JPS62145491A - Measuring method for number of holes - Google Patents

Abstract

PURPOSE:To measure the number of holes in a real time by storing a specific uneven pattern consisting of binary picture elements of the holes formed on a work and comparing a bright and dark pattern obtained by moving one by one picture element by a visual means and a memory pattern. CONSTITUTION:Light passing through the work 3 from a light source 1 is made incident to a line sensor 2 and the output thereof is inputted to a comparator 6. Every time when the sensor 2 is moved to the positions of the respective lines of the work 3, it scans respective columns by a signal from a driving circuit 5. From the comparator 6, the bright and dark pattern of binary signals of '1', '0' corresponding to the formation of the work is outputted. This bright and dark pattern is compared with the first and the second reference patterns indicating specific uneven corners of the holes stored in a counting circuit consisting of a memory controller 7, memories 8, 9, an up down gate 10, and an up down counter 11. The counter 11 outputs the difference between the numbers of the bright and dark patterns respectively coinciding with the first and the second reference patterns to a CPU4. The CPU4 displays the number obtained by adding '1' to the number of the inputs on a display device 12 as the number of the holes of the work.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a hole number counting method for counting the number of holes formed in a workpiece.

[Conventional technology]

Generally, in a manufacturing process in which holes are punched or drilled into a workpiece, defective products may be produced due to drill breakage or the like.

In the past, products were visually inspected by humans using sampling inspection to check for drilling errors, number of holes, and good/defective products.

However, if there are many holes, it may not be possible to inspect all the holes. Additionally, since the system cannot be stopped immediately during sampling inspection, a large number of defective products may be produced.

Therefore, the workpiece is captured by visual means such as a TV camera,
By performing image processing to recognize the hole pattern from the image data, or by detecting the contour line from the image data and determining that a hole is provided there if the contour line is closed. One possible method is to measure the number.

[Problem that the invention seeks to solve]

However, in the above method, one screen's worth of image data must be loaded into memory and then subjected to the above image processing, which requires a large memory capacity, the software for image processing is difficult, and real-time processing is not possible. be.

The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to provide a method for measuring the number of holes in a workpiece lζ that can measure the number of holes guarded in a work lζ in real time using visual means.

[Means for solving problems]

In order to solve such conventional problems, the present invention provides a bright and dark pattern consisting of at least 2 x 2 pixels, each pixel of which is binarized, indicating an area smaller than the hole formed in the workpiece. Among the light and dark patterns, a light and dark pattern indicating a specific concave corner of the hole is used as a first reference pattern, and a light and dark pattern indicating a convex corner protruding in the opposite direction to the concave corner is stored in advance as a second reference pattern. and capture a predetermined field of view in which the work exists using visual means,
Based on the output of the visual means, the bright and dark patterns are sequentially extracted while moving pixel by pixel within the predetermined field of view, and the sequentially extracted bright and dark patterns are compared with the first and second reference patterns, and the bright and dark patterns are compared with the first and second reference patterns. The number of holes formed in the workpiece is measured based on the number of bright and dark patterns that match the first reference pattern and the number of bright and dark patterns that match the second reference pattern.

〔Example〕

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

First, the principle of the present invention will be described.

Figure 1(a) shows the partial shape of the workpiece as a binary image of a 7x12 pixel matrix, where the bright parts (white parts) in columns 5 to 8 in rows 3 to 5 are the parts of the workpiece. showing the hole;
A dark area (shaded area) other than the bright area indicates the surface of the workpiece.

Here, as shown in Table 1, "1" is a bright pixel, and "0" is a bright pixel.
'' as a dark pixel, a first reference pattern and a second reference pattern each consisting of 2×2 pixels are set. This first reference pattern matches the light-dark pattern of the 4th and 5th columns in the 5.6th row of the binarized image, and specifies the lower left concave corner A related to these matrices.

Table 1 First Reference Pattern Second Reference Pattern However, in such a hole, the number of first reference patterns that indicate the lower left concave corner A is one.

In addition, in the hole shown in FIG. 2(b), the first reference pattern is in the 4th and 5th columns in the 4th and 5th rows related to the concave corner A1, and in the 5th and 6th columns in the 5th and 6th rows related to the concave corner A2. It exists in On the other hand, in such a case, a convex corner B is formed in the opposite direction to the concave corner. The bright and dark patterns in the 5th and 6th columns in the 4th and 5th rows related to this convex corner B match the second reference pattern shown in Table 1, and this second reference pattern specifies this convex corner B. . Therefore, by subtracting the number of second reference patterns (1) from the number of first reference patterns (2), 1 is obtained.

