JPH0367567B2 - - Google Patents

Description

[Detailed Description of the Invention] [Summary] The present invention provides a method for binarizing height data at a brightness peak of a light section image in a mounted component inspection apparatus that inspects the mounting state of a mounted component using a light section method. When extracting a part shape, the binarization is performed based on whether the height data is within a predetermined range and the brightness data at the peak is above an appropriate level corresponding to the type of part. This makes it possible to extract the correct part shape without being affected by shadows, regardless of the type of part.

[Industrial application field]

The present invention relates to a mounted component inspection device that automatically inspects the mounting state of electronic components (particularly chip portions) mounted on a printed board using an optical cutting method.

In recent years, surface-mounted components (chip components) have come into widespread use in order to miniaturize electronic devices. It is predicted that the use of chip components will continue to advance and the number of chips will increase rapidly in the future. In the manufacturing process of printed circuit boards using chip components, mounting is performed using automatic machines. However, automation of the visual inspection of the mounted state has been delayed and currently relies on human visual inspection. Automation of visual inspection is essential to improve the reliability of printed boards using chip parts. Against this background, there has been a strong desire to automate the visual inspection of chip component mounting.

[Conventional technology]

Conventional mounted component inspection equipment using optical cutting method is
A slit-shaped light beam (hereinafter referred to as slit beam) is irradiated onto the printed board from directly above, and a line sensor detects the light cut line formed there from diagonally above.The resulting light cut image is used to identify the part. Inspection is performed by extracting the shape and comparing it with a reference pattern.

One way to extract the shape of the part is to first find the peak of brightness in the height direction of the light-cut image, obtain the height data at this peak, and then obtain the height data (above a certain value). In some cases, the shape of the part is extracted by binarizing it as a "component" when it is present (or within a certain range) and as a "background" at other times.

Another method is to obtain the height data and brightness data at the peak value in the same way as above, and check if the brightness data is above a fixed specific value and the height data is above a certain value (or within a certain range). ), there are things that are binarized as "components" and other times as "background".

[Problem that the invention seeks to solve]

Generally, in a mounted component inspection apparatus using the optical cutting method, as shown in FIG.
The above slit beam l for line sensor LS
is in the shadow of component Q and cannot be detected.

Therefore, when the former method described above is used to extract the part shape, when the peak value of brightness is determined and the height data is obtained as described above, in the shaded part, the However, due to the influence of diffused light, etc., the height of the peak will vary randomly, resulting in inaccurate heights being obtained. Therefore, if the height data is simply binarized based on a certain height, there is a drawback that only an inaccurate part shape can be obtained in the shaded areas.

In addition, devices that use the latter method to extract part shapes take into consideration the fact that the luminance in the shaded areas is lower than other parts, and when the luminance data is lower than a certain fixed level, , since that part is determined to be a shadow undetectable part, the above-mentioned drawback is considerably improved. However,
In general, parts are either black or white depending on their type, and their light reflectances vary greatly. Therefore, the optimal value of the brightness level (hereinafter referred to as "required brightness level") required to remove the shaded portion as an undetectable portion differs depending on the type of component. Therefore, if the required brightness level is fixed as described above, for example, if the required brightness level is too high for a black component, the component itself will be detected as missing, and for a white component, the required brightness level will be too high. If the required luminance level is too low, there is a problem in that even shadow areas are treated as parts.

SUMMARY OF THE INVENTION In view of the above-mentioned problems, an object of the present invention is to provide a mounted component inspection apparatus that can obtain a correct component shape regardless of the type of component.

[Means for solving problems]

As a means for extracting the shape of a part, the present invention detects the brightness peak in the height direction of a light cut image obtained by sequentially detecting a slit beam with a line sensor, and obtains the brightness data and height data. and the height data is determined by determining whether the brightness data is equal to or higher than the optimal required brightness level predetermined corresponding to the type of component, and the height data is within a predetermined range. The present invention is characterized by comprising a binarization means for binarizing.

[Effect]

As mentioned above, if the optimal required brightness level corresponding to the type of component is set in advance, the optimal required brightness level can be used when comparing with the brightness data obtained by the peak detection means. can. Therefore, regardless of the type of part, it is possible to extract the correct part shape without being affected by shadows.

〔Example〕

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

FIG. 1 is a configuration diagram showing an embodiment of the present invention.

