JPH04337404A - Position measurement device - Google Patents

Abstract

PURPOSE:To accurately measure the position of a solder by comparing a peripheral distribution determined through a peripheral distribution part with a threshold so as to detect the end point position of each material to be measured. CONSTITUTION:Each emission diode 24 of a semi-spherical cover 22 is emitted by a control device 31. In this situation, electronic parts are photographed by a photographing device 25 and an image signal is output. The image signal is sent to an image process device 30. The end point of each lead 40, i.e., the root position of a fillet 41 is detected by the image process device 30. Picture element levels (concentration level) are accumulated in the vertical direction of each lead 40 and peripheral distribution for each lead 40 of the image data is determined by a peripheral distribution map 32. The accumulated concentration level and a threshold R are compared with each other from the root of each lead 40 for each peripheral distribution by a point detection part 34, so as to detect the end point of each lead 40. A luminescent spot detection region Q is set for each end point by the image process device 30, and the shape of solder is measured for each luminescent spot detection region Q.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention is a position measurement method applied to measuring the soldering position when measuring the shape of an object to be measured having a mirror surface, such as a soldered part of an electronic component soldered on a board. Regarding equipment.

0002

2. Description of the Related Art Inspection of the soldered state of electronic components has been mainly performed visually. In order to eliminate such manual inspection work, automated inspection devices and shape measuring devices are being developed. FIG. 26 is a configuration diagram of such a device, in which light from an illumination device 3 is irradiated obliquely to each pin 2 of an electronic component 1, and the reflected light is imaged by a camera 4 to inspect the soldering state. It is something to do. In this technique, as shown in FIGS. 27 and 28, if there is no solder, the light will not be reflected directly above, so if the camera 4 captures an image from directly above, the screen will be dark. Also, if there is solder, the screen will be brighter. In this way, the soldering condition is measured from the degree of brightness and darkness of the screen.

However, although such measurements can detect the presence or absence of soldering, it is difficult to measure excess or deficiency in the amount of solder, and the reliability is low due to the influence of the surface condition of the solder, making it impractical.

FIG. 29 is a configuration diagram of a shape measuring device using the optical cutting method. This device scans each lead 7 of an electronic component 6 mounted on a board 5 with slit light or spot light 9 from a slit light source 8, and captures an image at this time with a camera 10 to check the soldering condition. It is something to be measured. In this case, the camera 10 captures an image from a direction different from the direction of incidence of the spot light 9.

However, in this device, since the solder is in a mirror-like state, the reflected light of the obliquely incident spot light 9 cannot be recognized by one camera 10, and the measurement time becomes long.

[0006] Also, as another technique, Japanese Patent Application Laid-Open No. 63-7607
It is conceivable to apply the stereo photometric method described in Publication No. 3. In this stereo illuminance difference method, multiple light sources are sequentially switched to illuminate the solder to create an illuminance difference map, and from this illuminance difference map, the slope of the solder corresponding to each pixel is determined.
These are tied together to measure the shape of the solder.

However, since the stereo photometric method requires measurement on a completely diffused surface, it is difficult to measure on mirror-like solder. This is because the illuminance varies depending on the inclination of the solder surface with respect to the direction when the light is irradiated.
In other words, the shape is measured using the difference in illuminance, and as a result, almost no light enters the camera. Therefore, the image obtained by the camera is not an image with a difference in illuminance, but only a binarized image. However, it is difficult to measure the shape of a mirror-like object such as solder. Further, in the stereo photometric method, it is necessary to store in advance the illuminance for the slope of the solder. For this reason,
The reliability of measurement accuracy is low when the surface characteristics of solder change, for example, when the surface condition changes depending on the amount of flux.

On the other hand, a hemispherical cover with a large number of point light sources arranged inside is placed above the electronic component, and among the point light sources of this hemispherical cover, each point light source is arranged in the same direction as the lead of the electronic component. A camera captures each image when the lights are turned on in sequence. There is a technique for measuring the shape of the solder from the position of the reflected light from the solder when each light source in these images is turned on.

However, although this technique can measure the shape of the solder with high precision, there is a possibility that reflected light from areas other than the solder portion may be captured, and therefore, the soldering position of each lead cannot be accurately detected before measuring the solder shape. There is. Therefore, a portion other than the soldering position may be determined to be the soldering position and an erroneous measurement may be made.

