JPS61243304A - Visual inspection system for mounted substrate - Google Patents

Abstract

PURPOSE:To improve an accuracy in detecting the shift of a position by providing a reference for detecting the shift of the mounting position of a part to be mounted to a surface from a substrate mounting design pattern itself. CONSTITUTION:A window 40 is set to a position which does not cover two portions in a substrate background. In this case, when the bounds of the window 40 is set so as to leave a given space relative to the two portion as shown in the accompanying drawing, the one window 40 does not include two background portions even when the positioning of the substrate involves some errors. When one and the other portions of the window 40 are set within a land 15 and the bounds of a substrate material 1, respectively, the threshold of conversion to binary notation required when the position of a chip part is recognized in the respective portions of the window may be changed in accordance with the brightness level of the background. Thus, by limiting the background in the window 40 to one kind, discrimination between the portion of this kind and the chip part can be facilitated.

Description

[Detailed description of the invention] (Technical field) The present invention relates to an electronic component mounted on a printed circuit board.

In particular, the present invention relates to improvements in visual inspection equipment for inspecting the mounting position accuracy of surface-mounted components such as square chip components and flat pack ICs.

(Prior art and its problems) The density of component mounting on printed circuit boards has recently become higher and higher, and the mounting format has changed from the conventional lead insertion method using leaded components to the use of chip components and flat pack ICs. There is a strong tendency to shift to the traditional imposition format. By the way, in general, an automatic lead component insertion machine can easily detect whether or not there is an insertion error when inserting a component, whereas an automatic lead component loading machine can detect a deviation in the mounting position immediately after the component is loaded.

Alternatively, it is difficult to detect that an abnormality has occurred in the mounting position. For this reason, after mounting a nine-sided component, an external appearance inspection of the mounting position is usually performed using visual inspection or the like, and the need for this is becoming stronger as boards become more densely packed.

As described above, visual inspection not only causes extreme fatigue to the operator, but also tends to result in frequent mistakes in one operation, and therefore efforts have been made to mechanize this inspection. One example is a video comparator-type chip component mounting board appearance inspection apparatus shown in FIG. This device uses two sets of illuminators 23 and TV cameras 21, one of which 21-a inputs the image of the non-defective board 1-a, and the other 21-b inputs the image of the board to be inspected 1-b. If there is a component mounting error on the test board, the difference in the image level for that part will become noticeable.

Visual inspection can be easily performed by displaying this on a TV monitor (not shown), or can be automated by outputting it to the outside via a signal processing circuit (not shown), etc. . In addition, if the field of view of the TV camera is narrow, the inspection can be continued while moving the board by a constant value using an xy table 30 or the like (However, this device had the following drawbacks.

(Depending on the type of 11-step component, it is difficult to detect a one-position shift because the visual contrast between the background board and the chip component cannot be obtained.

(2) The standard for mounting position deviation is a good board, and it is difficult to demand perfection from this.
There is a limit to the accuracy of positional deviation detection.

(3) It is necessary to accurately match the field of view size, relative position, and posture of the two TV cameras, making adjustment difficult.

(Objective) The present invention eliminates these drawbacks and provides a recognition method capable of accurately detecting the mounting position of a nine-sided component regardless of the state of the board.

At the same time, another object of the present invention is to improve the accuracy of positional deviation detection by providing a standard for detecting mounting positional deviation of nine-sided components from the board mounting design pattern itself.

(Example) Before explaining the basic principle of the present invention, let us first explain the third reason why it is difficult to recognize the position of surface-mounted parts in the above method depending on the conditions.
This will be explained using FIGS. 4 and 5. FIG. 3 shows a model of the imaging part, where 1 is a board, 10 is a surface-mounted component mounted on the board, 21 is a TV camera, and 22 is a model of the imaging part.
Reference numeral 23 indicates a magnifying lens for a camera, and 23 indicates an illuminator, such as a ring-shaped light. On the other hand, FIG. 4 shows a model of a square chip component, which is an example of a surface-mounted component.
) shows the top surface, (b) shows the side surface, and (c) shows the back surface. In the figure, 11 is a ceramic base material, and 12 is a main body.