In the figure, the present embodiment first includes a light irradiation means consisting of a semiconductor laser 1, a collimating lens 2, and a cylindrical lens 3. This light irradiation means converts laser light _l1 outputted from a semiconductor laser 1 into parallel light l2 with a collimating lens ₂ ,
Furthermore, the cylindrical lens 3 converts it into a slit-shaped light beam (slit beam) _L3 , and this slit beam _L3 is irradiated from directly above onto the printed board P placed on a stage 4 movable in the X and Y directions. do. The printed board P has various parts (particularly chip parts) Q mounted on the board P, and a light cutting line L is formed on the board R and the parts Q by irradiation with the slit beam _L3 . It is formed.

Furthermore, an imaging lens 5, a galvanometer mirror 6, and a line sensor (such as a CCD line sensor)
It is equipped with a light detection means consisting of 7. This light detection means first guides the diagonally upwardly reflected light l ₄ from the light cutting line L to the galvano mirror 6 via the imaging lens 5 . Since the galvano mirror 6 is finely rotated (vibrated) back and forth within a certain angle range, the reflected light _l4 guided to the galvano mirror 6 is swung up and down along with the vibration. A line sensor 7 sequentially detects this reflected light. Then, the line sensor 7 sequentially detects the reflected light _l4 in the height direction, and by swinging the galvanometer mirror 6 once in one direction, the entire image of the light cutting line L can be seen. Hereinafter, this image will be referred to as a "light-section image."

As shown in FIG.
The images are obtained as approximately slit-shaped images (m ₁ , m ₂ , m ₃ ) with deviations corresponding to the height of the object, and these become multi-gradation gradation images corresponding to the brightness of the object. In the figure, images m ₁ , m ₂ , and m ₃ correspond to a high component, a low component, and a board, respectively. Such a light-cut image can be expressed by an X address and a Z address, where the X address corresponds to the line direction of the line sensor 7 and the Z address corresponds to the height.

In order to properly obtain the above-mentioned light cut images sequentially over the entire surface of the printed board P, the stage 4, the galvano mirror 6, and the line sensor 7 are operated by the stage drive circuit 9 and the galvano mirror drive, respectively, based on instructions from the signal processing circuit 8. They are driven in synchronization with each other by a circuit 10 and a line sensor drive circuit 11.

Next, this embodiment includes a peak detection circuit (not shown) as one of the image processing systems for the optically sectioned image. In the light section image shown in FIG. 2, the highest brightness point (peak) in the height direction (Z direction) exists for each position in the X direction. The peak detection circuit detects the peak and obtains height data (Z address) and brightness data (for example, 8-bit gradation data) at this peak. The height data and brightness data at the peak obtained in this manner accurately correspond to the height and brightness of the detection target image, except for the shaded portions described above.

The peak detection circuit further creates a height image and a brightness image along the XY plane by determining the height data and brightness data for all the light cut images sequentially obtained on the entire surface of the printed board P. Examples of these images are shown in Figure 3a and
Shown in b. In the figure a, images m ₁₁ and m ₁₂ correspond to the parts, and in the Y direction there are images (shaded parts) corresponding to the shadow parts of the parts. In figure b, the image
m _11a and m _12a correspond to the parts body, and images m _11b and m _12b
corresponds to the component electrodes, and images m ₂₁ , m ₂₂ , m ₂₃ , and m ₂₄ correspond to the substrate electrodes. Similar to figure a, the images corresponding to the shaded parts of the component (shaded areas) are shown in the Y direction. ) has been completed. Note that both the height image and the luminance image have multiple gradations. To facilitate understanding of these images, changes in brightness and height obtained by sequentially detecting objects as shown in FIG. 4a in the Y direction are shown in FIG. 4b and c. Then, in the shaded part B, the brightness level becomes lower than other parts, as shown in figure b, and furthermore, as shown in figure c.
As is clear from the figure, the height appears randomly due to the influence of diffused light. Even in the shaded areas shown in Figures 3a and b,
This shows a similar change.

Therefore, next, a binarization circuit (not shown) for accurately extracting the part shape from the height image without being influenced by shadows will be explained. In this circuit, the values of each image, the brightness image and the height image, are the following two.
2 of the height image based on whether or not the two conditions are met.
Perform value conversion.

The first condition is that the value of the brightness image (pixel) is equal to or higher than the required brightness level. Note that this required brightness level is the brightness level necessary to distinguish and remove the shaded part from other parts, and is set to the optimal value according to the type of part, and is preset in the system memory ( stored in the memory 12) in FIG.
The parts are extracted according to their type. As a specific value, for example, as shown in FIG. 4B, the brightness level is set to be lower than the brightness level of the main body of the part (portion A) and higher than the brightness level of the shadow (part B). In this way, the required luminance level will generally be a small value for black parts and a large value for white parts.

