JPH1117400A - Packaging part/inspecting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an accurate part image by teaching, and at the same time to accurately judge the quality of the packaging state of a part by comparing images. SOLUTION: A part-unloading image, obtained at an image input part 1 before a part B is mounted, is stored in a part 2 for storing an unloading substrate image, a part-loading image corresponding to the position of the part B is compared with a part-unloading image that is stored in the part for storing the unloading substrate image and corresponds to the part at an image comparison part 3, and the quality of packaging state of the part B is judged at a judgment part 4 based on the comparison result.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a mounted component inspection apparatus for capturing an image of a component mounted on a board and inspecting the mounted state.

[0002]

2. Description of the Related Art Conventionally, as a device for inspecting a mounted state of a component mounted on a substrate such as a printed wiring board, a mounted component inspection device for automatically inspecting by using image processing has been considered.

FIG. 26 is a block diagram for explaining a conventional mounted component inspection apparatus. That is, the mounted component inspection apparatus includes a board moving mechanism 10 that places a board S on which a component B is mounted and moves to a predetermined position, an image input unit 1 that captures an image of the component B on the board S, A part image storage unit 6 for cutting out only a designated part range from the input image and storing the extracted image, and an input image and a part image storage unit 6
An image comparison unit 3 that compares the image stored in the image comparison unit and outputs the similarity, and a determination unit that determines whether the component is correctly mounted based on the magnitude of the similarity output from the image comparison unit 3 4 is provided.

The procedure of component inspection using this mounted component inspection apparatus is roughly divided into teaching for acquiring a non-defective component image and actual inspection. FIG.
Is a flowchart of the teaching process, and FIG. 28 is a flowchart of the actual inspection process. In the following description, reference numerals not shown in FIGS. 27 and 28 refer to FIG.

First, as step S801 shown in FIG. 27, a non-defective board (one on which components are mounted at an accurate position) is set on the board moving mechanism 10. Next, step S80
In 2, in order to obtain an image of a component to be inspected, the board moving mechanism 10 is moved, and the component is arranged in the imaging area of the image input unit 1.

Next, in step S 803, an image of the part is captured by the image input unit 1. Thereafter, as shown in step S804, a cutout range is set so that only the image of the component can be cut out from the fetched image.

Next, in step S 805, an image within the previously set cutout range, that is, an image of only the component, is stored in the component image storage unit 6. In step S806,
It is determined whether there is a next part, and if so, steps S802 to S805 are repeated. If there is no next part, the teaching ends. As a result, images of all components on the board can be stored.

Next, an actual inspection process will be described. First,
As shown in step S901 in FIG. 28, a substrate to be inspected (substrate to be inspected) is set on the substrate moving mechanism 10.
Next, in step S902, the board moving mechanism 10 is moved to obtain an image of the component to be inspected, and the component is arranged in the imaging area of the image input unit 1.

Next, in step S 903, an image including the component is captured by the image input unit 1. Thereafter, as shown in step S904, the entire region of the image including the component captured in step S903 is sequentially compared with the inspection target component image stored in the component image storage unit 6 by the teaching described above, The most similar (highest similarity) part is searched for and the similarity at that position is output.

In step S905, it is determined whether or not the similarity is equal to or higher than a predetermined reference. If the similarity is equal to or higher than the reference, it means that the two images are very similar, and the process proceeds to step S906 to determine a good product. That is, the fact that the component to be inspected is very similar to the image of a non-defective product stored in advance indicates that the component to be inspected is mounted at an accurate position.

On the other hand, if the similarity does not reach the reference, it means that the two images are not similar, and the flow advances to step S907 to determine a defective product. That is, the fact that the component to be inspected is not similar to the image of a non-defective product stored in advance indicates that the component to be inspected is not mounted at an accurate position.

In step S908, it is determined whether there is a next part. If there is a next part, step S
902 to S905 are repeated, and a good product or a defective product is determined based on the similarity. When the determination of non-defective products and defective products for all components is completed, the inspection processing is completed.

[0013]

However, such a mounted component inspection apparatus has the following problems. FIG.
9 is a diagram showing a state of image comparison, where a teaching image (1) is a case where only an image of a component is accurately cut out, and a teaching image (2) is a case where a cutout including an image other than the component is performed.

In the teaching described with reference to the flowchart of FIG. 27, when only the image of the part is accurately cut out as in the teaching image (1), FIG.
If the parts are mounted correctly as shown in
As shown in (b), if there is some displacement but it is within the allowable range, the similarity becomes equal to or higher than a predetermined criterion in comparison with the teaching image (1), and it is possible to judge a good product.

On the other hand, as shown in the teaching image (2),
When clipping including an image other than a part is performed, an image serving as a reference for comparison includes an image other than the part to be inspected, so that even a slight displacement of the part does not satisfy the similarity. In other words, when the components are normally mounted as shown in FIG. 29A, the non-defective product can be determined, but as shown in FIG. Even in this case, a defective product is determined.

That is, in the mounted component inspection apparatus, it is necessary to accurately cut out only the image of the component in the teaching performed in advance. In a conventional mounted component inspection apparatus, it is necessary to manually set the cutout range in order to cut out the teaching accurately.

However, setting the cut-out range by referring to the component images one by one requires a very time-consuming operation.

It is also conceivable to automatically set the cut-out range based on the size and mounting position of components using design data and the like. In this case, there is a problem that an accurate cutout cannot be performed.

