JPH04278406A - Three-dimensional measuring method - Google Patents

Abstract

PURPOSE:To realize high-speed and highly accurate three-dimensional instrumentation applying the projecting method of a grating pattern through manipulation of stripes. CONSTITUTION:A data table 1 for obtaining the phase value of a grating pattern photographed onto the surface of an object from the lightness value of the surface and, a data table 2 for obtaining tone height of the object from a reference surface based on the phase value of the grating pattern photographed on the object surface and the measuring position in the axially same direction as that of the phase of the reference surface are calculated and formed beforehand. The height of tire object is obtained from the lightness value of the surface of the object by using the data tables 1 and 2.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention relates to height displacement of an object and
This invention relates to the field of measuring dimensional external shapes.

0002

2. Description of the Related Art A grating pattern projection method using fringe scanning has been considered as one of the methods for measuring the height displacement and three-dimensional shape of an object in a non-contact manner. This method projects a grid pattern (hereinafter referred to as grid pattern) with a sinusoidal illuminance distribution onto an object, and projects the brightness value of the height measurement point when performing stripe scanning in four stages with 1/4 period. Measurements are taken from an angle different from the direction, and the phase value of the grating pattern is calculated from each brightness value. Since the phase of the grating pattern is modulated according to the height displacement of the measurement point, by calculating the amount of this phase modulation and substituting it into the geometric relational expression of the optical device, the amount of height displacement of the object can be measured and the three-dimensional This is a method to find the shape.

0003

[Problems to be Solved by the Invention] Generally, when measuring the three-dimensional shape of an object, a configuration is used in which a grid pattern is projected onto the measurement surface from an oblique direction and the height of the object is measured from a direction perpendicular to the measurement surface. used. When a grating pattern is projected onto an object to be measured from an oblique direction, the length of the optical path from the principal point of the projection lens changes depending on the position of the measurement point, so the period of the grating pattern and focal position often change significantly. , it is not possible to project a uniform grid pattern. In addition, in the grid pattern projection method using fringe scanning, the illuminance distribution of the grid pattern is required to be in a sinusoidal state, but the method to achieve this state is to project a pulse waveform grid mask pattern. In many cases, the projected slit waveform is not necessarily a sine waveform, as the projected slit waveform is not necessarily a sine waveform. , the amount of change in phase is often not constant. In conventional technology, there is no solution to the problem that the grating pattern changes due to the influence of the optical system, and errors occur in calculating the amount of height displacement for the amount of phase change modulated by the height displacement of the object. There was a problem.

[0004] Also, generally when measuring the height of an object,
Three-dimensional shapes are often measured. In this case, since measurement is performed using a two-dimensional set of values of measurement points, there are a large number of measurement points, and the measurement time per point is required to be short. However, in the conventional technology, the phase amount is calculated using the arctan function from the brightness information of the grating pattern, and the height displacement amount is calculated from the phase amount using the geometric relational expression of the optical system, which requires a large amount of calculation and measurement. The problem was that it took a long time.

In view of the above, an object of the present invention is to provide a three-dimensional measurement method that can measure three-dimensional shapes at high speed and with high precision.

[0006]

[Means for Solving the Problems] The three-dimensional measurement method of the present invention projects a grid pattern in which the illuminance distribution of illumination is in a sine state from an oblique direction onto the object surface to be measured, and measures the brightness of the object surface by projecting the grid pattern onto the object surface to be measured. Height of object by grid pattern projection method using fringe scanning, measured from an angle different from the direction and
A three-dimensional measurement method for measuring dimensional shapes, which includes a data table that calculates the phase value of the projected grating pattern from brightness information of an object on which a grating pattern is projected and scanned, and the phase of the grating pattern projected onto a measurement reference plane. A data table is prepared in advance to calculate the phase in the reference plane and the amount of height displacement at the coaxial position from the value, and a grid pattern is projected during measurement, and the above two types are calculated from the brightness information of the object to be measured by stripe scanning. The feature is that the height displacement is determined using a data table.