The second condition is that the value of the height image (pixels) is
Must be within a predetermined range. This range is set, for example, as follows. First, the fifth
As shown in FIG. a, a region including a component Q to be inspected is set as a window W in the height image. Then, as shown in FIG. 5b, a histogram of heights within the window W is created. Pay attention to this histogram in order from the lowest height to the highest height, find the height where the frequency has a peak at a predetermined specific value a or more, and calculate this height h ₀
Let be the board height. Next, the value h _s is obtained by adding, for example, 1/2 value h ₁ of the general height of the component to the above board height h ₀
Let be the height slice level and this slice level
The predetermined range is above h _s . With this setting, the parts and the board can be accurately distinguished without being affected by the warpage of the board. Note that instead of the above value h ₁ , the permissible height range of the component may be added to the above board height h ₀ and the range within that range may be set as the above predetermined range.

The binarization circuit binarizes the height image as a "component" when both the first and second conditions are satisfied, and as a "substrate (background)" otherwise. As a result, for example, the shaded part (oblique part) of the height image shown in FIG. 3a is removed.
The correct part shape can be extracted without being affected by shadows. To explain this with reference to Fig. 4, the luminance shown in Fig. 4b is lower than the required luminance level in the shaded area, so the corresponding part in Fig. 4c is removed as the background, and A binarized pattern with an accurate part shape as shown in FIG. In other words, the required luminance level is too high and the main body of the component is chipped and removed, and the required luminance level is too low and the shaded part is not extracted as a component.

Finally, a judgment circuit (not shown) compares the binarized pattern obtained in the above manner with a standard normal pattern, and determines the mounting state of the component (for example, the component Determine whether there are any defects such as "missing parts", "misalignment", "defective height", etc.).

As described above, according to this embodiment, a correct part shape that is not affected by shadows can be obtained regardless of the type of part.

In the embodiment, the image processing described above is performed based on instructions from the CPU 13 shown in FIG.

〔Effect of the invention〕

According to the present invention, it is possible to obtain an accurate part shape that is not affected by shadows, regardless of the type of the part, and it is therefore possible to further improve inspection performance (defect detection rate).

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing one embodiment of the present invention, and FIG.
The figure shows an example of a light section image obtained in the same example, Figures 3a and 3b show examples of a height image and a brightness image, respectively, obtained in the same example, and Figures 4a to 4
d is a reference diagram for explaining the effect of the same embodiment, and FIGS. The figure shown in FIG. 6 is a diagram showing an undetectable part due to a shadow. 1... Semiconductor laser, 2... Collimating lens, 3
...Cylindrical lens, 6...Galvano mirror,
7...Line sensor.

Claims (1)

[Scope of Claims] 1. Light irradiation means 1, 2, 3 for irradiating a slit-shaped light beam onto a printed board on which components are mounted; light detection means 5, 6, 7 for sequentially detecting the light cutting line formed above in the height direction with a line sensor 7; peak detection means for detecting a brightness peak and obtaining brightness data and height data at the peak; binarization means for binarizing the height data depending on whether the height data is within a predetermined range; and normal data based on the binarization data obtained by the binarization means. and determining means for comparing the mounting state of the component and determining the mounting state of the component based on the comparison result. 2 The light irradiation means includes a semiconductor laser 1 that outputs a laser beam, a collimating lens 2 that converts the laser beam obtained by the semiconductor laser into parallel light, and a collimating lens that converts the parallel light obtained by the collimating lens into the slit shape. 2. The mounted component inspection apparatus according to claim 1, further comprising a cylindrical lens 3 for converting the light beam into a light beam. 3. The light detection means shakes the reflected light from the light cutting line in the height direction with a galvano mirror 6,
The mounted component inspection apparatus according to claim 1 or 2, wherein the reflected light is sequentially detected by the line sensor 7. 4. The binarization means binarizes it as a "component" when the luminance data is equal to or higher than the required luminance level and the height data is within the predetermined range, and otherwise binarizes it as a "background". A mounted component inspection apparatus according to any one of claims 1 to 3. 5. The mounted component inspection apparatus according to any one of claims 1 to 4, wherein the line sensor is a CCD line sensor.