Similarly, in the hole shown in FIG. 1(C), the first reference pattern has concave corners A3, A4, As and A
6, and the second reference pattern is present in two adjacent columns in each two adjacent rows according to B1, B2 and B3.
exists in two adjacent columns in each adjacent two rows. Therefore, 1 is obtained by subtracting the number 3 of the second reference pattern from the number 4 of the first reference pattern.

In this way, no matter what shape the holes formed in the workpiece have, the difference between the number of first reference patterns and the number of second reference patterns in one hole becomes one. Therefore, when a plurality of holes are formed in the workpiece, since the difference between each hole is 1, the difference between all the holes is equal to the number of holes. Therefore, by finding the difference, the number of holes can be measured.

Also, the upper left concave corner C as shown in Fig. 2(a)
The difference in the number of occurrences between a first reference pattern that specifies the concave corner C and a second reference pattern that specifies the convex corner E in the opposite direction to the concave corner C may be determined.゛The first reference pattern and the second reference pattern in this case are shown in Table 2.

Table 2 First Reference Pattern Similarly to the second reference pattern, the first reference pattern in Table 3 identifies the upper right concave corner F as shown in Figure 2 Φ), and what is this concave corner F? The difference in the number of occurrences of each pattern may be calculated from a second reference pattern that specifies a convex corner G in the opposite direction.

Table 3 First reference pattern Second reference pattern Furthermore, the first reference pattern in Table 4 that specifies the lower right concave corner H as shown in Figure 2 (CJ), and what is this concave corner H? The difference in the number of occurrences of each pattern may be calculated from the second reference pattern that specifies the convex corner I in the opposite direction.

Table 4 First reference pattern Second reference pattern Next, in the binarized image showing the entire shape of the workpiece 31 as shown in FIG. 3(a), the outer circumference i of the workpiece 31 is There is one corner B.

Therefore, when the number of existing second reference patterns is calculated based on the entire binarized image Iζ of the workpiece 3I, one second reference pattern related to a convex corner on the outer periphery of the workpiece 31 is included in the number of existing second reference patterns. In such a case, when calculating the difference between the numbers of each of the first and second reference patterns, one extra second reference pattern related to the convex corner on the outer periphery is subtracted, and the difference is less than the number of holes. is also one smaller, so the number of holes must be found by adding 1 to the difference.

Further, in the workpiece 32 having the shape shown in FIG. 3(b), the difference in the number of the first and second reference patterns that specify the concave corner A and the convex corner B is 1. Therefore, regardless of the shape of the workpiece, the difference between the numbers of the first and second reference patterns on the outer periphery of the workpiece is always 1. Therefore, when calculating the number of holes based on the binarized image of the entire workpiece, it is sufficient to add 1 to the difference in the number of holes in the first and second reference patterns for holes located on the right side of the outer circumference of the workpiece. .

Next, an embodiment of an apparatus to which the method of the present invention is applied will be described with reference to FIG.

A workpiece 3 is placed between a light source 1 and a line sensor 2 in this device, so that only light from the light source 1 that is not blocked by the workpiece 3 enters the line sensor 2. In addition, as shown in FIG. 5, the workpiece 3 has a hole 3.
It has a shape having a, 3b and 3C, and the line sensor 2 is sequentially moved to each position from 1st to 14th row.

The CPU 4 applies a trigger signal to the sensor drive circuit 5 each time the line sensor 2 is moved to the position of each row, and when the sensor drive circuit 5 receives this trigger signal, it sends a clock signal for driving the line sensor to the 2-in sensor 2. Add. As a result, the line sensor 2 scans from point P in the 1st column to point P in the 18th column in synchronization with the clock signal number, and sequentially sends a detection output proportional to the amount of light received at each point to the comparator 6. Add.

The comparator 6 is a binary circuit, and if the detection output is above a preset threshold value, it outputs a signal ``1'' indicating that light has been received at the above point; If so, a signal indicating that no light was received at the above point.''
Therefore, as shown in FIG. 6, each time the line sensor 2 scans at each row position, a binarized signal corresponding to the shape of the workpiece is output from the comparator 6).