[0007]

[Operation] In the three-dimensional measurement method of the present invention, for example, data table 1 is used to determine the phase value of the projected grating pattern from the brightness value of the grating pattern scanned by stripes, and the phase value of the grating pattern projected onto the measurement reference plane is calculated. Measure and calculate the data table 2 in advance to determine the amount of height displacement with respect to the phase value at the position in the coaxial direction, and at the time of measurement, use the data table 1 to calculate the phase amount from the brightness value of the grid pattern scanned by stripes. Then, the height displacement amount at the measurement point is calculated using the data table 2 from the obtained phase amount and the position of the measurement point.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an example of the configuration of a measuring instrument for carrying out the present invention. 1 is a measurement reference plane that serves as a reference for measurement; 2 is an imaging camera that images the measurement plane; 3 is a projection device that projects a lattice pattern and shifts it in 1/4 period increments; 4
5 is a projection control device that outputs a signal that shifts the grid pattern by 1/4 period; 5 is an image memory that stores four images taken by the imaging camera 2 one at a time when designated; 6
is a computer used for measurement, which gives instructions to the projection control device 4,
It is equipped with a memory that instructs the image memory to store images, reads the brightness value of the image, and stores a data table.

FIG. 2 illustrates the geometric relationship of the optical system in the configuration shown in FIG. X is the horizontal coordinate system of the measurement reference plane 1 which is coaxial with the phase of the grating pattern, O is the origin of the coordinate system X at the center position of the screen imaged by the imaging camera 2, and P is the projection lens of the projection device 3. The principal point, C is the principal point of the imaging lens of the imaging camera 2, x is an arbitrary measurement point, Z is the amount of height displacement of the object at the measurement point x, and x' is the lattice pattern projected on the measurement point x. The point position on the coordinate system X when projected onto the reference plane 1, Lpop is the distance from the measurement reference plane 1 to the projection lens principal point P, Lpo is the distance from the projection lens principal point P to the origin O, and Lco is Imaging lens principal point C
to the origin O, Lpc is the distance in the X coordinate direction from the projection lens principal point P to the imaging lens principal point C, a is the projection angle of the lattice pattern on the object surface at measurement point x, and β is the object at measurement point x This is the imaging angle when imaging the surface.

FIG. 3 is a diagram showing the relationship between the brightness and phase of the grating pattern projected onto the measurement reference plane 1 and the coordinate system X. Ф is the phase coordinate system, φx is the phase of the grating pattern at the measurement point x on the measurement reference plane 1, and φx' is the point x on the measurement reference plane 1
cyc is the periodic length of the sine waveform of the grating pattern in the coordinate system X direction, and I is the projected brightness of the grating pattern.

FIG. 4 is a diagram showing the relationship between the coordinate system of the image taken by the imaging camera 2 and the projected grid pattern. X
In the figure, Y is a horizontal coordinate system, which is a coordinate system in the same direction as the phase coordinate axis, and Y is a vertical coordinate system. In the figure, the X coordinate value is 0 to 511, there are 512 sample points, and the Y coordinate value is 0 to 511.
This is an image with 480 sample points and 481 sample points.

FIG. 5 is a diagram showing data table 1. In is the brightness value when the phase of each grating pattern is shifted by 1/4 period, and I3-I1 and I4-
The phase value φ corresponding to I0 is calculated in advance,
This is an example of a case where the quantization level of the image is 0 to 255.

FIG. 6 is a diagram of the base table, which is used when creating the data table 2, and is a table of the phase of the grating pattern on the X coordinate axis of the grating pattern projected onto the measurement reference plane 1.

FIG. 7 is a diagram showing the relationship between the phase φ at an arbitrary coordinate position x on the X coordinate axis of the base table.

xn is the X coordinate point of the base table closest to or at the same position in the direction larger than x of the X coordinate, xn-
1 is the X coordinate point of the base table one position before xn.

φn is the phase value of point xn of the base table, φn-1 is the phase value of point xn-1, and φ is φn, φn-
This is the phase value at x when the change in phase between 1 is a straight line.

FIG. 8 is a diagram showing data table 2. This is a table in which the height Z from the measurement reference plane 1 is calculated from all the coordinate points of the coordinate system X and the phase value of the lattice pattern at each coordinate point in the image coordinate system X coaxial with the phase change.

