JPH02141603A - Method for detecting width and difference in level - Google Patents

Abstract

PURPOSE:To perform accurate and stable detection by irradiating an object with light and calculating the positions of the side wall surfaces of a difference-in-level part on a change of quantity of light received from the sampling point group of the irradiated light and the calculating the width between the side wall surfaces from said positions. CONSTITUTION:When the waveform formed by the group of the sampling points 42 of irradiated light has a shape 61, the number of signals having a definite level or more are checked with respect to the quantity of light of the points 42 at first and, thereafter, the number of the points 42 of a difference-in-level part, that is, a dark part 62 are counted. Next, the x-coordinates XL, XR of both ends of the dark part 62 (gap part) are calculated with respect to the quantity of light of the points 42 to calculate the width W=XL-XR of said gap part. Subsequently, the left regression curve 72 of a lefthand stable region part (region reduced in irregularity) 71 wherein the points 42 can be trusted is calculated. In the same way, the right regression curve 74 of a righthand stable region part 73 is calculated. Next, the y-coordinates of the intersecting point 76 of the longitudinal direction straight line of the coordinates XL and the line 72 are calculated and the y-coordinates YR of the intersecting point 78 of the longitudinal direction straight line f the coordinates XR and the line 74 are calculated. Then, difference in level H=YR-YL is obtained.

Description

Detailed Description of the Invention [Industrial Application Field] The present invention uses a two-dimensional (X , Z) relates to a method for detecting the width and level difference between works using a position detection sensor.

[Conventional technology]

The appearance of a two-dimensional position detection sensor equipped with a two-dimensional optical position detector using the optical cutting method is as shown in FIG. 8th
In the figure, (1) is a light source using a laser or a light emitting diode, and the light from this light source (1) is passed through a condensing lens (2).
Mirror (3), scanning mirror (4)
A portion of the reflected light (7) (scattered light) that enters the object (6) through the mirror (5) and is received by the light receiving lens (9) provided on the lens holder (8). After that, the light is focused on a two-dimensional optical position detector (10). This optical position detector (10) is attached to the sensor head (13) via a mounting bracket (11), and an adjustment screw (12)
The position can be adjusted. The sensor head (13) houses each of the above components.

FIG. 9 is a block diagram showing the hardware for processing the signal from the two-dimensional optical position detector (10), in which the data normalized by the divider (27) is transferred to the A/D converter (
30) and is taken into the CPU bus as processing data. In the figure, (21) is a preamplifier that amplifies the signal from the optical position detector (lO), (22) is a bandpass filter (BPF) composed of an operational amplifier, etc.
23) is a half-wave rectifier circuit that rectifies the output signal of the band-pass filter (22), and (24) is a half-wave rectifier circuit (23).
It is a low-pass filter (LP) that converts the output signal of the current signal into a continuous analog signal of the current signal.
4) are provided for each axis X, Xa, Y, and Ya. Note that although x and Z are used in the measurement, in this case they are used as X and Y for calculation purposes. Here Xa.

Ya indicates the actual axis relative to the ideal X and Y axes.
(25) and (26) are a subtracter and an adder each composed of an operational amplifier, etc., (27) is a divider that receives the output from the subtracter (25) and the adder (26), and performs division; 28) is a sample and hold circuit (S/H) for sampling the analog signal resulting from the operation of the divider (27) and holding it for the time required for signal conversion. (29) is an analog switch (S/H) composed of field effect transistors, etc. A
SW) (30) is an A/D converter that outputs the converted digital signal to the CPU bus as processing data.

A sensor using a two-dimensional optical position detector using the optical cutting method can detect the distance M (Z) from the reference position and the width direction displacement (X). Based on this detection result, the step and width values can be calculated from the processed and normalized signal (i.e., the output signal from the A/D converter (30)) as shown in FIG. is determined from the intersection of data tangents. Regarding this conventional method, Figures 10 and 11 (a
) to (f). Figure 1θ is a flowchart of an algorithm showing signal processing for detecting width and level difference, and Figures 11(a) to (f) are output signals X and Y (described later) in the block diagram of Figure 9 for each axis. This shows the signal waveform taken, and the width and step difference are calculated in the same process using the segment method based on the differential data array.