The binarized signal is applied to the memory controller 7.

The memory controller 27 controls the first and second memories 8 and 9.
are selected alternately, and the first memory 8 is selected when the above-mentioned scanning is performed on odd-numbered rows, and the second memory 9 is selected when scanning is performed on even-numbered rows. Then, the above-mentioned binary signal is added to the selected memory, and this binary
The value signal is also applied to the up/down gate 10.

The first and second memories 8 and 9 respectively store binarized signals for one column of adjacent odd-numbered rows and even-numbered rows, as shown in FIG. When the 2-in sensor 2 moves and a new binarized signal is added, the memory is erased and the binarized signal is stored.

On the other hand, the up/down gate 10 is connected to the memory controller 7
When n binarized signals based on the current scan are input from the first or second memory, n lines based on the previous scan are input from the first or second memory.
- Input one row of binarized signals. FIG. 8 shows the configuration of this up-down gate 10, in which the binary signals of the n rows are sequentially input to the terminal D of the D flip-flop 81 starting from the first column.
At the same time, the binary signals of the n-1 rows are sequentially applied to the terminal D3 starting from the first column power 1. Then, the binary signals of n rows output from terminal Q1 are output from terminal D! The n-1 row of binarized signals output from the terminal Q3 are applied to the terminal D4. Therefore, terminals Q1, Q2 . Q
From s and Q4, (n-1, m) and (n-
Binarized signals at positions 1, m-1), (''+m) and (n, m-1) are output.

The binarized signals output from the terminals Qs + Qz + Qs and Q4 are sent to AND circuits 82 and 8, respectively.
Added to 3. The AND circuit 82 connects each of the above terminals Q1 #
When each of the binary signals related to Q2.93g and Q4 matches the first reference pattern in Table 1 described above, the AND condition is satisfied and an up count signal 11'' is output,
If the pattern matches the second reference pattern in Table 1, the AND condition is satisfied and the AND circuit 83 outputs a down count signal "1".

Each of the above count signals 11'' is an up/down counter 11
added to. The up/down counter 11 counts up the up count signal ``1'' and counts down the down count signal ``1''.

By the way, as can be seen from the binarization process in FIG. 7, the same pattern as the first reference pattern is the concave corner A, M-A.
Exists in y. Moreover, the same pattern as the second reference pattern exists in the convex corners B1 to B. By the way,
The number of patterns that are the same as the first reference pattern is 7, and the number of patterns that are the same as the second reference pattern is 5.
It is. Therefore, the up/down counter 11 calculates the difference 2 between the numbers of each line when the line sensor 2 finishes scanning each row.
is counted, and a signal indicating this difference of 2 is added to the CPU 4. As is clear from FIG. 7, the value obtained by subtracting the number of patterns identical to the second reference pattern from the number of patterns identical to the first reference pattern in each hole 3a, 3b, and 3C is 1. be. On the other hand, the above value at the outer circumference of the workpiece is -1.

The CPU 4 adds 1 to the difference 2 to obtain the difference 3 between the concave corner and the convex corner, which is related only to the holes 3a, 3b, and 3C in the workpiece 3. Then, a signal indicating the determined number of holes, 3, is applied to the display 12, and the number of holes, 3, is digitally displayed on the display 12.

Now, once the number of holes in one workpiece has been measured, another workpiece is placed between the light source 1 and the line sensor 2, and the above measurements are performed one after another. However, if the number of holes in the workpiece is other than 3, the CPU 4 determines that the workpiece is a defective product, sends a stop signal to the workpiece manufacturing system 13 to stop this system 13, and defective products continue to be manufactured. prevent this from happening.

FIG. 9 shows another embodiment of the device to which the method of the invention is applied, in which the memory controller 7, first and second memories 8 and 9 in the device shown in FIG. It is configured by inserting a register 14. In the apparatus of this embodiment, one row of binary signals is stored in the shift register 14, so that n-1 rows of binary signals can be obtained.

In the above embodiment, the line sensor 2 is moved to the position of each row, but the position of the workpiece 3 may of course be moved.

By the way, for example, a signal @l due to noise may be mixed into two adjacent columns in two adjacent rows that should all be formed with the signal ""0", and a pattern identical to the first reference pattern may be formed by mistake. Or a signal ``'' due to noise in two adjacent columns in two adjacent rows that should all be formed by the signal ``'1''.
If a pattern identical to the second reference pattern is erroneously formed due to the mixing of 0'', the existence of these erroneous patterns will also be counted.