First, an example of a method for measuring the height displacement Z of an object at an arbitrary measurement point x in the geometric relationship diagram of FIG. 2 will be described below. From Figure 2, Lpc=(Lco-Z)×tanβ+(Lpo
p-Z)×tanα...Formula 1 tanβ=x/(
Lco-Z)
...Formula 2 ∴Lpc=
x+(Lpop-Z)×tanα
...Equation 3 Therefore, the measurement point x of the measurement object
The height Z from the measurement reference plane 1 at
...Equation 4 and tanα=(Lpc-x')/Lpop
...Equation 5
∴Z=Lpop-(Lpc-x)/((Lpc-
x')/Lpop)

...Equation 6
Therefore, by calculating x', the height Z of the object surface at the measurement point x can be calculated.

In FIG. 3, if a grating pattern projection method using fringe scanning is used, the phase φ of the projected grating pattern
is φ=arctan((I3-I1)/(I0-
I2)) ...can be calculated using equation 7. At this time, I0, I1, I2, and I3 are the brightness values of the image when the phase of the projected grating pattern is shifted by 0, 1/2π, π, and 3/2π, respectively.

Since φx' is the phase value projected onto the object plane, it can be calculated from the brightness value at point X' using equation 7.

By setting the optical system so that the origin of the phase Ф is the same as the point O, x'=(φx'/2π)×cyc

...Equation 8. By substituting the result of Equation 8 into Equation 6, the height Z of the object surface can be found. With the above steps, the height of an object at any measurement point can be measured.

Next, a measurement method using a data table, which is a feature of the present invention, will be explained using the above-mentioned geometric relational expression. First, create a data table. Data table 1 is a two-dimensional table for obtaining the result of Equation 7. The quantization level of the image memory 5 is 0 to 255, and the values of I3-I1 and I0-I2 are -255 to +255.
becomes. Create -25 in I3-I1, I0-I2 of formula 7
This is done by substituting values from 5 to +255, calculating and storing φ for each value.

Next, as a preliminary preparation, a reference object having the same plane as the measurement reference plane 1 is placed. A grating pattern is projected onto this reference object plane, and an image is captured by shifting the phase by 1/4 period. As shown in FIG. 4, the horizontal direction of the image is an X coordinate system, and the vertical direction is a Y coordinate system. Here, the X coordinate system is the same as the X coordinate system in FIG. The grating pattern is projected to have the same phase value in the Y coordinate direction. As a base table, the phase value of the grating pattern projected on the reference object plane is calculated for each coordinate point of the X coordinate system, from the image brightness value I0 to
Calculations are made using data table 1 from I3 to create the table shown in FIG. In this case, the Y coordinate can be anywhere,
Image brightness values not only at one point but at multiple points in the Y coordinate direction,
The reliability may be increased by using the phase value, for example, by averaging. Note that the range of the phase value is 0≦φ<2π.

Data table 2 is created by calculating the amount of height displacement when the phase of the grid pattern projected at each coordinate point of the X coordinate system changes stepwise depending on the object and becomes 0≦Δφ<2π. This is done by calculating and storing everything. To explain the calculation method using FIG. 7, at the coordinate point x of the The phase value of the base table is searched in the direction in which the phase increases from the point, and as shown in Fig. 7, φxn-1≦φ
Find the position of xn where x'≧φxn, and from xn-1 x
If the phase change leading to n is a straight line, the phase value is φx
'The x' point of the X coordinate system is calculated using a linear interpolation formula such as Equation 9. x'=(xn-xn-1)/(φxn-φxn
-1)×φx'+xn

...Equation 9 Here, the coordinate value of the x' point is calculated using a value finer than the coordinate pitch of the X coordinate system. By substituting the value of x' obtained here into Equation 6, the height displacement amount Z when the phase changes φx' from the measurement reference plane can be found.

The creation of data tables 1 and 2 is thus completed. Next, as an actual measurement method, measure point x in Figure 2.
We will explain how to measure the height Z of an object in .