[0019]

SUMMARY OF THE INVENTION The present invention is an apparatus for inspecting a mounted component to solve such a problem. That is, the present invention provides an unmounted image storage unit that stores a component unmounted image corresponding to the position of a component on the board before the component is mounted, and an unmounted image storage unit that stores the component on the board after the component is mounted. Image comparing means for comparing the corresponding component mounted image with the component non-mounted image corresponding to the component stored in the non-mounted image storage means, and judging the quality of the component mounting state based on the result of the comparison by the image comparing means Determining means for performing the determination.

An unmounted image storage means for storing an unmounted image corresponding to the position of the component on the board before the component is mounted, and an unmounted image storage means for storing the unmounted image corresponding to the position of the component on the board after the component is mounted. Mounted image storage means for storing a component mounted image to be mounted, and an area unit corresponding to the planar appearance of the component, based on the component unmounted image stored in the non-mounted image storage means and the component mounted image stored in the mounted image storage means The mounting component inspection apparatus further includes an image comparing unit that sequentially compares the components, and a position determining unit that determines the position of the component based on a result of the comparison by the image comparing unit.

A first mounted image storage means for storing a first mounted image captured from a predetermined angle of the component mounted on the board, and a first mounted image storage means for storing the first mounted image at a different angle from the first mounted image. A second mounted image storage unit for storing a second component mounted image, a first component mounted image stored in the first mounted image storage unit, and a second mounted image storage unit stored in the second mounted image storage unit. A mounted component inspection apparatus that includes an image comparison unit that sequentially compares a component mounted image with an area corresponding to the external shape of the component in a plan view and a position determination unit that determines a position of the component based on a result of the comparison by the image comparison unit. is there.

A cross image projecting means for projecting a cross image on a peripheral area centered on the component mounted on the board; and an image of a portion of the cross image projected on the component and a portion projected on the peripheral area. And a position determining means for determining the position of the component from the distortion of the mounted component.

In the present invention, since the component unmounted image on the board before the component is mounted and the component mounted image on the board after the component is mounted are compared by the image comparison means, the mounting state of the component is affected. This makes it possible to perform image comparison based on a component-unmounted image that is not affected by the image.

Further, by comparing the images of the component-unmounted image and the component-mounted image sequentially in units of regions corresponding to the external shape of the component in plan view, the component is mounted in a portion where the image difference is the largest. This makes it possible to accurately determine the position of the reference component.

When comparing the first component-mounted image and the second component-mounted image, the angle at which the image is taken is different from each other, so that the position at which the component image appears in each image is different. That is, by sequentially comparing the two images in the area unit corresponding to the external shape of the component in plan view, it can be understood that the component is mounted on a portion where the difference between the images is the largest.

Further, when the cross image is projected by the cross image projecting means on the peripheral region centering on the component, image distortion occurs in a portion of the cross image projected on the component and a portion projected on the peripheral region. In other words, the fact that the cross image projected on the part of the component is distorted with respect to the surrounding area depending on the height of the component can be used to calculate the position of the component based on the distorted part.

[0027]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a perspective view showing an embodiment of a mounted component inspection apparatus according to the present invention. FIG. 1 is a block diagram illustrating the first embodiment. That is,
The mounted component inspection apparatus according to the first embodiment includes a board moving mechanism 10 on which a board S on which a component B is mounted is mounted;
And the like, and an image of a portion where the component B is not mounted (hereinafter, referred to as an “unmounted image”).
, An image comparing unit 3 for comparing images, a determining unit 4 for performing a pass / fail determination based on the comparison result, and a component mounting position database DB for storing the position of the component B. Have.

Among such configurations, the substrate moving mechanism 10
Obtains component mounting position data from the component mounting position database DB, and moves the substrate S placed in the XY direction so that the component B to be inspected enters the imaging area of the image input unit 1.

The image input unit 1 is, for example, a CCD (Char
ge Coupled Device), which transmits a captured image to the non-mounted substrate image storage unit 2 and the image comparison unit 3.

The unmounted board image storage 2 stores a component B on the board S
Is not mounted, an unmounted image of a portion corresponding to the position of the component B is obtained from the image input unit 1 and stored as a reference in a later image comparison.

The image comparison unit 3 obtains an image of the component B to be inspected (hereinafter, referred to as “mounted image”) from the image input unit 1 in a state where the component B is mounted on the substrate S, and A comparison is made with an unmounted image of the corresponding component B obtained from the image storage unit 2. In the image comparison, a portion that is most similar to the unmounted image is searched for from the input image, and is output as the similarity at that position.

The judging section 4 judges whether or not the component B to be inspected is correctly mounted based on the similarity obtained from the image comparing section 3. That is, the similarity obtained from the image comparison unit 3 is compared with a predetermined reference value, and a good product and a defective product are determined based on the comparison result.

Next, a component inspection process in the mounted component inspection apparatus according to the first embodiment will be described with reference to the flowcharts of FIGS. 2 and FIG. 3 refer to FIG.

The component inspection process is divided into teaching of an unmounted image shown in the flowchart of FIG. 2 and actual inspection shown in the flowchart of FIG.

First, in step S101, the part B
Is set on the substrate moving mechanism 10. Next, in step S102, the board moving mechanism 10 is moved so that the position at which the component B to be inspected is mounted enters the imaging area of the image input unit 1. At this time, the board moving mechanism 10
The position (X, Y coordinates) of the target part B is obtained from B,
The movement is performed based on this.

Next, at step S103, the image input unit 1 takes in the unmounted image at that position.
The captured unmounted image is sent to the unmounted board image storage unit 2 and stored therein.