Therefore, if each reference pattern in four adjacent columns in two adjacent rows containing the first and second reference patterns as shown in table #I5 is preset, these reference patterns are formed by one noise signal. Without this, the number of holes can be measured based on the difference in the number of these reference patterns.

Table 5 Reference pattern including the first reference data pattern Reference pattern including the second reference data pattern By the way, in this embodiment, the first and second reference data patterns shown in any of Tables 1 to 4 above are Based on the reference pattern of
The number of holes is measured by adding 1 to the difference between the numbers of the first and second reference patterns in the binarized image of the workpiece. Therefore, the number of existence a of all the first reference patterns shown in Tables 1 to 4 and the number of existence of all the second reference patterns are determined, and the operation shown in the following equation q) is calculated. You can also measure the number of holes C by going there.

−b C= −+ 1 ・−(1) Note that the above 2×
The size of the second pixel must be smaller than the size of the hole to be searched.

Further, the present invention is not limited to the line sensor, but can also use other visual means such as a television camera that captures two-dimensional images.

〔Effect of the invention〕

As explained above, according to the present invention, the hole detection process is made hardware, so the number of holes can be measured in real time, and the apparatus becomes simpler. Furthermore, this measurement method can measure the number of holes in any workpiece by the same process, regardless of the installation direction, shape, or hole shape of the workpiece.

[Brief explanation of drawings]

1, 2, and 3 are diagrams showing the binarization of a workpiece using bright and dark patterns, and FIG. 4 is a block diagram showing an embodiment of a device to which the method according to the present invention is applied. Figure 5 is the fourth
Figure 6 is a diagram showing the workpiece and the scanning position of the line sensor that captures the workpiece in the embodiment shown in the figure.
A timing chart showing the binarized signal of each scanning line in the embodiment shown in the figure, and FIG. 7 shows the binarized image in the embodiment shown in FIG. 4 with ``''l'' and @O''. Figure, 8th
This figure is a block diagram showing the configuration of the up-down gate in the embodiment shown in FIG. 4, and FIG. 9 is a block diagram showing another embodiment to which the method of the present invention is applied. l...Light source, 2...Line sensor, 3...Work, 4...-CPU, 5...Sensor drive circuit, 6...
Comparator, 7... Memory controller, 8... First memory, 9... Second memory, 10... Up/down gate, 11... Up/down counter, 12...
・Display device, 13... Manufacturing system, 14... Shift register applicant's agent Takahisa Kimura No. 114 Figure 2

Claims (2)

[Claims] (1) Indicates an area smaller than the hole formed in the workpiece,
A light and dark pattern consisting of at least 2 x 2 pixels binarized for each pixel, and among the light and dark patterns, a light and dark pattern indicating a specific concave corner of the hole is used as a first reference pattern, and the light and dark pattern is opposite to the concave corner. A bright and dark pattern showing a convex corner protruding in the direction of is stored in advance as a second reference pattern, a predetermined visual field in which the workpiece is present is captured by visual means, and the bright and dark pattern is determined based on the output of this visual means. Sequentially extracting while moving pixel by pixel within the predetermined visual field, comparing the sequentially extracted light and dark patterns with the first and second reference patterns, and determining the number of light and dark patterns that match the first reference pattern. A method for measuring the number of holes, characterized in that the number of bright and dark patterns that match the second reference pattern is subtracted from the second reference pattern, and 1 is added to the subtracted value to determine the number of holes formed in the workpiece. (2) Indicates an area smaller than the hole formed in the workpiece,
A light and dark pattern consisting of at least 2 x 2 pixels binarized for each pixel, a light and dark pattern indicating a concave corner among the light and dark patterns as a first reference pattern,
A bright and dark pattern indicating a convex corner is stored in advance as a second reference pattern, a predetermined visual field in which the workpiece is present is captured by a visual means, and the bright and dark pattern is determined based on the output of this visual means within the predetermined visual field. The sequentially extracted light and dark patterns are compared with the first and second reference patterns, and the first and second reference patterns are sequentially extracted while moving one pixel at a time.
A first number a of light and dark patterns that match the reference pattern of A method for measuring the number of holes, characterized in that the number c of holes formed in a workpiece is determined by the following formula: C=(a-b)/4+1.