JPH0560635B2 - - Google Patents

Description

[Detailed Description of the Invention] [Summary] A pattern inspection device that calculates the slope of a brightness distribution curve of a pattern to be inspected along a predetermined inspection line, and calculates the number and number of peaks of a differential curve obtained from the calculated slope. The quality of the pattern to be inspected is determined based on the value of each peak.

[Industrial application field]

The present invention relates to a pattern inspection apparatus, and more particularly to an apparatus for inspecting the pattern of a two-color molded product.

Currently, parts such as key tops, dials, and knobs used in electronic devices are generally made by resin molding, but patterns such as letters and figures are displayed on their surfaces to identify their functions. ing. In recent years, a two-color molding method has been developed as a method for manufacturing resin molded products having display patterns, and is widely used.

In this two-color molding method, the display pattern part is molded in the first step, and the background part, that is, the main body part, except for the display pattern part is molded in the second step, whereby the display pattern is formed on the background. A resin molded product with a raised appearance is obtained.

By the way, when a product is manufactured using a two-color molding method, a defect called "fogging" may occur in the product, in which the resin of the display pattern part is covered with the resin of the background part. For this reason, it is necessary to inspect products manufactured using the two-color molding method to see if "fogging" has occurred.

[Conventional technology]

Conventionally, therefore, it has been determined whether or not "fogging" has occurred by converting the pattern to be inspected into a binary image and inspecting the line width of the display pattern portion in the obtained binary image.

[Problem that the invention seeks to solve]

By the way, for example, if the display pattern part and the background part are "black" and "white" respectively, and the degree of "fogging" is small, the "fogging"
However, in the conventional method above, depending on the binarization level used when converting the pattern to be inspected into a binary image, the "gray" part is determined to be "black" as a result. , as a result, a “gray” “fog”
is incorrectly determined to be part of the normal "black" display pattern. The binarization level needs to be set to a luminance level approximately midway between the respective luminance levels of the pattern display area and the background area in order to avoid the effects of noise in the analog video signal or changes in the external environment regarding lighting. Therefore, the conventional method has a problem in that the detection accuracy deteriorates when the degree of "fogging" is small.

The present invention was made in view of the above problems, and
It is an object of the present invention to provide a pattern inspection device with improved detection accuracy of "fogging".

[Means for solving problems]

As shown in FIG. 1, the pattern inspection apparatus of the present invention that achieves the above object includes an imaging means 1 for imaging a pattern to be inspected, and an A/D converter for converting an analog video signal outputted from the imaging means 1 into a digital signal. A converting means 2, a storage means 3 for storing multivalued data of the digital signal converted by the A/D converting means 2, and a pixel group corresponding to a predetermined inspection line set on the pattern to be inspected. a smoothing means 4 for smoothing the multi-value data of each constituent pixel stored in the storage means 3; and the inspection line obtained from the multi-value data smoothed by the smoothing means 4. A differentiating means 5 calculates the slope of the brightness distribution curve of the pattern to be inspected along the master pattern, and the number of peaks of the differential curve obtained from the slope calculated by the differentiating means 5 and the peak value of each peak are calculated from the master pattern. The present invention is characterized by comprising a determining means 6 that determines the quality of the pattern to be inspected by comparing it with a predetermined value of the number of peaks and a threshold value of the peak value of the peaks, respectively, which are set based on the following.

[Effect]

The brightness distribution curve of the pattern to be inspected along a predetermined inspection line, which is obtained through the imaging means 1, A/D conversion means 2, storage means 3, and smoothing means 4, has its slope calculated by the differentiating means 5. be done. Then, the determining means 6 calculates the number of peaks of the differential curve obtained from the slope of the brightness distribution curve calculated by the differentiating means 5 and the peak value of each peak, based on the number of peaks set based on the master pattern. By comparing the predetermined value of
Determine whether or not this has occurred.

〔Example〕

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

FIG. 2 is a block diagram showing the configuration of one embodiment of the present invention. In the figure, an imaging device 10 images a pattern to be inspected 20 and supplies an analog video signal to an A/D converter 11. A/D converter 11
converts the analog video signal into a digital signal and supplies it to the frame memory 12.
stores the multi-value data of the supplied digital signal. Note that the imaging device 10, A/D converter 11, and frame memory 12 correspond to the imaging means 1, A/D conversion means 2, and storage means 3 in FIG. 1, respectively.

As will be described in detail later, the smoothing circuit 13 converts the multivalued data stored in the frame memory 12 of each pixel constituting a pixel group corresponding to a predetermined inspection line set on the pattern to be inspected 20. The data is smoothed and supplied to the smoothing pattern memory 14, and the smoothing pattern memory 14 stores the smoothed multivalued data. Note that the smoothing circuit 13 and the smoothing pattern memory 14 correspond to the smoothing means 4 in FIG.