Thereafter, in step S104, a cutout range corresponding to the external shape of the component B in plan view is set. Then, in step S105, the image of the portion corresponding to the cutout range of the unmounted image is stored. by this,
An image in a non-mounted state only at the position where the component B is to be mounted is stored.

In step S106, it is determined whether or not the next part B is present.
S105 is repeated. If there is no next part B, the teaching ends. As a result, unmounted images corresponding to the positions of all the components B on the substrate S can be stored.

After storing the unmounted image, an actual inspection is performed. First, as shown in step S201 of FIG. 3, a substrate to be inspected is set on the substrate moving mechanism 10. Next, in step S202, the substrate moving mechanism 10 is moved,
The component B to be inspected enters the image pickup area of the image input unit 1. At this time, the board moving mechanism 10 obtains the position (X, Y coordinates) of the target component B from the component mounting position database DB, and moves based on this.

Next, in step S203, the mounted image of the component B at that position is captured by the image input unit 1. Then, in step S204, the mounted image of the component B captured earlier is compared with the unmounted image corresponding to the position of the component B obtained from the unmounted board image storage unit 2. By this comparison, the similarity is output from the image comparison unit 3.

In step S205, the determination unit 4 determines whether or not the similarity is equal to or less than a predetermined reference. First
In the embodiment, when inspecting the component B, an unmounted image corresponding to the position where the component B is mounted is captured, and the unmounted image is compared with the mounted image of the component B. Therefore, when the similarity is equal to or less than a predetermined reference value (not similar), it means that the component B is correctly mounted, and conversely, when the similarity is not equal to or less than the predetermined reference value (similar), This means that the component B is not mounted or is not mounted at an accurate position.

That is, if the similarity is equal to or smaller than the predetermined reference in step S205, the result is Yes, and the non-defective item is determined in step S206. On the other hand, if the similarity is not equal to or less than the predetermined reference in step S205, the result is No, and the defective product is determined in step S207.

In step S208, it is determined whether or not the next part B is present.
S205 is repeated. If there is no next part B, the inspection ends.

FIG. 4 is a diagram showing a state of image comparison.
(A) is an unmounted image (within the dashed line in the figure), (b) is an inspection image (normal), (c) is an inspection image (normal: there is some displacement), and (d) is an inspection image (out of stock). is there.

When an unmounted image shown in (a) is stored, the images are compared based on this. For example,
In the case of the inspection image shown in (b), since the part B is mounted on a portion corresponding to the unmounted image in (a), (a)
The similarity between the unmounted image and the mounted image in (b) is very small.

In the case of the inspection image shown in FIG.
The part B is mounted on the part corresponding to the unmounted image of (a) with a slight shift, and the unmounted image of (a) and the mounted image of (c) are partially similar but mostly different. However, the similarity is also equal to or less than a predetermined reference.

On the other hand, in the case of the inspection image shown in FIG.
No component is mounted on the portion corresponding to the unmounted image in (a), and almost completely matches the unmounted image in (a). That is, since the degree of similarity is extremely large, it can be seen that the component B is not mounted.

As described above, in the first embodiment, since the unmounted image in which the component B is not mounted by the teaching is stored, if the component B to be inspected is not mounted, the image is displayed. And the unmounted image almost completely match. This makes it possible to reliably determine that the component B is not mounted.

Further, since the comparison is made with the mounted image to be inspected based on the unmounted image which is not affected by the mounted state of the component B, the mounting state of the component B (tolerable range) is determined based on the similarity between the two images. ) Can be accurately determined.

Next, a second embodiment will be described. FIG. 5 is a block diagram illustrating the second embodiment. That is,
The mounted component inspection apparatus according to the second embodiment includes a board moving mechanism 10, an image input unit 1, an unmounted board image storage unit 2, an image comparison unit 3, a determination unit 4, and a component mounting position database D.
B is the same as the mounted component inspection apparatus of the first embodiment, except that the transmitted light illumination device 11 is provided at the position of the substrate S in the substrate moving mechanism 10.

The transmitted light illuminating device 11 irradiates the lower surface of the substrate S with light, and when the image input unit 1 captures an image, the component B
The hole (through hole) for inserting the lead is raised in white.

FIG. 6 is a diagram for explaining the difference between unmounted images.
(A) shows the case where the transmitted light illuminating device is not used, (b) shows the case where the transmitted light illuminating device is used, and (c) shows the case where the flow soldering is performed in the state of the missing part.

That is, as shown in FIG. 6A, the image input unit 1 does not use the transmitted light illumination device 11 (see FIG. 5).
When an unmounted image is captured by (see FIG. 5), the hole H for inserting the lead of the component provided on the substrate S becomes a black image.

On the other hand, as shown in FIG. 6B, when an unmounted image is captured by the image input unit 1 in a state where light is irradiated from the lower surface of the substrate S using a transmitted light illuminating device, light is emitted from the hole H. The image at that portion is transmitted and appears as a white image.

As shown in FIG. 6C, when the flow soldering is performed in a state where the solder is not mounted, the image of the solder put in the hole H appears as a white image.

That is, when no component is actually mounted, the portion of the hole H of the substrate S appears as a white image. In the second embodiment, the transmitted light illumination device 11 is used as a teaching image in FIG. The image shown in FIG. This makes it possible to match the teaching image with the image when the component is not actually mounted, and it is possible to reliably determine whether the component is not mounted.

The component inspection processing in the mounted component inspection apparatus of the second embodiment is the same as that of the first embodiment shown in the flowcharts of FIGS. 2 and 3, but is performed in step S103 in the teaching of FIG. As an image input, an image obtained by irradiating light from below the substrate S by the transmitted light illumination device 11 shown in FIG. As a result, it becomes possible to capture an unmounted image in which a hole H portion as shown in FIG. 6B becomes a white image.