0010

As described above, even if the shape can be measured with high precision, the solder position cannot be detected accurately. SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a position measuring device that can accurately measure the position of solder, etc.

[0011]

[Means for Solving the Problems] The present invention provides a position measuring device that detects the tip position of each object to be measured from an image obtained when light is irradiated from a predetermined area above each object to be measured in a plurality of strips. , a peripheral distribution section that calculates the peripheral distribution by accumulating the pixel levels of each measured object, and a tip that detects the tip position of each measured object by comparing the peripheral distribution determined by this peripheral distribution section with a threshold value. The present invention is a position measuring device including a detecting section to achieve the above object.

The present invention also provides a position measuring device for detecting the tip position of each object to be measured from an image obtained by irradiating light from a predetermined area above each object to be measured in a plurality of strips. A peripheral distribution section that calculates the peripheral distribution by accumulating the pixel levels of the body, a maximum and minimum detection section that calculates the maximum and minimum values of the peripheral distribution determined by this peripheral distribution section, and a maximum and minimum detection section that calculates the maximum and minimum values of the peripheral distribution determined by this peripheral distribution section. A position in which the above objective is achieved by including a tip detection unit that searches for a trail in the peripheral distribution from the maximum value to the minimum value and detects the intersection of this trail and a threshold value as the tip position of each measured object. It is a measuring device.

The present invention also provides a position detection device for detecting the tip position of each object to be measured from an image obtained when light is irradiated from an oblique direction onto each of the objects to be measured in a plurality of strips. A position measuring device that aims to achieve the above purpose by setting a long region in the direction, determining the peripheral distribution in the longitudinal direction, and detecting the side edges of each object to be measured from this peripheral distribution. be.

In this case, the edge detection means includes a peripheral distribution detecting section which sets a long region in the longitudinal direction of the object to be measured and obtains a peripheral distribution in this longitudinal direction, and a peripheral distribution detected by the peripheral distribution detecting section. and an edge detection section that performs pattern matching on the object to be measured determined by the distribution width judgment means to detect side edges of the object to be measured. are doing.

[0015]

[Operation] By providing such a means, the pixel level of each object to be measured in an image when light is irradiated from a predetermined area above each of the plurality of strip-shaped objects to be measured is accumulated by the peripheral distribution part. A peripheral distribution is obtained using the method, and the tip detection section compares this peripheral distribution with a threshold value to detect the tip position of each object to be measured.

[0016] Also, the maximum and minimum values of the marginal distribution are determined by the maximum/minimum detection section, and then the tip detection section searches for the trail of the marginal distribution from the maximum value to the minimum value, and then calculates the trail between this trail and the threshold value. Detect the intersection point as the tip position of each object to be measured.

Furthermore, by providing the above means, the edge detection means detects a long area in the longitudinal direction of the object to be measured in an image obtained when light is irradiated from an oblique direction onto each of the plurality of strip-shaped objects to be measured. Then, the peripheral distribution in the longitudinal direction is obtained, and the side edges of each object to be measured are detected from this peripheral distribution.

In this case, the edge detection means sets a long area in the longitudinal direction of the object to be measured, determines the peripheral distribution in this longitudinal direction, determines the object to be measured from the distribution width of this peripheral distribution, and determines the area to be measured. Pattern matching is performed on the object to be measured to detect side edges of the object to be measured.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of a shape measuring device to which the position measuring device of the present invention is applied. A board 21 on which electronic components are mounted by soldering is placed on the XY table 20. A hemispherical cover 22 that blocks external disturbance light is supported by a support mechanism 23 and disposed above the XY table 20. This hemispherical cover 22 has a plurality of light emitting diodes (LEDs) 24 arranged inside it. The light emitting diodes 24 are arranged every 5 degrees in the latitude direction and every 10 degrees in the longitude direction. This hemisphere cover 22
An imaging device 25 is arranged above the substrate 21 and almost directly above the substrate 21 . An image output terminal of this imaging device 25 is connected to an image processing device 30. In addition, XY table 2
0 and the hemispherical cover 22 are connected to a control device 31, which controls the movement of the XY table 20 and the light emission of each light emitting diode 24.