Further, 13 is a metallic electrode portion. Such chip components are mounted on a substrate with either the top surface (a) or the back surface (c) facing upward depending on the conditions. Next, FIG. 5 shows the video signal level when such a chip component or the board itself is imaged by a TV camera in the state shown in FIG. 3. The horizontal axis in Figure 5 represents the aperture of the lens, or in other words, the brightness of the object as seen from the TV camera.
Ru. In FIG. 5, C indicates a chip component.

Since C3 is a ceramic portion and is white, the video signal level is extremely high. However, the main bodies of chip components come in various colors depending on the component, and there are large differences in level, for example, C2 for light brown and C8 for black. Furthermore, the level variation in the electrode portion is very large as shown by the shaded area C8. This is because the electrode portion has somewhat specular reflection characteristics and at the same time the surface is not flat. Furthermore, as shown in B, the background substrate also shows various levels depending on the location. For example, B is a dark green resisted base material part, and although the level is low, B2
The solder-plated land portion IB is the resisted copper pattern portion, and the copper-plated land portion B4 has a considerably high level.

In this way, since the chip component to be detected and the substrate portion as the background overlap in terms of the video level, it may be difficult to identify them using a black and white image. For example, if a chip component is mounted with the ceramic part facing upward, its presence or absence and misalignment can be easily recognized, whereas if a light brown chip component is mounted on a copper-plated land, both Since the difference in image level is small, it is difficult to recognize the 9-position shift. In this way, the background substrate has different brightness levels depending on its parts, and the chip components themselves have various brightness levels depending on their parts or types, making recognition difficult. These problems cannot be solved simply by replacing TV cameras with color cameras;
It is necessary to create some kind of difference between the image of the chip component itself and the image of the board so that the position of the chip component can be clearly recognized.

In order to make this possible, in the present invention, as shown in FIG.
Recognition is performed using the signal processing unit 26 that can convert the information into information S3 at an appropriate threshold. Among these, the function of the signal processing section 26 may be implemented by hardware, software, or a combination of both. The image memory 25 also divides the TV screen into two-dimensional meshes and determines the brightness level of each mesh (pixel) portion.

For example, as shown in FIG. 7, any configuration may be used as long as the video signal S is converted into gradation levels such as 0, 1.2 to 15 and stored according to its luminance level. good.

By the way, a model of the state in which the rectangular chip component shown in Fig. 4 is mounted on the land portion on the board is shown in Fig. 1 (
It will be like a). A point P within the xy coordinates on the board. Component] 0 is implemented correctly. land 1
5 and the electrode portion 13 of the component are properly soldered, and the occurrence of solder bridges between adjacent components and lands can be prevented. in this case.

Previously P. Using the four vertices of the chip part as a reference, a window 40 as shown by the broken line is set on the screen, and the position of the chip component within this window is detected to determine that the mounting position is normal. Detect whether or not (・ta.

However, in this method, as the board background in the window,
Two parts were included: the land 15 and the base material part 1, and it was necessary to distinguish these two parts from the chip component. Therefore, the first point of focus of the present invention is to change the window settings as shown in Figure 1 (b).
), (c) or (d),
The purpose is to set the window 4o at a position that does not straddle the two parts K in the background of the board.In this case, the range of the window should be set with a certain margin from the two parts as shown in the figure.

Even if there is some error in the positioning of the board, two background parts will not be included in one window (・.

Note that when one part of the window is set within the land 15 and the other part is set within the range of the base material 1 as shown in FIG. The binarization threshold for position recognition may be changed depending on the brightness level of the background.

By limiting the background within the window to one type in this manner, it becomes possible to provide an auxiliary means to further facilitate the identification of chip components from this portion.

