JPH0410001B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention moves the inspected object by driving an X-Y table on which the inspected object is placed, and images the surface of the inspected object on a camera. , an effective edge detection method and object to be inspected in which this video signal is binarized, scanned in the X direction or Y direction with a scanning line, and the minimum and maximum edges of the object to be inspected are determined based on the scanning length data. The present invention relates to a method for measuring the effective width of an object to be inspected by calculating the width of the object.

[Prior art and its problems]

Conventionally, when recognizing the minimum or maximum edge of an object to be inspected, such as a magnetic head, a method has been adopted in which a cursor line or a measurement frame is brought into contact with the object to be inspected and arithmetic processing is performed.

Therefore, if the object to be inspected is tilted with respect to the cursor line or measurement frame, it may not be possible to recognize which part is the minimum or maximum, and if the object is forced to recognize the minimum or maximum edge, the object to be inspected is It is necessary to correct the inclination by rotating the cursor line by the inclination angle, which not only results in a mechanically complicated structure, but also the measurement method of moving the cursor line wastes camera operation time each time the cursor line is moved. , not efficient.

[Purpose of the invention]

The first object of the present invention is to enable easy and simple determination of the minimum and maximum effective edges even when the object to be inspected is inclined with respect to the cursor line or measurement frame. The purpose is to easily and simply measure the effective width of the object to be inspected even if the object is inclined with respect to the cursor line or measurement frame.

[Structure of the invention]

The configuration taken to achieve the above object is that in the first invention, a video signal of the object to be inspected in a coordinate system having two axes in the X and Y scanning directions is stored in a binarized memory, and the stored video signal is stored in a binary memory. A video signal is sequentially scanned in the X or Y direction with a scanning line of 1 pixel pitch, and the edge addressed to each pitch is recognized using the black and white inverted coordinate values obtained.
The purpose is to determine the minimum and maximum edges of the object to be inspected based on the difference between the coordinate values and the reference line coordinate values having a specified angle.

In the second invention, a video signal of the object to be inspected in a coordinate system having two axes in the X and Y scanning directions is stored in a binarized memory, and the stored video signal is Alternatively, the edge for each pitch is recognized using the black and white inverted coordinate values obtained by scanning sequentially in the Y direction, and the maximum edge of the object to be inspected is determined by the difference between the coordinate value and the reference line coordinate value with the specified angle. Or, detect the minimum effective edge and detect the other maximum or minimum effective edge necessary to measure the effective width by the same operation from the opposite direction, and use the calculation formula H = 1
Difference between both effective edges 1 x COS θ x gauge (G) [μm/
This is to calculate the effective width of the object to be inspected using pixels.

〔Example〕

Next, embodiments of the present invention will be described based on the drawings.

Figure 1 shows the edge measuring device, 1 and 2 are X-
A Y table, 3 is an object to be inspected mounted and supported on the tables 1 and 2, 4 is a camera that takes an image of the object to be inspected 3, 5 is an A/D converter that converts a video signal into a digital signal, and 6 is an A/D converter that converts a video signal into a digital signal. An image processing device performs arithmetic processing such as binarization on the video information that has passed through the A/D converter 5, 7 is a binarization memory, 8 is a central processing unit (CPU), 9 is a timing control, 10 is a cursor control, 11 12 is a RAM, 12 is a ROM that stores a system program, 13 is a keyboard for writing required setting data into the RAM 11, 15 is a synthesis circuit, 16 is an A/D converter, and 17 is a monitor. Note that the X-Y tables 1 and 2 may be horizontally rotatable as required.

The image processing device 6 binarizes the video signal at the threshold level, and has a measurement frame 14 in which the processing range (measurement target area) is set in the video screen. The data to be binarized of the image of the object 3 is written into the binarization memory 7.

The measurement frame 14 can change its frame range by moving each side thereof.

At the start of the edge detection sequence, the binarization memory 7 is scanned in a coordinate system having two axes in the X and Y scanning directions.

The scan is run from the reference point 0,0 of the measurement frame 14 that sets the processing range (measurement target area) binarized by the image processing device 6, and this scanning line A...
The edge is detected as a coordinate value that is converted into black and white to 1 or 1-0, but in the first invention, as described above, the scanning line A is a coordinate value that is converted into black and white to 0-1 or 1-0. The edge of each pixel pitch is recognized, and the maximum (Fig. 3) and minimum (Fig. 4) of the edge of the object to be inspected 3 is determined by the difference between the coordinate value and the reference line coordinate value with the specified angle. It is linked to the central processing unit 8 (CPU) so that the detection can be performed, and the detection of the minimum and maximum edges of the object 3 to be inspected will be explained using a flowchart. Note that the threshold level, reference line coordinate values with a specified angle, and other input data are input in advance.