At the measurement point x, each brightness value I0 to I3 of the grid pattern obtained by stripe scanning on the object plane is measured by taking an image. Calculate I3-I1, I0-I2, extract the phase value φx' on the object plane using these two values from data table 1, and set the fractional value of φx' to match the stepwise increase in phase in data table 2. Adjust. Next, the height Z of the object surface is extracted from the data table 2 using the measurement point x and the extracted phase φx'. In this case, the Y coordinate may be placed anywhere.

With the above steps, the height Z of the object at the measurement point x can be measured. If high measurement accuracy is required, use φx' without adjusting, and use x in equation 9 when referring to data table 2.
As in the method used to calculate ', each Z value may be extracted from φn-1 and φn where φn-1≦φx'<φn, and the accurate Z value may be calculated using equation 10, which is a linear interpolation formula. . Z=(zφn-zφn-1)/(φn-φn-
1)×φx'+zφn

...Formula 10
When measuring a three-dimensional shape, the measurement is not of a point but of a two-dimensional area, but even in this case, the above-mentioned measurement may be performed for each measurement point. Although equations 9 and 10 use a linear interpolation equation, if higher measurement accuracy is required, a curved interpolation equation such as a spline curve may be used.

In the above example, the direction in which the phase of the grating pattern changes is coaxial with the coordinate system X of the captured image, but if the direction in which the phase changes in the same direction as the coordinate system Y, then , the value related to X in the above example may be replaced with Y. Data table 1 is a two-dimensional table in the example, but if there is enough memory to store the table, I3
-I3, I2, I, not from the binary values of I1, I0-I1
A four-dimensional table configuration may be used in which the phase is determined from four values of 1 and I0. As with data table 1, the amount of stepwise change in phase in data table 2 is more accurate if there is sufficient memory, and in some cases it can be measured with sufficient accuracy even without using equation 10. .

In the example, the arrangement of the optical system is shown as a configuration in which the image is projected obliquely onto the measurement reference plane 1 and an image is taken from the vertical direction, but the arrangement is not limited to this example, and other configurations may be used. In that case,
The data table 1 and the base table remain unchanged, and the data table 2 can be created by changing equations 1 to 6 to equations corresponding to the configuration of each optical system.

[0030]

[Effects of the Invention] According to the present invention, high-speed and highly accurate height measurement and three-dimensional measurement are possible using simple calculations and a measurement method that corresponds to the projected grid pattern.

Even when the period of the grating pattern projected onto the object to be measured is not constant, highly accurate measurement is possible, and measurement can be performed in a short time without actually performing a large number of calculations.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing an example of the configuration of a measuring device for implementing the present invention.

FIG. 2 is a diagram showing the geometrical relationship of the optical system in the configuration shown in FIG. 1;

FIG. 3 is a diagram showing the relationship between the brightness and phase of the grating pattern projected onto the measurement reference plane 1, and the coordinate system X. FIG.

FIG. 4 is a diagram showing the relationship between the coordinate system of an image captured by the imaging camera 2 and a projected grid pattern.

FIG. 5 is a diagram showing data table 1;

FIG. 6 is a diagram of a base table.

[Figure 7] Arbitrary coordinate position X on the X coordinate axis of the base table
FIG. 3 is a diagram showing the relationship between phase φ in .

8 is a diagram showing data table 2. FIG.

Claims (1)

[Claims] 1. A grid pattern having a sinusoidal illuminance distribution is projected onto the object surface to be measured from an oblique direction, and the brightness of the object surface is measured from an angle different from the projection direction of the grid pattern. In a three-dimensional measurement method that measures the height and three-dimensional shape of an object using a grid pattern projection method using fringe scanning, a grid pattern is projected and the phase value of the projected grid pattern is determined from the brightness information of the object that has been scanned with stripes. Prepare in advance a data table and a data table that calculates the amount of height displacement at a position coaxial with the phase on the reference plane from the phase value of the grating pattern projected on the measurement reference plane, and when measuring, project the grating pattern, A three-dimensional measurement method characterized in that a height displacement is determined from brightness information of a scanned object to be measured using the above two types of data tables.