FIG. 7 is a diagram showing a state of image comparison.
7A shows an image of a substrate that has not been mounted using transmitted light illumination, FIG. 7B shows an inspection image (normal), and FIG. 7C shows an inspection image (missing item).

That is, in the non-mounted image in the broken line in (a), the hole H appears as a white image due to the transmitted light illumination. Since the component B is mounted in the image shown in (b), the similarity of the image in the portion corresponding to the non-mounted image shown in (a) is small, and it can be seen that the component B is mounted.

On the other hand, in the image shown in (c), the similarity of the image in the portion corresponding to the unmounted image shown in (a) is extremely large, and it can be seen that the component B is not mounted. That is, in the unmounted image shown in (a), the image of the portion corresponding to the hole H is converted into a white image by the transmitted light illumination. This can be made almost the same as the state where the part is shown, and the state where the component B is not mounted can be accurately determined.

Next, a third embodiment will be described. FIG. 8 is a block diagram illustrating the third embodiment. That is,
The mounted component inspection apparatus according to the third embodiment includes a board moving mechanism 10, an image input unit 1, an unmounted board image storage unit 2, an image comparison unit 3, a determination unit 4, and a component mounting position database D.
B is the same as the mounted component inspection apparatus of the first embodiment and the second embodiment in the point that B is provided. However, a board hole mask generation unit 21 is provided between the image input unit 1 and the unmounted board image storage unit 2. Is different.

The board hole mask generation section 21 applies a predetermined mask to a hole portion of the board S in the unmounted image captured by the image input section 1. That is, a predetermined mask is applied to the hole portion of the substrate S in the unmounted image captured in the teaching, and when comparing the images in the actual inspection, the masked portion is excluded and compared, so that the component is actually The similarity with the image when it is not mounted is increased.

Next, a component inspection process in the mounted component inspection apparatus according to the third embodiment will be described with reference to the flowcharts of FIGS. 9 and 10 refer to FIG. 8. FIG. 9 shows the teaching, and FIG. 10 shows the actual inspection processing.

First, in step S301, the component B
Is set on the substrate moving mechanism 10. Next, in step S302, the board moving mechanism 10 is moved so that the position where the component B to be inspected is mounted enters the imaging area of the image input unit 1. At this time, the board moving mechanism 10
The position (X, Y coordinates) of the target part B is obtained from B,
The movement is performed based on this.

Next, in step S303, the image input unit 1 captures the unmounted image at that position.
The captured unmounted image is sent to the unmounted board image storage unit 2 and stored therein.

Thereafter, in step S304, a cutout range corresponding to the external shape of the component B in plan view is set. Then, in step S305, a predetermined mask is generated based on the data of the position and size of the hole in the substrate S, and is added to the portion corresponding to the hole of the unmounted image in the range cut out in step S304. This process is performed by the substrate hole mask generation unit 21.

In step S306, the image in the cutout range, that is, the unmounted image in which the mask is added to the hole portion is stored in the unmounted substrate image storage unit 2.

In step S307, it is determined whether or not the next part B is present.
S306 is repeated. If there is no next part B, the teaching ends. As a result, unmounted images corresponding to the positions of all the components B on the substrate S (images obtained by adding a mask to the positions of holes) can be stored.

After storing the unmounted image, an actual inspection is performed. First, as shown in step S401 of FIG.
The substrate to be inspected is set on the substrate moving mechanism 10. Next, in step S402, the board moving mechanism 10 is moved so that the component B to be inspected enters the imaging area of the image input unit 1. At this time, the board moving mechanism 10 obtains the position (X, Y coordinates) of the target component B from the component mounting position database DB, and moves based on this.

Next, in step S403, the image input unit 1 captures the mounted image of the component B at that position. Then, in step S404, the mounted image of the component B previously captured and the unmounted image corresponding to the position of the component B obtained from the non-mounted board image storage unit 2 (the mask is added at the position of the hole). Compare. By this comparison, the similarity is output from the image comparison unit 3.

In step S405, the determination unit 4 determines whether the similarity is equal to or less than a predetermined reference. When the similarity is equal to or less than a predetermined reference value (not similar), it means that the component B is correctly mounted. On the other hand, when the similarity is not equal to or less than the predetermined reference value (similar), the component B is not mounted. This means that B is not implemented or is not implemented correctly.

If it is determined in step S405 that the similarity is equal to or less than the predetermined reference, the determination is Yes, and the non-defective item is determined in step S406. On the other hand, step S405
In the case where the similarity is not less than the predetermined reference,
Then, a defective product determination in step S407 is performed.

In step S408, it is determined whether or not the next part B is present.
Step S405 is repeated. If there is no next part B, the inspection ends.

FIG. 11 is a diagram showing a state of image comparison.
(A) is an unmounted image with a mask added, (b) is an inspection image (normal), (c) is an inspection image (out of stock, before soldering),
(D) shows an inspection image (out of stock, after soldering).

In the third embodiment, a mask M is added to a portion of a hole H of an unmounted image as shown in FIG. For example, in the case of the inspection image shown in (b), since the part B is mounted on a portion corresponding to the unmounted image in (a),
The degree of similarity between the non-mounted image in (a) and the mounted image in (b) is very small.