As shown in FIG. 5: First, sequence processing (step 18) is performed to perform each of the following flows.

・Next, set whether to display the image on the monitor 17 as negative or positive (step 19), and set the threshold so that points exceeding the threshold are white and points below the threshold are black so that binary image processing can be performed. Set the level (step 20).

- Then, after the X-Y tables 1 and 2 move from the predetermined origin position in the X and Y directions as shown in the figure and stop, the stage T imaged by the camera 4 is placed in the measurement frame 14. , and sets the cursor line 21 (step 22).

- After the stage T is determined, the binarized data subjected to the image processing is written into the binarized memory 7 as shown in FIG. 8 (step 23).

- The reference point 0 of the measurement frame 14 is stored in the binarization memory 7,
A point corresponding to 0 is determined and made scannable (step 24).

- Next, set the angle reference line 25 parallel to the inclination of the inspected object 3 in the stage T (step 26).

- After setting the angle reference line 25, it is determined whether to measure the effective edge or width of the object to be inspected (step 27).

- When measuring the effective edge (maximum edge) of the object to be inspected, determine whether to measure the effective edge of the object 3 in the X direction or the Y direction (step 28), and also measure the effective edge in the left and right directions. Edge detection, vertical edge detection (step 29)
(Step 30) After determining whether the edges are the same or not, a desired edge is detected among them. Here, the edge detection in the horizontal direction and the edge detection in the vertical direction described above are performed in the horizontal direction as shown in FIG. That is, to explain based on the edge detection flow in the left and right directions: - First, it is determined whether to detect the maximum edge point or, for example, the minimum edge point (step 31), and when the maximum edge point is detected. Input and set the initial minimum value -999 as comparison data,
Initial maximum value when detecting edge minimum
Enter 999 (Step 32) (Step
33).

- Next, it is determined whether to detect an edge or a right edge (step 34), and if a left edge is to be detected, the left cursor 21a is aligned with the angle reference line 25, and if a right edge is to be detected. Then, the right cursor 21b is aligned with the angle reference line 25 to enable scanning (step 35) (step 36).

- Then, the white point counter is cleared (step 37), and the scanning line A is scanned from the angle reference line 25 in the X direction at a pitch of one pixel (step 38) to detect the white point (step 30). .

- At this time, continue scanning, scan further from the white point detection point, and if the white point count is larger than the noise value programmed in advance considering the presence of noise (step 40), The earliest white point detection point is set as the edge point (step 41), and if the white point count is smaller than the pre-programmed noise value by scanning further from the white point, the white point count is cleared (step 41).
42) and continue scanning, then detect and if the white point count (step 43) is greater than the programmed noise value as described above, the first detected white point is set as the edge point.

・Then, calculate the distance from the angle reference line 25 to the edge point (step 44), and judge whether the distance is longer than the comparison data (step 44).
45), and if it is long, coordinates are calculated as the maximum comparison data (step 46). Then, repeat this process pixel by pixel (step 47), scan until VEND, determine whether the edge detection point is the maximum edge measurement (step 48), and check whether that detection point is the maximum detection point (step 49). After that, that detection point is set as the maximum edge point (step 50).

・If stage T is present (step 51)
is converted into extended coordinates X, Y (step 52) as shown in FIG.
x′+x(bh×G [μm/pixel (pixel)]))
(Y=y'-y(bu×G [μm/pixel)]) is calculated by the CPU 8, and the calculated value is used as the edge position.

Next, to explain the second invention, this second invention allows the edge of each pixel pitch to be recognized by the coordinate values where the scanning line A is black and white inverted to O-1 or 1-O, and the coordinate values of the edges at both ends are recognized. The effective width H (Fig. 3) (Fig. 4) of the object to be inspected 3 is detected based on the difference between the edge and the angle reference line 25 parallel to the edge, and calculation processing can be performed via the CPU 8. It is.

In other words, this second invention has X as shown in FIG.
After determining whether to measure the effective width in the X direction or the Y direction (step 27), select whether to scan in the X direction or
It is determined whether the scanning is in the Y direction (step 53), and the maximum rightward edge detection (step 54), maximum leftward edge detection (step 55) is performed based on the edge detection flow shown in FIG. Maximum edge detection (step 56) Maximum downward edge detection (step 57) is performed.