This image processing device 30 stores the image signal from the image pickup device 25 as image data, measures the soldering position of each lead of the electronic component from this image data, and determines the soldering position from this image data of the soldering position. It has the function of measuring shape. Specifically, it has the functions of a peripheral distribution section 32, a maximum/minimum detection section 33, a tip detection section 34, and an edge detection section 35.

The peripheral distribution unit 32 has a function of accumulating the pixel levels of each lead to obtain a peripheral distribution, and the maximum/minimum detection unit 33 determines the maximum and minimum values of the peripheral distribution determined by the peripheral distribution unit 32. It has a function.

The tip detection section 34 also has a first function of detecting the tip position of each lead by comparing the peripheral distribution obtained by the peripheral distribution section 32 with a threshold value, and a maximum/minimum detection section 33.
It has a second function of searching the trail of the peripheral distribution from the maximum value to the minimum value determined by , and detecting the intersection of this trail and the threshold value as the tip position of each lead.

The edge detection section 35 has a function of setting a long region in the longitudinal direction of each lead of the electronic component, determining the peripheral distribution in this longitudinal direction, and detecting the side edge of each lead from this peripheral distribution. ing. Specifically, the edge detection unit 35 includes a peripheral distribution detection unit that sets a long area in the longitudinal direction of the lead and calculates the peripheral distribution in this longitudinal direction, and a peripheral distribution detection unit that determines the peripheral distribution from the distribution width of the peripheral distribution determined by this peripheral distribution detection unit. It has a distribution width determination unit that determines leads, and a function of performing pattern matching on the leads determined by the distribution width determination unit to detect side edges of the leads. Note that the image processing device 30 includes a monitor television 36.
is connected. Next, the operation of position measurement in the device configured as described above will be explained.

The control device 31 controls each light emitting diode 24 shown in FIG. 2 among the light emitting diodes 24 of the hemispherical cover 22;
That is, from 70 degrees to 90 degrees upward when viewed from each lead of the electronic component.
50 degrees to 7 degrees when viewed from each lead of the electronic component.
Each light emitting diode 24 emits light in an oblique direction of 0 degrees. In addition, in FIG. 2, each light emitting diode 24 indicated by a black circle emits light.

In this state, the imaging device 25 images the electronic component and outputs the image signal. This image signal is sent to the image processing device 30, which digitizes the image signal and stores it as image data shown in FIG. By the way, this image data is from 70 degrees shown in FIG.
Image data when each light emitting diode 24 emits light in a 90 degree area (FIG. 5) and image data when each light emitting diode 24 emits light in an oblique direction of 50 degrees to 70 degrees shown in FIG. 6 (FIG. 7) It is a composite of. The image data shown in FIG. 3 is for each lead 40 and fillet (
The soldering part) 41 is shown.

Next, the image processing device 30 detects the tip position of each lead 40, that is, the base position of the fillet 41. As shown in FIG. 8, the peripheral distribution unit 32 accumulates pixel levels (shade levels) in the vertical direction of the leads 40 for each lead 40 of the image data to determine each peripheral distribution. Next, the tip detection section 34 compares the cumulative density level and the threshold value R for each peripheral distribution from the base of each lead 40 to detect the tip position of each lead 40.

When the tip positions of each lead 40 are detected in this way, the image processing device 30 sets bright spot detection areas Q at these tip positions. Thereafter, the shape of the solder is measured for each bright spot detection area Q.

In this way, the peripheral distribution of the lead 40 is determined,
Since the cumulative density level of this peripheral distribution and the threshold value R are compared to detect the tip position of each lead 40,
The soldering position of each lead 40 can be detected accurately. Next, other position detection will be explained.

The peripheral distribution section 32 of the image processing device 30 is shown in FIG.
As shown in the figure, the peripheral distribution is obtained by accumulating the gray level in the vertical direction of each lead 40 in one image data. Next, the tip detection section 34 compares the cumulative density level of this peripheral distribution with the threshold value S from the base of the lead, and detects the point of change in the cumulative density level as the tip position of each lead 40. Note that the threshold value is the maximum value Ma and minimum value M of the cumulative gray level.
i, number of leads n, threshold value = (Ma-Mi)/
n+Mi
...(1) It is determined by n>1. When the tip positions of each lead 40 are detected in this way, the image processing device 30 sets bright spot detection areas Q at these tip positions in the same manner as described above. Thereafter, the shape of the solder is measured for each bright spot detection area Q.