On the other hand, in the case of the inspection image shown in FIG.
No component is mounted on the portion corresponding to the unmounted image in (a), and it substantially matches the unmounted image in (a). In particular, in the inspection image of (c), the position of the hole H is a black image since it is before soldering. In the non-mounted image of (a), a mask M similar to the state before soldering is added to the hole H portion, and the hole H portion is excluded from the comparison target, so that the component B is not mounted. The degree of similarity with the image can be greatly increased.

Also, in the case of the inspection image shown in (d), no part is mounted on a portion corresponding to the unmounted image in (a), and the image substantially matches the unmounted image in (a). In particular, in the inspection image of (d), the position of the hole H is a white image since it is after soldering. By adding a similar mask M in a state after soldering to the hole H in the unmounted image of FIG. 7A, the similarity with the image in the case where the component B is not mounted can be greatly improved. Can be enhanced.

As described above, in the third embodiment, the board B to be inspected is mounted because the board hole mask generation unit 21 additionally stores the predetermined mask at the hole portion in the unmounted image. The similarity with the image of the hole H projected when there is no component can be increased, and it can be reliably determined that the component B is not mounted.

Next, a fourth embodiment will be described. FIG.
FIG. 9 is a block diagram illustrating a fourth embodiment. That is, the mounted component inspection apparatus according to the fourth embodiment includes a board moving mechanism 10, an image input unit 1, an image comparison unit 3, a determination unit 4,
The system includes a component automatic cutout unit 5, a component image storage unit 6, and a component mounting position database DB.

Of the above configurations, the substrate moving mechanism 1
0, an image input unit 1, an image comparison unit 3, a determination unit 4, and a component mounting position database DB are the same as those in the first embodiment.

The component automatic cutout unit 5 calculates the component B in the image based on the difference between the unmounted image obtained from the board on which the component B is not mounted and the mounted image obtained from the board on which the component B is normally mounted. Is determined.

The component image storage unit 6 stores the component images in the cutout range determined by the automatic component cutout unit 5.

The mounted component inspection apparatus of the fourth embodiment is characterized in that a reference component image is automatically and accurately cut out and stored when teaching is mainly performed.

FIG. 13 is a block diagram for explaining the structure of the automatic component cutout unit 5. That is, the component automatic cutout unit 5 includes an unmounted image storage unit 51, a mounted image storage unit 52,
It comprises an image comparison unit 53 and a component image cutout unit 54.

The non-mounted image obtained by the image input unit 1 is stored in the non-mounted image storage unit 51, and the mounted image is stored in the mounted image storage unit 5.
2 is stored. The image comparison unit 53 obtains component size data from the component mounting position database DB (see FIG. 12), compares the unmounted image with the mounted image in units of the component size, and outputs cutout position data. further,
The part image cutout unit 54 that has obtained the cutout position data
Outputs the component image cut out based on the cutout position data to the component image storage unit 6.

In the fourth embodiment, the exact component position is obtained by sequentially comparing the unmounted image of the component and the mounted image in units of the component size, and the image of the component is obtained by cutting out the image at that position. It becomes possible to perform teaching with only extracted.

Next, a teaching process in the mounted component inspection apparatus of the fourth embodiment will be described with reference to the flowchart of FIG. Reference numerals not shown in FIG.
2 and FIG.

First, in step S501, the component B
Is set on the substrate moving mechanism 10. Next, in step S502, the board moving mechanism 10 is moved so that the position at which the component B to be inspected is mounted enters the imaging area of the image input unit 1. At this time, the board moving mechanism 10
The position (X, Y coordinates) of the target part B is obtained from B,
The movement is performed based on this.

Next, in step S 503, the image input unit 1 captures an unmounted image at that position.
The captured unmounted image is stored in the unmounted image storage unit 51 of the component automatic cutout unit 5.

Next, in step S504, a non-defective substrate S on which the component B is correctly mounted is set on the substrate moving mechanism 10. Then, in step S505, the board moving mechanism 10 is moved so that the component B on the board S is
In the image pickup area.

Next, in step S506, the image input unit 1 captures the mounted image of the component B at that position. The captured mounting image is stored in the mounting image storage unit 52 of the automatic component cutout unit 5. With the processing so far, the unmounted image and the mounted image on which the component B is accurately mounted are captured.

Next, in step S507, the image comparison unit 53 of the automatic component extraction unit 5 acquires the size data of the component B from the component mounting position database DB. In the component mounting position database DB, a component size composed of the external shape in plan view of each component B is stored corresponding to the position of the component B. The part size data is acquired by the image comparison unit 53.

Thereafter, as shown in step S508,
The image comparison unit 53 compares the acquired images in units of the component size.
Do with. That is, the image comparison unit 53 compares the unmounted image corresponding to the position of the component B stored in the unmounted image storage unit 51 with the mounted image corresponding to the position of the component B stored in the mounted image storage unit 52. Then, the image comparison is sequentially performed on the obtained component size unit.

In step S509, a position where the difference is the largest is obtained from the results of the image comparison in parts size units compared by the image comparing unit 53. In step S510,
The component image in the component size at that position is cut out by the component image cutout unit 54, and the component image is stored in the component image storage unit 6.

Then, in step S511, it is determined whether or not the next part B is present.
Steps 501 to S510 are repeated. If there is no next part B, the inspection ends. This makes it possible to accurately cut out the position of the component B during teaching and obtain only the component image.

FIG. 15 is a view for explaining a state in which the cutout position is determined by the image comparison performed by the image comparison section 53. In FIG. 15, (a1), (b1) and (c1) indicate mounted images, and (a2), (b2) and (c2) indicate non-mounted images.