That is, as shown in the figure, when the waveform on the surface of the workpiece has a shape (image waveform> (41)), first the light intensity at the sampling point (42) is checked (step 101), and then the point sequence is A data array is created from the set (step 1o2) (see Figure 11(a)).Next, based on the differential data array, the data group is segmented (43) (step 103) (see Figure 11(a)).
b)), from which horizontal and vertical representative segments (
44) (step 104) (Fig. 11(C)
)). Then, a regression line (45) is determined using the least squares method based on the data belonging to the representative segment (44) (step 105) (FIG. 11(d)). lastly,
Calculate the intersection point (46) of the two straight lines in the horizontal and vertical directions (Step 10B) (Fig. 11 (e)), and each coordinate (in the measurement, the output signals are x and z, but for calculation purposes, the output signal is , y. Note that x = (X-Xa) /
(X+Xa), y= (Y - Ya) / (Y+Y
From a)), the difference in the coordinate values of the intersection point (46) is calculated as the width W of the object in the X direction and the height difference H of the object in the Y direction (step 107) (FIG. 11(f)).

(Problems to be Solved by the Invention) The conventional method of irradiating an object with light and receiving reflected light from the object to detect the width and step difference is configured as described above, and the output signal x , y (Fig. 9) as shown in Fig. 12 for defining the width. Therefore, even if the same width is measured, the waveforms will have shapes like (41a), (41b), and (41c), so the width dimensions will be Wa, Wb, Wc
There was some variation. As shown in FIG. 13, the height difference H also varied as shown in Ha, )lb, and Hc.

Furthermore, the conventional method described above is good for detecting depressions, but when trying to detect something like a bottomless through hole, it is difficult to judge how much to trust the data, making it difficult to detect accurately. there were.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a width and step detection method that can accurately and stably detect widths and steps.

[Means to solve the problem]

The width and step detection method according to the present invention includes sampling of the irradiated light in the width and step detection method in which the width and step are detected by irradiating light onto an object and receiving reflected light from the object. A first step of determining the position of the side wall surface of the stepped portion by changing the amount of light received from the point group, and then determining the width between the side wall surfaces from this position, and determining a straight line in each stable region portion of the sampling points, respectively. Next, the method includes a second step of finding each intersection between each of these straight lines and the positions of the above-mentioned side wall surfaces, and then finding the level difference between these intersections.

(Function) In this invention, since changes in the amount of light received from the workpiece are detected, the position of the side wall surface of the stepped part can be detected, and the vertical line at this position can be determined without variation, thereby accurately determining the width. Furthermore, by finding the intersection of the vertical line and the regression line with no variation in the stable region of the sampling points, the step difference can also be found accurately.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. In this example, the width and step difference (for example, the width and step difference between works) are detected by checking the amount of light and determining the position at the bottom, not from image data (normalized data) but as a pre-normalization step. Detection is performed by counting the number of pieces of data, and step detection is performed using image data. In this method, a straight line is determined by the least squares method as in the prior art, but the step is determined from the intersection of the vertical line indicating the width and the straight line.

Figure 1 is a flowchart of the width detection algorithm, Figure 2
Figures (a) and (b) are diagrams showing signal waveforms similarly to Figure 11. If the waveform formed by the group of sampling points (42) of the irradiated light has the shape (61), first the sampling points After checking the number of lights whose signal is above a certain level with respect to the light intensity of (42) (step 201),
Counting the number of sampling points (42) that are step portions, that is, dark (to DAR) portions (62) (step 20)
2). Next, the sampling point (42) has a dark light intensity (signal level) (DARK) part (62) (for example, a gap part). ) (Figure 2(a)),
Next, the width of this gap (W=XLXR) is determined (step 204) (FIG. 2(b)). However, steps 201 to 204 are the sampling points of the irradiated light (
42) Determine the positions XL and XR of the side wall surfaces (83) and (64) of the stepped portion based on changes in the amount of light received from the group,
Next, from these positions XL and XR, the side wall surface (63),
(64) constitutes the first step (205) of determining the width W between the lines.