As will be described in detail later, the differentiation circuit 15 calculates the slope of the brightness distribution curve of the pattern to be inspected along the inspection line obtained from the multivalued data stored in the smoothing pattern memory 14, and calculates the calculated value, that is, the differential. supplying the value to the differential pattern memory 16;
The differential pattern memory 16 stores the supplied differential values. Note that the differentiating circuit 15 and the differentiating pattern memory 16 correspond to the differentiating means 5 in FIG.

The dictionary memory 17 stores the differential value of the brightness distribution curve along the inspection line in the master pattern,
A determination reference value based on the stored differential value is supplied to the fog determination circuit 18. Fogging determination circuit 1
8 is composed of, for example, a microcomputer, and stores the number of peaks and the peak value of each peak in the differential curve of the pattern to be inspected obtained from the differential values stored in the differential pattern memory 16 in the dictionary memory 1.
It is compared with the determination reference value supplied from 7 to determine whether or not "fogging" has occurred. In addition,
The dictionary memory 17 and the fogging determination circuit 18 correspond to the determining means 6 in FIG.

Next, the operation of the above configuration will be explained.

Hereinafter, as an example of the pattern to be inspected, the alphabetical letter "E" displayed on the upper surface of a two-color molded key top as shown in FIG. 3 will be explained. Note that the display pattern section and the background section are black and white, respectively.

First, the imaging device 10 images the display pattern "E" on the keytop. FIG. 4 shows an imaging device 10.
An image captured by is shown. This image is actually formed by the shading of square pixels with sides of 35 μm. An analog signal representing the brightness of each pixel is converted into a digital signal by an A/D converter 11, and the multivalued data is stored in a frame memory 12. In addition, the broken line in Figure 4
L ₀ , L ₁ , and L ₂ are inspection lines, and the inspection lines are set for each display pattern.

The smoothing circuit 13 smoothes multivalued data of pixels on the inspection line. The reason is that the raw data contains noise components from the illumination or imaging system, and when a brightness distribution curve on the inspection line _L0 is created using the raw data, the brightness distribution curve as shown in Figure 5. This is because fine irregularities occur due to noise components, and these irregularities may cause an erroneous determination in determining whether or not "fogging" has occurred. Therefore, the multivalued data of each pixel is smoothed, and the method of smoothing is as follows.
In this embodiment, the data value of the pixel is the arithmetic average value of the pixel to be smoothed and surrounding pixels in contact with the pixel. For example, the 6th
In the figure, if A to I are each pixel and E is the pixel to be smoothed, then the arithmetic mean value of the data values of each pixel of A to I is the smoothed data value of pixel E. . By performing this smoothing, the brightness distribution curve on the inspection line _L0 no longer includes fine irregularities, as shown in FIG.
In other words, the noise component has been removed.
Then, smoothed by the smoothing circuit 13,
The data value of each pixel on the inspection line is stored in the smoothing pattern memory 14.

The differentiation circuit 15 calculates the slope of the smoothed luminance distribution curve as shown in FIG. In this embodiment, as shown in FIG. 8, if E _-1 , E ₀ and E ₊₁ are each of three consecutive pixels on the inspection line, the pixel E ₊₁ The value obtained by subtracting that of pixel E _-1 from the smoothed data value of is used as the differential value of pixel E ₀ . and,
By plotting the absolute value of this differential value against the pixel position, a differential curve indicating the absolute value of the slope of the brightness distribution curve can be obtained. FIG. 9 shows a differential curve corresponding to the brightness distribution curve shown in FIG. 7.
Note that the differential pattern memory 16 includes the differential circuit 1.
The absolute value of the differential value of each pixel calculated in step 5 is stored.

The dictionary memory 17 calculates the number of peaks and the peak value of each peak in the differential curve from the differential value of the master pattern, and uses the reference value M _t of the number of peaks and the threshold value P _t of the peak value as determination reference values. The signal is supplied to the fogging determination circuit 18. Then, the fog determination circuit 18
is calculated by using the program described below to calculate the number of peaks and the peak value of each peak in the differential curve from the differential value of the pattern to be inspected, and compare them with the reference value M _t and threshold value P _t , respectively. fog”
Determine whether or not there is.

Next, using the flowchart shown in Figure 10,
A fog determination routine executed by the fog determination circuit 18 will be described.

FIG. 10 is a flowchart showing a fog determination routine for an inspection line consisting of k pixels. In this flowchart, the pixel position is represented by a count value i, and the count value i of the pixel position at the "start point" and "end point" of the inspection line is "1" and "k", respectively.

First, in step 101, count values "i", 10 "M" and "j" are set to "1" and "0" respectively.
and set to “0”.