First, in (a1) and (a2), image comparison is started for each component size indicated by a broken line frame in the figure. In this example, image comparison is sequentially performed on each image from the upper left in the figure to the right. The images of each component size shown in (a1) and (a2) almost match, and the difference between them is very small.

Next, image comparison in parts size unit is (a
2) and (b2), there is a slight difference in the image of each component size in each image. As the region of the component size unit to be compared with the image reaches the position of the component B, the difference between the images gradually increases.

Then, the image comparison further proceeds, and (a)
When the process proceeds to 3) and (b3), a large difference occurs between the images in component size units in each image. That is,
In the image of the component size indicated by the broken line frame in (a3), there is an image of the component B of exactly the same size, whereas in the image of the component B, the image is (b3)
There is no component image in the component size image indicated by the broken line frame. The difference between the mounted image and the non-mounted image has the largest difference in the image of each component size in the broken-line frame shown in (a3) and (b3).

The image comparing section 53 calculates the position where the difference between the images of the respective component sizes is the largest, and notifies the component image cutout section 54 that the component B exists at this position.
Accordingly, the component image cutout unit 54 can cut out an image at the position of the component B accurately, and obtain an image of only the component B.

In the fourth embodiment, by cutting out such a component image, only the image of each component B can be cut out automatically and accurately, and the time required for teaching can be reduced.

Next, a fifth embodiment will be described. FIG.
FIG. 14 is a block diagram illustrating a fifth embodiment. That is, the mounted component inspection apparatus according to the fifth embodiment includes a board moving mechanism 10, an image input unit 1, an image comparison unit 3, a determination unit 4,
In addition to the configuration of the fourth embodiment including the component automatic cutout unit 5, the component image storage unit 6, and the component mounting position database DB, a perspective image input unit 1a is provided.

The perspective image input section 1a captures an image of the component B from an oblique direction while the image input section 1 captures the image from almost directly above the target component B. For this reason, the component automatic cutout unit 5 includes an image of the component B captured by the image input unit 1 (hereinafter, referred to as a “normal image”) and an image of the component B captured by the perspective image input unit 1a (hereinafter, referred to as “normal image”).
This is called a "perspective image". ) And the part B in the image based on
Is calculated.

FIG. 17 is a block diagram for explaining the configuration of the automatic component cutout unit 5. As shown in FIG. That is, the component automatic cutout unit 5 includes a normal image storage unit 55, a perspective image storage unit 56, an image comparison unit 53, and a component image cutout unit 54.

The normal image of the part B captured by the image input unit 1 is stored in the normal image storage unit 55,
The perspective image of the part B captured in a is stored in the perspective image storage unit 56.
Is stored. Further, the image comparing section 53 obtains component size data from the component mounting position database DB (see FIG. 12), compares the normal image with the perspective image in units of the component size, and outputs cutout position data. Further, the component image cutout unit 54 that has obtained the cutout position data,
The component image cut out based on the cutout position data is output to the component image storage unit 6.

FIG. 18 is a diagram showing the positional relationship between the image input unit and the oblique image input unit. The image input unit 1 is disposed almost directly above the component B to be teaching mounted on the board S, and captures an image from almost directly above the component B, that is, a normal image via the lens L1 by the imaging unit. The perspective image input unit 1a is disposed obliquely above the component B to be taught, and captures an image from obliquely above the component B, that is, a perspective image, via the lens L2 by the imaging unit.

In the fifth embodiment, an accurate component position is obtained by comparing a normal image and a perspective image of a component B in units of the component size, and an image of the component is obtained by cutting out the image at that position. Can be performed.

Next, a teaching process in the mounted component inspection apparatus according to the fifth embodiment will be described with reference to the flowchart of FIG. The reference numerals not shown in FIG.
6 and FIG.

First, in step S601, the component B
A non-defective substrate S on which is accurately mounted is transferred to the substrate moving mechanism 1.
Set to 0. Next, in step S602, the board moving mechanism 10 is moved so that the position at which the component B to be inspected is mounted enters the imaging area of the image input unit 1. At this time, the board moving mechanism 10 obtains the position (X, Y coordinates) of the target component B from the component mounting position database DB, and moves based on this.

Next, in step S603, the normal image of the component B at that position is captured by the image input unit 1, and the captured normal image is stored in the normal mounted image storage unit 55 of the automatic component cutout unit 5.

Next, in step S604, a perspective image of the same part B is captured by the perspective image input section 1a, and the captured perspective image is stored in the perspective image storage section 56 of the automatic component cutout section 5. By the processing up to this point, the normal image and the perspective image of the correctly mounted component B are captured.

Next, in step S605, the image comparison unit 53 of the component automatic cutout unit 5 acquires the size data of the component B from the component mounting position database DB, and compares the images in units of component size.

That is, the image comparison unit 53 obtains the part size unit obtained from the normal image of the part B stored in the normal image storage unit 55 and the perspective image of the part B stored in the perspective image storage unit 56. Are sequentially compared.

In step S606, the position where the difference is the largest is obtained from the results of the image comparison in parts size units compared by the image comparison unit 53. In step S607,
The position is corrected by the offset amount estimated from the height of the part B, the part image in the part size at the corrected position is cut out by the part image cutout unit 54, and the part image is stored in the part image storage unit 6. Remember.

Then, in step S608, it is determined whether or not the next part B is present.
Steps 601 to S607 are repeated. If there is no next part B, the inspection ends. This makes it possible to accurately cut out the position of the component B during teaching and obtain only the component image.

FIG. 20 is a diagram for explaining a state in which the cutout position is determined by the image comparison performed by the image comparison section 53. 20, (a1), (b1), and (c1) show normal images, and (a2), (b2), and (c2) show perspective images.

