It is an object of the present invention to provide an electronic component lead direction discriminating method and apparatus capable of discriminating it.

  In order to achieve the above object, according to the present invention, there is provided a method for determining whether the lead extraction direction of the electronic component is two directions or four directions based on the image data of the electronic component captured by the image data capturing means. A normalization process for normalizing the coordinates of the reference position of each lead of the electronic component based on the image data of the component, and calculating a variation index value (for example, variance or standard deviation) of the normalized coordinates of the reference position of each lead. A variation index value calculation process to be performed, and a lead direction determination process to determine whether the lead extraction direction of the electronic component is two directions or four directions by comparing the variation index value with a predetermined threshold value. It is a thing. Here, the reference position of the lead may be a position determined by a constant reference, such as the center, tip, base, or position away from the tip or base by a predetermined [%].

  The present inventors normalize the coordinates of the reference position of each lead with respect to various electronic components having different lead extraction directions, the number of leads, the package shape, etc., and the variation index value of the normalized coordinates of the reference position (for example, (Variance or standard deviation) is calculated and the distribution of the variation index values is examined. As shown in FIG. 4, the distribution index values (for example, dispersion) of the two directions lead and the variation index of the four directions lead are calculated. It was found that the distribution of values (for example, variance) appears separately. Specifically, it has been found that the distribution index value distribution in the two directions leads to a larger distribution index value distribution than the distribution index value distribution in the four directions.

  Focusing on this characteristic, in the present invention, a threshold value is set in advance between the distribution index value distribution in two directions and the distribution index value distribution in four directions, and image data capturing means The variation index value calculated from the image data of the electronic component captured by the above is compared with a threshold value, and it is determined whether the lead extraction direction is two directions or four directions depending on whether the variation index value is larger than the threshold value. It is what you do. In this way, it is possible to automatically and accurately determine whether the lead extraction direction of the electronic component is two directions or four directions based on the image data of the electronic component.

  In this case, it is considered that the orientation (horizontal direction and vertical direction) of the electronic component image captured by the image data capturing unit is often shifted from the X direction or the Y direction of the reference coordinates. In addition, after correcting the rotation angle of the image of the electronic component so that the lead extraction direction of the electronic component coincides with the X direction or the Y direction of the reference coordinate, the X coordinate and Y coordinate of the reference position of each lead are normalized. It is good to make it. As a result, the X and Y coordinates of the reference position of each lead can be normalized without being affected by the orientation of the image of the electronic component captured by the image data capturing means. Then, the respective variation index values of the normalized X coordinate and Y coordinate are calculated, and if either the X coordinate variation index value or the Y coordinate variation index value is equal to or greater than the threshold value, the lead extraction direction It is good to determine that is in two directions. In this way, when the lead take-out direction is two directions, the lead take-out direction is accurate to determine whether the lead take-out direction is two directions or four directions, regardless of which direction is the X direction or the Y direction. It can be automatically distinguished well.

  In addition, the method for normalizing the coordinates of the reference position of each lead is not limited to a specific normalization method, and various normalization methods can be used. However, as shown in FIG. It is preferable to normalize the X coordinate and Y coordinate so that they fall within the range of -1.0 to +1.0. In this way, the variation index value (dispersion, standard deviation, etc.) falls within the range of 0 to +1.0, and determination is easy.

Hereinafter, an embodiment embodying the best mode for carrying out the present invention will be described.
The lead direction discriminating apparatus used in the first embodiment uses a component data creation system that creates component data representing the characteristics of an electronic component from image data of the electronic component, and as shown in FIG. A computer 11, an image data capturing means 12 such as a camera for capturing image data of parts, an input device 13 such as a keyboard and a mouse, a display device 14 such as a liquid crystal display and a CRT, and programs and data described later. It has a configuration including a storage device 15 for storing.

  The computer 11 executes a read direction discriminating program shown in FIG. 2 to be described later, thereby serving as normalization means, variation index value calculating means, and read direction discriminating means in the claims. To determine whether the lead-out direction of the electronic component taken in is two directions or four directions. The determination of the lead extraction direction is performed through rotation angle correction processing, normalization processing, variation index value calculation processing, and lead direction determination processing. Hereinafter, the contents of these processes will be described.

[Rotation angle correction processing]
It is considered that the orientation (horizontal direction and vertical direction) of the electronic component captured by the image data capturing unit 12 often deviates from the X or Y direction of the reference coordinates. Therefore, before performing the normalization process, the rotation angle of the image of the electronic component is corrected so that the lead extraction direction of the electronic component matches the X direction or the Y direction of the reference coordinates. Regarding this rotation correction, an operator looks at the image of the electronic component displayed on the screen of the display device 14 to determine the rotation angle of the electronic component, and operates the input device 13 to correct the rotation angle of the electronic component. However, it is desirable to automatically correct the rotation angle of the electronic component by calculating the rotation angle of the electronic component (lead row) with respect to the reference coordinate (XY coordinate) by image processing technology. When the lead extraction direction is two directions, the rotation angle may be corrected so that the lead extraction direction coincides with either the X direction or the Y direction of the reference coordinates.