Next, in step 102, it is determined whether the absolute value P _i of the differential value of the i-th pixel is greater than the absolute value P _i+1 of the differential value of the (i+1)-th pixel. At this time, assuming that the differential curve is as shown in FIG. 11 as an example, if the pixel is within the area, the determination result in step 102 is negative (NO).
, and proceed to step 103. In step 103, it is determined whether the count value i is equal to the set value k, but if the pixel is within the area (FIG. 11), the determination result is negative, and in step 104, the count value i is Increment by "1" and return to step 102. Then, as long as the determination result in step 102 is negative, the loop of steps 102→103→104→102 is repeated, and the count value i is incremented each time the loop is repeated.

On the other hand, if the determination result in step 102 is affirmative (YES), it is determined that the absolute value P _i of the differential value is the peak value of the mountain, and the count values M and j are each incremented by "1" in step 105. Then, in step 106, the value P _i is set as the peak value D _j . Therefore, the count value M indicates the number of peaks in the differential curve, and the peak value D _j indicates the peak value at the jth peak.

Next, the process proceeds to step 107, where it is determined whether or not the value P _i is smaller than the value P _i+1 . However, if the pixel is within the area shown in FIG. move on. step 108
In step 109, it is determined whether the count value i is equal to the set value k or not, but if the pixel is within the area, the determination result is negative, and the count value i is incremented by "1" in step 109 and the step is continued. Return to 107. Then, while the pixel is within the area, steps 107→108→109→107
This loop is repeated, and when the determination result in step 107 becomes positive, it is determined that the pixel has moved from one area to another, and the count value i is set to "1" in step 110.
The process returns to step 102.

On the other hand, when the pixel is within the area, as the loop of steps 107→108→109→107 is repeated, the determination result at step 108 finally becomes affirmative. Then, it is determined that the investigation of the values P ₁ to _{P k} of all k pixels on the inspection line has been completed, and the process proceeds to step 111 onwards to determine whether or not "fogging" has occurred based on the results of the above investigation. Determine whether Note that the control result in step 103 is affirmative only when there is no peak in the differential curve.

In step 111, it is determined whether the count value M is equal to the preset reference value _Mt , and if the determination result is negative, that is, the number of peaks in the pattern to be inspected and the number of peaks in the master pattern are different. If the number of peaks does not match, the process proceeds to step 116 and it is determined that there is "fogging".
On the other hand, if the control result in step 111 is affirmative, the process advances to step 112.

In step 112, it is determined whether or not the peak value D _j is greater than or equal to the preset threshold value P _t , and if the determination result is negative, that is, the peak value D _j is equal to or greater than the threshold value P t . If it is smaller than _Pt , the process proceeds to step 116, and it is determined that there is "fogging". On the other hand, if the determination result in step 112 is affirmative, the process proceeds to step 113, where it is determined whether the count value j is "1" or not. If the determination result is negative, the count value j is decremented by "1" in step 114 and the process returns to step 112, while if the result is affirmative, the peak values of all the mountains are equal to the threshold value P _t It is determined that the above is the case, and the process proceeds to step 115, where it is determined that there is no "fogging".

〔Effect of the invention〕

As explained above, according to the present invention, it is determined whether or not "fogging" has occurred using the slope of the luminance distribution curve that responds sharply to "fogging". The detection accuracy of "fogging" is improved.

[Brief explanation of drawings]

Fig. 1 is a block diagram of the principle of the present invention, Fig. 2 is a block diagram showing the configuration of an embodiment of the invention, Fig. 3 is a diagram showing the appearance of the key top, Fig. 4 is a diagram showing an image, and Fig. 5 is a diagram showing the appearance of the key top. The figure is a diagram showing the brightness distribution curve on the inspection line _L0 , and Figure 6 is a diagram explaining the smoothing method.
Figure 7 is a diagram showing the brightness distribution curve after smoothing, Figure 8
The figure is a diagram explaining the method of calculating the differential value, Figure 9 is a diagram showing the differential curve on the inspection line _L0 , Figure 10 is a flowchart showing the fogging determination routine, and Figure 11 is a diagram showing other differential curves. It is a figure which shows an example. 1... Imaging means, 2... A/D conversion means, 3...
...Storage means, 4...Smoothing means, 5...Differentiation means, 6...Judgment means.

Claims (1)

[Scope of Claims] 1: an imaging means 1 for imaging a pattern to be inspected; an A/D conversion means 2 for converting an analog video signal outputted from the imaging means 1 into a digital signal; a storage means 3 for storing the multivalued data of the digital signal thus converted; a smoothing means 4 for smoothing the multivalued data smoothed by the smoothing means 4; and calculating the slope of the brightness distribution curve of the pattern to be inspected along the inspection line obtained from the multivalued data smoothed by the smoothing means 4. Differentiating means 5
and the number of peaks of the differential curve obtained from the slope calculated by the differentiating means 5 and the peak value of each peak,
A pattern inspection device comprising: a determining means 6 that determines the quality of the pattern to be inspected by comparing it with a predetermined value of the number of peaks and a threshold value of the peak value of the peaks, which are set based on a master pattern.