First, in (a1) and (a2), image comparison is started for each component size indicated by a broken-line frame in the figure. In this example, image comparison is sequentially performed on each image from the upper left in the figure to the right. The images of each component size shown in (a1) and (a2) almost match, and the difference between them is very small.

Next, the image comparison for each component size is (b)
When the image proceeds to the positions shown in (1) and (b2), there is a slight difference between the images in the component size unit in each image. As the area of the part size unit to be compared with the image reaches the position of the part B, the difference between the images gradually increases.

Then, the image comparison further proceeds, and (c)
When the process proceeds to 1) and (c2), a large difference occurs in the image of each component size in each image. The difference between the normal image and the perspective image is the largest in the image of each component size in the broken line frame shown in (c1) and (c2).

The image comparing section 53 calculates a position where the difference between the images of the respective component sizes is the largest, applies a predetermined correction to this position, and transmits it to the component image cutout section 54. The position where the difference becomes maximum is an intermediate position between the position of the component in the normal image and the position of the component in the oblique image. Therefore, the position where the difference is maximum is corrected based on the height of the part B and the angle of the perspective image with respect to the vertical direction. Accordingly, the component image cutout unit 54 can cut out an image at the position of the component B accurately, and obtain an image of only the component B.

In the fifth embodiment, by using the normal image and the oblique image, it is possible to cut out the component image only by using the board S on which the component B is correctly mounted, and to automatically and accurately extract each component image. It is possible to perform a teaching process of cutting out only the image of the part B.

Next, a sixth embodiment will be described. FIG.
FIG. 14 is a block diagram illustrating a sixth embodiment. That is, the mounted component inspection apparatus according to the sixth embodiment includes a board moving mechanism 10, an image input unit 1, an image comparison unit 3, a determination unit 4,
In addition to the configuration of the fourth embodiment including the component automatic cutout unit 5, the component image storage unit 6, and the component mounting position database DB, a cross image projection unit 1b is provided.

The cross image projection section 1b projects a cross image (one in which only the cross portion is irradiated with light) onto the component B on the substrate S and its peripheral area. The image input unit 1 captures the cross image and transfers it to the automatic component cutout unit 5, and the automatic component cutout unit 5 uses the component B based on the cross image.
Is calculated.

FIG. 22 is a block diagram for explaining the structure of the automatic component cutout section 5. As shown in FIG. That is, the component automatic cutout unit 5 includes the cross image analysis unit 57 and the component image cutout unit 54.

The cross image analyzing unit 57 obtains the cross image captured by the image input unit 1, calculates the position of the component B from the distortion of the portion of the cross image that hits the component B, and uses it as cutout position data. Output to the component image cutout unit 54. The part image cutout unit 54 that has obtained the cutout position data
Outputs the component image cut out based on the cutout position data to the component image storage unit 6.

FIG. 23 is a view for explaining the projected state of the cross image. That is, the cross image projection unit 1b is arranged next to the image input unit 1, and projects a cross image from an obliquely upper part of the component B mounted on the substrate S from the issuing unit via the lens L3. The image input unit 1 captures the cross image projected from the cross image projection unit 1b onto the component B and the surrounding area by the imaging unit via the lens L1.

In the sixth embodiment, the component automatic cutout unit 5 automatically calculates the position of the component B based on the cross image taken in by the image input unit 1, and cuts out the image at that position, so that only the image of the component is obtained. Can be performed.

Next, a teaching process in the mounted component inspection apparatus of the sixth embodiment will be described with reference to the flowchart of FIG. The reference numerals not shown in FIG.
1 and FIG. 22.

First, in step S701, the part B
A non-defective substrate S on which is accurately mounted is transferred to the substrate moving mechanism 1.
Set to 0. Next, in step S702, the board moving mechanism 10 is moved so that the position at which the component B to be inspected is mounted enters the imaging area of the image input unit 1. At this time, the board moving mechanism 10 obtains the position (X, Y coordinates) of the target component B from the component mounting position database DB, and moves based on this.

Next, in step S703, the image input unit 1 captures the cross image projected from the cross image projection unit 1b onto the part B and the surrounding area. Then, in step S704, the captured cross image is analyzed by the cross image analysis unit 57 of the automatic component cutout unit 5, and the position of the component B is calculated from the distortion of the cross.

Next, in step S705, the image input unit 1 captures a normal image (normal image) of the component B.
This normal image is sent to the component image cutout unit 54. After that, in step S706, a part image of a normal image is cut out using a region corresponding to the position of the part B obtained by the cross image analysis unit 57 in the previous processing as a cutout position, and stored in the part image storage unit 6.

Then, in a step S707, it is determined whether or not the next part B is present.
Steps 701 to S706 are repeated. If there is no next part B, the inspection ends. This makes it possible to accurately cut out the position of the component B during teaching and obtain only the component image.

FIG. 25 is a diagram for explaining the determination of the cutout position based on the cross image performed by the cross image analysis unit 57. FIG.
3A shows a normal image, FIG. 3B shows a cross image, and FIG. 3C shows a cross image distortion detection.

That is, when a cross image is projected on the component B correctly mounted as shown in (a) and its surrounding area, the state shown in (b) is obtained. The part B has a predetermined height, and as shown in FIG.
By projecting the cross image from obliquely above from above, the part of the cross image that hits the part B is shifted from the part that hits the peripheral area.

As shown in (c), the cross image analyzer 57
Reads the signal values along the vertical and horizontal lines of the cross image projected on the peripheral area of the component B (see the broken line in the figure), and calculates the location (distortion position) where a large change has occurred.