This point is the second point of interest of the present invention, and its contents are as follows. In order to clearly identify the background and chip components within the window, the information S3 obtained through the signal processing unit 26 in FIG.
It would be nice if it had a shape like C) (・This is a model converted to two levels: ``1 level'' (white level) and ``0'' level (black level) for Remo S3. (a) Crawling/
Doe 5 is 1″, other shaded areas are 0″, and (
In b), only land 15 is O'' as indicated by diagonal lines, and the others are 1
", and (c) is the case where only the tip electrode is "1" and the others are "0". in this way.

A method for obtaining information for clearly identifying chip components and electrodes will be explained using FIG. 9. FIG. 9(a) shows the conversion relationship between the information SI and 82 in FIG. The symbols C, , B, etc. on the horizontal axis correspond to those in FIG.
2 is a level 1o signal.

The electrode part of the chip component is level 5 to level 12 as shown in C6.
is converted to a signal in the range of . Therefore, as shown by the solid line in FIG. 9(b), the conversion processing characteristics from signal S2 to signal S83 in FIG. 6 are set so that only the vicinity of level 10 is converted to the white level (1'' level).

As for S, the result shown in FIG. 8(a) is obtained, and if the transformation is completely reversed, the result shown in FIG. 8(b) is obtained.

Furthermore, if the conversion characteristics are set as shown by the broken line in FIG. 9(b), the state shown in FIG. 8(c) can be obtained.

In this way, by focusing on the video level of the board background in the area where the window is set and separating the information around that level from the video level corresponding to the chip component, it is possible to easily identify the two. Such processing is highly effective if the background within the window is of one type, and is advantageous when the distribution of image levels is wide, as in the case of the electrode portion of the chip component shown in FIG. Of course, on the other hand, if the distribution of image levels in the land area is wider.

The process described above may be reversed.

Above, the present invention has been described in detail using the example of FIG. 9 in which a chip component is mounted on a copper-plated land.
Even if the land is plated with solder, a similar method can be used by focusing on the image level of that part.

Also, considering the color tone of the board and the color tone of the parts used, processing the video signal into a convenient shape using optical filters, etc.

It is possible to extend the idea of the present invention by using a color camera and utilizing its color information.In general, the brightness of the video signal of the land portion is determined by There is a considerable difference depending on whether you irradiate from the direction or from the X direction, so consider the illumination direction and make two
It is important to determine the value threshold, but this can also be used to illuminate dark-colored chip components from the direction where the lands become brighter (for example, light-colored chip components It is also possible to make identification easier by performing reverse illumination.Furthermore, the idea of the present invention is that a video comparator type inspection device shown in Fig. 2 or a single TV camera can be used to It is clear that this method can also be used for devices that image a board and perform inspection by comparing the converted signal with pre-stored information. Also, the target parts are not limited to square chip parts.

(Effects) By using the present invention as described above, it is possible to easily recognize the mounting state of components mounted on a board under nine different conditions.
A general-purpose visual inspection device can be realized.

[Brief explanation of drawings]

FIG. 2 is a configuration diagram of a conventional visual inspection device, FIGS. 3 to 5 are diagrams explaining the limits of the inspection ability of the conventional device, FIG. 6 is an embodiment of the inspection method of the present invention, and FIG. FIG. 8 and FIG. 9 are diagrams for explaining the details of the recognition processing operation for visual inspection in the embodiment of the present invention. board, 10: surface-mounted parts, 15: 1yy)", 21: T
V right camera 23: illuminator 525: image memory. 26: Signal processing section, 40: Window. Figure 1 Figure 2 Size Figure 3 Figure 4 Figure 5 Procedural amendment (method) August 23, 1939

Claims (1)

[Claims] In an external appearance inspection method for a mounted board in which the position of a mounted component on a board is recognized through a plurality of windows provided in an input image of a TV camera, the set position of the window is as follows:
A method for inspecting the appearance of a mounted board, characterized in that the window is configured so that two or more portions of the board background that are different in terms of image level are not included within the range of the window.