  If this rotation angle correction process is performed, the X and Y coordinates of the reference position of each lead can be normalized without being affected by the orientation of the image of the electronic component captured by the image data capturing unit 12.

[Normalization processing]
In the normalization process, the X coordinate and Y coordinate of the reference position of each lead of the electronic component are extracted based on the image data of the electronic component after the rotation angle correction, and the X coordinate and Y coordinate are normalized. Here, the reference position of the lead may be a position determined by a constant reference, such as the center, tip, base, or position away from the tip or base by a predetermined [%].

  In addition, the method for normalizing the coordinates of the reference position of each lead is not limited to a specific normalization method, and various normalization methods can be used. However, as shown in FIG. It is preferable to normalize the X coordinate and Y coordinate so that they fall within the range of -1.0 to +1.0. In this way, the variation index value (dispersion, standard deviation, etc.) falls within the range of 0 to +1.0, and determination is easy.

[Variation index value calculation processing]
In the variation index value calculation process, each variation index value of the normalized X coordinate and Y coordinate is calculated. In this embodiment, the variances σx ² and σy ² are calculated by the following equations as the variation index values.

σx ² = 1 / n · {(x1 -xav) ² + (x2 -xav) ² + ... + (xn -xav) ² }
σy ² = 1 / n · {(y1 −xav) ² + (y2 −xav) ² + …… + (yn −xav) ² }
Here, n is the total number of leads, x1, x2,..., Xn is the X coordinate of the reference position of each lead, y1, y2, ..., yn are the Y coordinates of the reference position of each lead, and xav is the lead of each lead. The average value of the X coordinate of the reference position, yav, is the average value of the Y coordinate of the reference position of each lead, and is calculated by the following equation.

xav = 1 / n · {x1 + x2 + …… + xn}
yav = 1 / n. {y1 + y2 + ... + yn}
As distribution index value, variance sigma] x ^2, instead of .sigma.y ^2, the standard deviation sigma] x, may be calculated .sigma.y (the square root of the variance).

The present inventors normalize the coordinates of the reference position of each lead for various electronic components having different lead extraction directions, the number of leads, the package shape, and the like, and the variances σx ² and σy of the normalized coordinates of the reference position. to calculate the ^2, its variance sigma] x ^2, was examined the distribution of .sigma.y ^2, the variance of the two directions of the lead as shown in FIG. 4 sigma] x ^2, variance sigma] x ² distribution and 4 direction of the lead of .sigma.y ^2, .sigma.y It was found that the distribution of ² appears separately. Specifically, the two directions of the lead dispersion sigma] x ^2, variance sigma] x ² distribution in the four directions lead .sigma.y ^2, compared to the distribution of .sigma.y ^2, sigma] x ^2, one increases the distribution of .sigma.y ² Turned out to be.

Focusing on this characteristic, in the present embodiment, in advance, the two directions of the variance sigma] x ² lead, variance sigma] x ² of the distribution and the four-way lead .sigma.y ^2, intermediate the distribution of .sigma.y ^2, 2 direction of the lead And a threshold value (for example, 0.7) for discriminating between four directions of leads.

[Lead direction discrimination processing]
In the read direction determination process, the image data capturing means 12 by the dispersion sigma] x ² from the image data of the electronic component was calculated through the above-mentioned respective processes incorporating the .sigma.y ² each compared with a threshold value, any sigma] x ² and .sigma.y ² Whether or not one of them is larger than the threshold value determines whether the lead extraction direction is two directions or four directions.

  The determination of the read direction described above is executed by the computer 11 as follows according to the read direction determination program of FIG. When this program is started, first, in step 101, the rotation angle of the image of the electronic component so that the lead extraction direction of the electronic component captured by the image data capturing means 12 coincides with the X direction or the Y direction of the reference coordinates. Correct. Thereafter, the process proceeds to step 102, where the X and Y coordinates of the reference position of each lead of the electronic component are extracted based on the image data of the electronic component. In the next step 103, as shown in FIG. The X coordinate and the Y coordinate of the reference position are normalized so as to fall within the range of −1.0 to +1.0, respectively.

Then, the procedure proceeds to step 104, each of the variance sigma] x ² of X and Y coordinates normalized to calculate the .sigma.y ^2. In the next step 105, the variance sigma] x ^2, to compare the .sigma.y ² respectively with a threshold (e.g. 0.7), it is greater than one threshold of sigma] x ² and .sigma.y ^2, step 106 the process proceeds, it is determined that the two directions of the leads, if both sigma] x ² and .sigma.y ² is equal to or smaller than the threshold, the process proceeds to step 107, determines that the four-way lead.