Figure 3 is a flowchart of the step detection algorithm, Figures 4 (a) to (e) are Figures 11 (a) to (f).
This is a diagram showing the signal waveform similarly to the sampling point (42).
If the waveform formed by the group has the shape (61), first find the left regression line (72) of the left stable region (stable, low-variance region) (71) where the sampling point (42) is reliable. (Step 301) (Figure 4(a)
)), then determine the X coordinates of both ends (for example, the left end is X and the right end is characterized by X coordinates (step 302) (FIG. 4(b)),
Next, find the right regression line (74) of the right stable region part (73) where the sampling point (42) is reliable (step 303) (Fig. 4(c)), and then find the vertical straight line ( 75) and the left regression line (72) (7
6) Calculate the Y coordinate YL of
The Y coordinate YR of is calculated (step 304) (FIG. 4(d)).

Next, stage M H (H=YR-YL) is determined (step 305) (FIG. 4(e)). Therefore, these steps 301 to 305 are performed on the straight line (72) at each stable region portion (71), (73) of the sampling point (42).
), (74), then this straight line (72),
Find the intersections (76) and (78) of (74) and the XR vertical lines (75) and (77), and find the intersections (76) and
(78) constitutes a second step (306) for determining the step H between the steps.

Fig. 5 shows the signal waveform of the output signals x and y on each axis, and also defines the width r, and the width (w
= XL-XR) can be detected. The value of this width W is determined by the shape of the unstable signal waveform (61a), (61b) or (
61C) is not affected.

Figure 6 is a diagram that defines a "step", where the positions xL and xR of both ends of the dark part (62) and the vertical lines (75) and (77) are found, and then regression is performed from the stable region parts (71) and (73). Find straight lines (72) and (74) on the left and right, and find the intersection of both lines (76).

The level difference H is determined from the Y coordinates YL and YR of (78).

By the way, in the above embodiment, as shown in FIG.
R) We will explain the case where the positions of both ends of the part (62) are detected, and the dark part (62) is defined as the sample point group where the signal level is the lowest (the valley part as seen in the waveform). On the other hand, the same effect can be obtained by setting a threshold value C for the signal level and detecting the sampling point at the edge when the signal level is above (or below) a certain level, as shown in FIG.

In the above embodiment, the case where the width of the concave portion is determined is shown, but the present invention can also be applied to the case where the width of the convex portion is determined. Further, the level difference may be zero.

(Effects of the Invention) As described above, according to the present invention, widths, steps, and their respective positions can be detected stably and accurately.

[Brief explanation of the drawing]

FIG. 1 is a flowchart for width detection showing an embodiment of the present invention, and FIGS. 2(a) and 2(b) correspond to FIG. 1.
An explanatory diagram showing the signal waveform, FIG. 3 is a flowchart for step detection, and FIGS. 4(a) to (e) correspond to FIG. 3. An explanatory diagram showing the signal waveform, FIGS. FIG. 7 is an explanatory diagram showing another embodiment of the present invention, FIG. 8 is an external view of a two-dimensional position detection sensor,
FIG. 9 is a block diagram, FIG. 10 is a flowchart showing a conventional width and step detection method, and FIGS. 11(a) to (
f) Corresponds to FIG. 10 and is an explanatory diagram showing the signal waveform, and FIGS. 12 and 13 are explanatory diagrams defining the conventional width and step, respectively. (6)...Target, (7)...Reflected light, (42)...Sampling point, (83), (84)...Step side wall surface, (71)
, (73)... Stable region part, (72), (
74)...straight line, (78), (78)...intersection, (205)...first process, (306)...second process, W...width, H...step , XL+XR...Position of the side wall surface of the stepped portion. Note that the same reference numerals in each figure indicate the same or corresponding parts.

Claims (1)

[Claims] In a width and step detection method that detects the width and step by irradiating light onto an object and receiving reflected light from the object, the step is detected by changing the amount of light received from the sampling point group of the irradiated light. a first step of determining the position of the side wall surface, and then determining the width between the side wall surfaces from this position;
The method is characterized by including a second step of determining a straight line in each stable region of the sampling point, then determining each intersection of each straight line and the position of the both side wall surfaces, and then determining a step difference between the intersection points. Width and step detection method.