In this way, the peripheral distribution is obtained by accumulating the gray level in the vertical direction of each lead 40 in one image data, so the soldering position of each lead 40 can be detected accurately, and the image noise is reduced. The influence of can be removed. That is, when the cumulative density level is determined for each lead 40 as shown in FIG. 8, the tip position of the lead 40 may be incorrectly detected due to the influence of image noise as shown in FIG. The influence of image noise can be removed by accumulating the gray levels. Next, further position detection will be explained.

Similarly to the above, the peripheral distribution section 32 is connected to each lead 40.
Find the marginal distribution by accumulating the pixel levels of . At this time,
If there is an insufficiently extracted portion of the image on the root side of each lead 40, the density level of the peripheral distribution will be low on the root side, as shown in FIG. Therefore, there is a possibility that the position of the tip of the lead 40 may be erroneously detected. In such a case, the maximum/minimum detection unit 33 detects the maximum value Ma and the minimum value Mi of the marginal distribution as shown in FIG.
seek. In this case, the maximum/minimum detection unit 33 detects the maximum value Ma
is on the lead 40 side with respect to the minimum value Mi, based on the cumulative density level change of the entire peripheral distribution. Next, the tip detection section 32 searches the edges of the peripheral distribution from the maximum value Ma toward the minimum value Mi to find the intersection K with the threshold value S, and detects this intersection K as the tip position of the lead 40. Note that the threshold value S is as shown in the above equation (1).

In this way, the maximum value Ma and the minimum value Mi of the marginal distribution are determined, and the trail of the marginal distribution is searched from the maximum value Ma toward the minimum value Mi, and the intersection K with the threshold value S is found at the tip of the lead 40. Since it is detected as a position, not only can the lead tip position be accurately detected, but also the lead tip position can be accurately detected even if there is a portion of the root side of each lead 40 where the image is insufficiently extracted. Next, detection of the side edge position of each lead 40 will be described.

Detection of this side edge position is performed by the edge detection section 35 of the image processing device 30. The peripheral distribution detecting section of the edge detecting section 35 sets a long region W in the longitudinal direction of each lead 40 for each lead 40, as shown in FIG. Find the marginal distribution. Figure 14 shows one lead 40
shows the marginal distribution of Next, the distribution width judgment unit determines the maximum and minimum values of the peripheral distribution, sets a threshold value F from the minimum value, detects the intersection between this threshold value F and the peripheral distribution, and detects the side edge of the lead 40. Remove unnecessary data during detection. That is, as shown in FIG. 15, if the peripheral distribution is mountain-shaped, the distribution width determining unit determines it as a lead, and if it shows a one-sided slope, it determines it as noise and removes it. Next, as shown in FIG. 16, the distribution width determination unit determines the width A of the mountain-shaped portion in the peripheral distribution from the intersection of the cumulative density level of the peripheral distribution and the threshold value F, and compares it with the arbitrarily determined width H. . As a result of this comparison, if the width A of the chevron-shaped portion is narrower than the width H as shown in FIG. 17, the distribution width determination section removes it as noise. Next, the edge detection unit performs one-dimensional pattern matching on the leads 40 determined by the distribution width determination unit to detect side edges of the leads 40. That is, the edge detection section sets an area measurement area E having the same width H as the width H of the lead 40, as shown in FIG. 18, and calculates the area of the peripheral distribution that falls within this area measurement area E. In addition, in FIG. 18, the area of the peripheral distribution that falls within the area measurement area E is zero. Then, the edge detection section moves the area measurement area E as shown in FIG. 19, and along with this movement, sequentially determines the area of the peripheral distribution. In this way, the edge detection section determines the area of the peripheral distribution that falls within the area measurement area E, and determines the area where this area is maximum, that is, the area where the lead 40 and the area measurement area E match, as shown in FIG. . Thus, the edge detection section detects the side edge L of the lead 40 from this one-dimensional pattern matching.