There are various methods for calculating the distortion position. For example, a change point of a read value of a vertical and horizontal cross image is found, or a signal value is added along the vertical and horizontal directions of the image. You may make it find based on the part in which the addition value becomes 0 among the added values.

Then, by calculating the distortion position of the cross image in the vertical and horizontal directions, the cutout position in the image of the component B is determined. The cross image analysis unit 57 obtains the cutout position data by such calculation, and outputs the data to the component image cutout unit 54. As a result, the component image cutout unit 54 cuts out the image at the position of the component B accurately from the normal image,
Only images can be obtained.

In the sixth embodiment, by using the cross image and the normal image, it is possible to cut out the component image only by using the board S on which the component B is accurately mounted. It is possible to perform a teaching process of cutting out only the image of the part B.

[0139]

As described above, the mounted component inspection apparatus of the present invention has the following effects. That is, according to the present invention, it is possible to accurately determine whether the component mounting state is good or not, based on the component unmounted image that is not affected by the component mounting state. In addition, the comparison between the component-unmounted image and the component-mounted image or the comparison between the first component-mounted image and the second component-mounted image, each having a different capture angle, is sequentially performed in units of areas corresponding to the external shape of the component in plan view. Accordingly, the position of the component can be accurately calculated from the difference. Further, the position of the component can be accurately calculated based on the distortion of the cross image projected on the component and its surrounding area.

As a result, an accurate image of the component can be automatically taken in during the teaching before the inspection of the mounted component, and it is possible to accurately judge the quality of the mounted state based on the image.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a first embodiment.

FIG. 2 is a processing flowchart (1).

FIG. 3 is a processing flowchart (No. 2).

FIG. 4 is a diagram showing a state of image comparison.

FIG. 5 is a block diagram illustrating a second embodiment.

FIG. 6 is a diagram illustrating a difference between unmounted images.

FIG. 7 is a diagram showing a state of image comparison.

FIG. 8 is a block diagram illustrating a third embodiment.

FIG. 9 is a process flowchart (3).

FIG. 10 is a processing flowchart (No. 4).

FIG. 11 is a diagram showing a state of image comparison.

FIG. 12 is a block diagram illustrating a fourth embodiment.

FIG. 13 is a block diagram illustrating an automatic component cutout unit.

FIG. 14 is a processing flowchart (No. 5).

FIG. 15 is a diagram for explaining determination of a cutout position.

FIG. 16 is a block diagram illustrating a fifth embodiment.

FIG. 17 is a block diagram illustrating an automatic component cutout unit.

FIG. 18 is a diagram illustrating a positional relationship of an image input unit.

FIG. 19 is a processing flowchart (part 6).

FIG. 20 is a diagram illustrating how to determine a cutout position.

FIG. 21 is a block diagram illustrating a sixth embodiment.

FIG. 22 is a block diagram illustrating an automatic component cutout unit.

FIG. 23 is a diagram illustrating a projected state of a cross image.

FIG. 24 is a processing flowchart (No. 7).

FIG. 25 is a diagram for explaining determination of a cutout position.

FIG. 26 is a block diagram illustrating a conventional example.

FIG. 27 is a flowchart of a conventional teaching process.

FIG. 28 is a processing flowchart of a conventional inspection.

FIG. 29 is a diagram illustrating a state of image comparison.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Image input part 2 Unmounted board image storage part 3 Image comparison part 4 Judgment part 10 Board moving mechanism B Parts DB Parts mounting position database S Board

Claims (6)

[Claims] 1. An unmounted image storage means for storing an unmounted image corresponding to a mounting position of a component on a board before the component is mounted, and mounting the component on the board after the component is mounted. An image comparison unit that compares a component mounted image corresponding to the position with a component non-mounted image corresponding to the component stored in the non-mounted image storage unit; and A mounting component inspection apparatus comprising: a determination unit configured to determine whether a mounting state is good or bad. 2. The mounting component inspection apparatus according to claim 1, wherein the non-component mounted image comprises a transmitted light image of the board. 3. A method for generating a mask image for filling the hole based on a position of a hole in the board provided corresponding to a mounting position of the component, and a method for forming a mask image on the board in the non-component mounted image. 2. The mounting component inspection apparatus according to claim 1, further comprising a hole mask generation unit that adds the mask image at a corresponding position. 4. An unmounted image storage means for storing a component unmounted image corresponding to the position of the component on the board before the component is mounted, and storing the component unmounted image on the board after the component is mounted. Mounted image storage means for storing a corresponding component mounted image; and a plan view outline of the component, wherein the component non-mounted image stored in the non-mounted image storage means and the component mounted image stored in the mounted image storage means are stored. And a position determining unit for determining a position of the component based on a result of the comparison by the image comparing unit. 5. A first mounted image storage means for storing a first component mounted image captured from a predetermined angle of a component mounted on a board, and said first component mounted image captured at an angle different from the first component mounted image. A second mounted image storage unit for storing a second component mounted image of the component, and the first component mounted image stored in the first mounted image storage unit and stored in the second mounted image storage unit Image comparing means for sequentially comparing the obtained second component mounting image with the area unit corresponding to the external shape of the component in plan view; and position determination for determining the position of the component based on the result of the comparison by the image comparing means. Means for inspecting a mounted component. 6. A cross image projecting means for projecting a cross image on a peripheral region centered on a component mounted on a substrate, and a portion of the cross image projected on the component and a portion projected on the peripheral region. And a position determining means for determining the position of the component from the distortion of the image in the above.