・The effective width H between both edges is (1 left - right 1 + 1) x COS θ x gauge (G) in terms of width.
[μm/pixel] In vertical width (1
It is calculated using the formula (lower - upper 1+1) x COS θ x gauge (G) [μm/pixel (pixel)] (step 58) (step 59).

If the stage T is included, the coordinates are converted to extended coordinates (X, Y) (step 60).

Incidentally, the maximum edge detected in both the first and second inventions is confirmed through the confirmation flow shown in FIG. 7, and then processed based on the flow described above. To explain the confirmation flow: - First, set the angle reference line 25 so that it passes through the maximum edge point, scan the scanning line A for each pixel, and make sure that the intersection of the scanning line A and the angle reference line 25 is white (step 61), scan line A level pixel by pixel (step 62) VEND
(Step 63), all scanning lines A and
If the intersection with the angle reference line 25 is a white point, the measurement judgment is confirmed to be correct (step 64).

- By the way, when scanning line A pixel by pixel, if a black point is detected, it is determined whether the scanning direction is right or left (step 65), and when scanning from left to right, the angle reference is used. The line 21 is shifted by adding one pixel (step 66), and when scanning from right to left, the angle reference line 25 is minus one pixel and shifted (step 67), and in this state, the scanning line A is further moved. Scan to see if a white point is detected, and if so, VEND
All the pixels are scanned one pixel at a time until reaching (step 63), and if a white point is confirmed in scan A on all pixels, that point is determined to be normal (step 64).

- To confirm the detection of a left edge where the angle reference line 25 shifts from right to left, the right cursor 21b
If the angle reference line 21 exceeds the angle reference line 21, the measurement judgment is treated as an error (step 68). On the other hand, even in the case of right edge detection confirmation in which the angle reference line 25 is moved in the left-right direction, if the angle reference line 25 crosses the left cursor 21b, the measurement determination is treated as an error (step 68).

Incidentally, in the above description, both the first and second inventions were explained assuming that the object to be inspected is tilted.
The first and second inventions of the present invention can also be used when the slope, that is, θ is O.

In addition, the first and second inventions in such embodiments can eliminate noise and accurately detect the length when detecting the maximum or minimum edge, and also confirm the maximum detection of the edge, so mistakes can be avoided. This makes it possible to perform precise measurements.

〔Effect of the invention〕

In the first invention, even if the object to be inspected is inclined, effective edges such as minimum and maximum edges of the object to be inspected can be easily determined and measured.

In the second invention, even if the object to be inspected is inclined, the effective width of the object to be inspected can be easily and simply measured.

Thus, the intended purpose can be achieved.

[Brief explanation of drawings]

The drawings show an embodiment of the effective edge detection method and effective width measurement method of an object to be inspected according to the present invention, FIG. 1 is a block diagram of an edge detection device, and FIG. 2 shows a stage and an object to be inspected to explain the calculation process. Relationship diagram, 3rd
The figure is an explanatory diagram showing the maximum point and effective width of an edge, FIG. 4 is an explanatory diagram showing the minimum point and effective width of an edge,
FIG. 5 shows the effective edge detection method of the first invention and the second invention.
A program flow showing the effective width measuring method of the invention,
FIG. 6 is a horizontal edge detection flow, FIG. 7 is a maximum point confirmation flow, and FIG. 8 is a data processing diagram showing a state in which binarized data is written in the binarized memory. In the figure, 3...Object to be inspected, A...Scanning line, 1,
2...X-table, 14...measuring frame, 4...camera, 25...angle reference line.

Claims (1)

[Claims] 1. A video signal of an object to be inspected in a coordinate system having two axes in the X and Y scanning directions is stored in a binarized memory, and the stored video signal is converted into a scanning line of 1 pixel pitch. The edge for each pitch is recognized using the black-and-white inverted coordinate values obtained by scanning sequentially in the A method for detecting effective edges of an object to be inspected, characterized by determining the minimum and maximum of . 2. Store the video signal of the object to be inspected in a coordinate system with two axes in the X and Y scanning directions in a binarized memory, and scan the stored video signal in the X or Y direction with a scanning line of 1 pixel pitch. The edge for each pitch is recognized using the black-and-white inverted coordinate values obtained by sequential scanning, and the maximum or minimum effective edge of the edge of the object to be inspected is determined by the difference between the coordinate value and the reference line coordinate value with the specified angle. At the same time, detect the other maximum or minimum effective edge necessary to measure the effective width by the same operation from the opposite direction, and calculate the calculation formula: H = 1 Difference between both effective edges 1 × COS θ × Gauge (G) [ A method for measuring the effective width of an inspected object in which the effective width of the inspected object is calculated using [μm/pixel (pixel)].