In this way, the peripheral distribution in the longitudinal direction of the lead 40 is determined, the lead 40 is determined from the distribution width of this peripheral distribution, and one-dimensional pattern matching is performed on the lead 40 to detect the side edge L of the lead 40. As a result, the side edges of the leads 40 can be detected accurately. Next, the measurement of the shape of the solder after detecting the tip position of each lead 40 and setting each bright spot detection area Q will be described.

The control device 31 drives the XY table 20 to create a bright spot detection area Q at the center of the screen imaged by the imaging device 25.
In other words, set it so that the soldering part will fit inside. Since the board 21 has been positioned in advance on the XY table 20, this setting is based on this positioning data.
Move 0.

After this positioning, the control device 31 arranges one row of the light emitting diodes 24 of the hemispherical cover 22, which are arranged along the longitudinal direction of the leads of the electronic components, on the upper part of the hemispherical cover 22. The light emitting diodes 24 that have been installed are caused to emit light one by one. At the same time, the imaging device 25 images the electronic component and the solder portion each time the light emitting diode 24 emits light, and outputs an image signal thereof. This image signal is sent to the image processing device 30 and stored as image data. FIG. 21 is a schematic diagram of such image data. In this image data, A to D indicate reflected light portions when the light emitting diodes 24 are sequentially caused to emit light.

This image processing device 30 calculates the centers of gravity (representative points) G1, G2, G3 of each reflected light portion A to D of image data.
... and each reflected light portion A to D shown in FIG.
Find the slopes Θ1, Θ2, Θ3... In addition,
These inclinations Θ1, Θ2, Θ3, . . . are obtained from the arrangement position of each light emitting diode 24 at the time of imaging. Next, the image processing device 30 calculates the midpoints C1, C2, etc. between the respective centers of gravity G1, G2, G3, etc., and also calculates the midpoints C1, C2, etc.
, Θ2... is found. Next, the image processing device 30
Distance L1 between and C1, distance L2 between C1 and C2
Find the height position Z1 of each midpoint C1, C2...
, Z2... is found. These heights Z1, Z2...Zi are

[0039]

[Math 1]

It is obtained by: As a result, the shape of one line of the soldered portion is measured. However, by sequentially causing the other rows of light emitting diodes 24 to emit light, the shape of the entire soldered portion shown in FIG. 23 can be determined.

As described above, according to the above-mentioned embodiment, the position of the soldering part can be detected with high accuracy, and even if this position detection is affected by image noise or there is insufficient image extraction, the lead can be accurately performed. The soldered part can be detected from the tip position. As a result, the area where image data is captured when measuring the shape of the soldered part is reduced to the bright spot detection area Q.
It is possible to measure shapes with high accuracy by removing the effects of other image noise.

It should be noted that the present invention is not limited to the above-mentioned embodiment, and may be modified without changing the gist thereof. For example, when detecting a side edge of a lead, the threshold value F at that time may not be fixed, but may be variable as shown in FIG. In this case, it is possible to deal with insufficient image extraction.

[0043] In addition, the imaging device is arranged in a manner shown in Fig. 2 to avoid blind spots.
A plurality of them may be arranged as shown in 5. These imaging devices 5
0 to 53 is an angle of 90 in the horizontal direction centered on the hemispherical cover 22
They are arranged at angles of 30 degrees. These imaging devices 25, 50 to 53 may be of epi-illumination type by providing a half mirror on the hemispherical cover 22. Furthermore, instead of providing a plurality of light emitting diodes of the hemispherical cover 22, one light emitting diode may be moved.

[0044]

As described in detail above, according to the present invention, it is possible to provide a position measuring device that can accurately measure the position of solder, etc.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing an embodiment of a position measuring device according to the present invention.

FIG. 2 is a diagram showing light emission from a light emitting diode in a hemispherical cover during position measurement with the same device.

FIG. 3 is a schematic diagram of image data during position measurement with the same device.

FIG. 4 is a diagram showing light emission from a light emitting diode when obtaining a fillet position in the same device.

FIG. 5 is a schematic diagram of image data when obtaining fillet positions.

FIG. 6 is a diagram showing light emission of a light emitting diode when obtaining a side edge.