According to the embodiment described above, the coordinates of the reference position of each lead from the image data of the electronic component taken by the image data capturing means 12 is normalized, variance sigma] x ² normalized coordinates of the reference position, .sigma.y ² is calculated, since one of sigma] x ² and .sigma.y ² are as extraction direction of the lead on whether greater than the threshold value to determine whether two-way or four-way, based on the image data of the electronic component Thus, it is possible to automatically determine whether the lead extraction direction of the electronic component is two directions or four directions with high accuracy. This eliminates the possibility of erroneously specifying the lead take-out direction due to an operator's operation error, satisfies the requirement for automation of component data creation, and saves labor and time required for component data creation. It can be omitted.

It is a block diagram which shows the structure of the read direction discrimination | determination apparatus used in one Example of this invention. It is a flowchart which shows the flow of a process of a read direction discrimination | determination program. (A) is a figure which shows the capture image of an electronic component, (b) shows the state normalized so that X-coordinate and Y-coordinate of the reference position of each lead may fall in the range of -1.0 to +1.0, respectively. FIG. 2 direction of the lead of the variance sigma] x ^2, distribution and 4 direction of the lead of the variance sigma] x ² of .sigma.y ^2, a diagram illustrating an example of the distribution of .sigma.y ^2. (A) is a figure which shows the capture image of an electronic component, (b) is a figure explaining an example which misdetermines the lead row of 2 directions as a lead of 4 directions.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Computer (normalization means, dispersion | variation index value calculation means, lead direction discrimination means), 12 ... Image data capture means, 13 ... Input device, 14 ... Display device, 15 ... Storage device

Claims (8)

In the electronic component lead direction determining method for determining whether the lead extraction direction of the electronic component is two directions or four directions based on the image data of the electronic component captured by the image data capturing means.
Normalization processing for normalizing the coordinates of the reference position of each lead of the electronic component based on the image data of the electronic component;
A variation index value calculation process for calculating a variation index value of the normalized coordinates of the reference position of each lead;
A lead direction discriminating process comprising: comparing the variation index value with a predetermined threshold value and discriminating whether the lead extraction direction of the electronic component is two directions or four directions. Method. Before performing the normalization process, the rotation angle of the image of the electronic component is corrected so that the lead extraction direction of the electronic component matches the X direction or the Y direction of the reference coordinates,
In the normalization process, the X coordinate and the Y coordinate of the reference position of each lead of the electronic component are respectively normalized,
In the variation index value calculation process, each variation index value of the normalized X coordinate and Y coordinate of the reference position of each lead is calculated,
In the lead direction determination process, when one of the variation index value of the X coordinate and the variation index value of the Y coordinate is equal to or greater than the threshold value, the lead extraction direction of the electronic component is two directions. The method according to claim 1, wherein the electronic component lead direction is determined.   3. The electronic component according to claim 2, wherein in the normalization process, normalization is performed so that an X coordinate and a Y coordinate of the reference position of each lead fall within a range of −1.0 to +1.0, respectively. Lead direction discriminating method.   4. The electronic component lead direction determination method according to claim 1, wherein the variation index value is a variance or a standard deviation. In an electronic component lead direction discriminating apparatus for discriminating whether the lead extraction direction of the electronic component is two directions or four directions based on the image data of the electronic component captured by the image data capturing means.
Normalizing means for normalizing the coordinates of the reference position of each lead of the electronic component based on the image data of the electronic component;
Variation index value calculating means for calculating a variation index value of a normalized coordinate of the reference position of each lead;
A lead direction discrimination means for comparing the variation index value with a predetermined threshold value and discriminating whether the lead extraction direction of the electronic component is two directions or four directions. apparatus. The normalizing unit corrects the rotation angle of the image of the electronic component so that the lead extraction direction of the electronic component matches the X direction or the Y direction of the reference coordinate before performing the normalization process. Normalize the X and Y coordinates of the reference position of each lead of the electronic component,
The variation index value calculation means calculates a variation index value of each normalized X coordinate and Y coordinate of the reference position of each lead,
The lead direction discriminating means is such that when one of the variation index value of the X coordinate and the variation index value of the Y coordinate is not less than the threshold value, the lead extraction direction of the electronic component is two directions. 6. The electronic component lead direction discriminating apparatus according to claim 5, wherein the electronic component lead direction discriminating device is discriminated.   The electronic component according to claim 6, wherein the normalizing unit normalizes the X coordinate and the Y coordinate of the reference position of each lead so as to fall within a range of −1.0 to +1.0, respectively. Lead direction discriminating device.   8. The electronic component lead direction determination device according to claim 5, wherein the variation index value calculation means calculates a variance or a standard deviation as the variation index value.

Images