FIG. 7 is a schematic diagram of image data when obtaining side edges.

FIG. 8 is a diagram showing the peripheral distribution for each lead.

FIG. 9 is a diagram showing the peripheral distribution according to the cumulative density level of each lead.

FIG. 10 is a diagram showing the peripheral distribution when image extraction is insufficient for each lead.

FIG. 11 is a diagram showing the peripheral distribution when image extraction is insufficient due to cumulative gray levels of multiple leads.

FIG. 12 is a diagram showing the operation when detecting the lead tip position from the maximum and minimum density levels.

FIG. 13 is a diagram showing an area when detecting a side edge of a lead in the device of the present invention.

FIG. 14 is a diagram showing the cumulative gray level of the same area.

FIG. 15 is a diagram showing image noise removal from the cumulative gray level.

FIG. 16 is a diagram showing the measurement effect of the lead width from the cumulative density level.

FIG. 17 is a diagram showing image noise removal from the cumulative gray level and an arbitrarily determined lead width.

FIG. 18 is a diagram showing the setting of area measurement regions used for one-dimensional pattern matching.

FIG. 19 is a diagram showing movement of area measurement area during pattern matching.

FIG. 20 is a diagram showing a state when matching is achieved by the same pattern matching.

FIG. 21 is a diagram showing reflected light from solder when each light emitting diode emits light during shape measurement.

FIG. 22 is a diagram showing a shape determined from light emission from one row of light emitting diodes in shape measurement.

FIG. 23 is a diagram showing a three-dimensional shape obtained from light emission from multiple rows of light emitting diodes in shape measurement.

FIG. 24 is a diagram showing a modified example of lead side edge detection.

FIG. 25 is a diagram showing a configuration when a plurality of imaging devices are used.

FIG. 26 is a configuration diagram of a conventional shape measuring device.

FIG. 27 is a diagram showing the operation of the device.

FIG. 28 is a diagram showing the operation of the device.

FIG. 29 is a configuration diagram of a conventional shape measuring device.

[Explanation of symbols]

20...XY table, 21...board, 22...hemisphere cover,
24... Light emitting diode, 25... Imaging device, 30... Image processing device, 31... Control device, 32... Peripheral distribution section, 33... Maximum minimum detection section, 34... Tip detection section, 35... Edge detection section.

Claims (4)

[Claims] 1. A position measuring device that detects the tip position of each of the objects to be measured from an image obtained when light is irradiated from a predetermined area above each of the objects to be measured in a plurality of strips arranged. A peripheral distribution section that calculates a peripheral distribution by accumulating pixel levels, and a tip detection section that detects the tip position of each object to be measured by comparing the peripheral distribution obtained by the peripheral distribution section with a threshold value. A position measuring device characterized by: 2. A position measuring device that detects the tip position of each of the objects to be measured from an image obtained when light is irradiated from a predetermined area above each of the objects to be measured in a plurality of strips arranged. A marginal distribution section that calculates a marginal distribution by accumulating pixel levels, a maximum and minimum detection section that determines the maximum and minimum values of the marginal distribution determined by this peripheral distribution section, and a maximum value determined by this maximum and minimum detection section. A position measuring device comprising: a tip detection unit that searches for a trail in the peripheral distribution from 0 to a minimum value and detects an intersection between this trail and a threshold value as a tip position of each of the objects to be measured. . 3. A position detection device that detects the tip position of each of the plurality of strip-shaped objects to be measured from an image obtained by irradiating light from an oblique direction to each of the plurality of strip-shaped objects to be measured. 1. A position measuring device comprising an edge detection means for determining a peripheral distribution in the longitudinal direction by setting a long area with respect to the object, and detecting a side edge of each of the objects to be measured from the peripheral distribution. 4. The edge detection means includes a peripheral distribution detecting section that sets a long region in the longitudinal direction of the object to be measured and obtains a peripheral distribution in the longitudinal direction, and a peripheral distribution detected by the peripheral distribution detecting section. distribution width determining means for determining the object to be measured from the distribution width of the distribution width; and an edge detecting section for detecting side edges of the object by performing pattern matching on the object to be measured determined by the distribution width determining means. The position measuring device according to claim (3).