JPH09178463A - Multidimensional coordinate measuring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a multidimensional coordinate measuring device which can simply be manipulated and can prevent an erroneous recognition of a shape caused by erroneous operation during measuring, and which can enhance the working efficiency. SOLUTION: A measuring device moves an object to be measured and a detecting probe, relative to each other, so as to measure a shape of the object to be measured. In this arrangement, the measuring device further comprises a data input part 11 for introducing coordinates and an edge normal vector of an edge part at each measuring point, a shape recognizing part 13 for selecting measuring items in accordance with a measuring point number, a shape computing part 12 for computing at least one of shapes included in the selected measuring items and for computing errors in computed shapes, and a normal vector computing means 13 for computing normal vectors at the measuring points. Further, the shape recognizing part 13 recognizes one of the shapes in accordance with at least one of a result of comparison between both normal vectors and a result of comparison between errors of the shapes. Thereby it is possible to eliminate the necessity of moving the probe in the normal direction of the object to be measured if the selected measuring items include a plurality of shapes.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multi-dimensional coordinate measuring machine such as a two-dimensional coordinate measuring machine or a three-dimensional coordinate measuring machine for measuring the shape of a measured object by moving the measured object and a detector relative to each other. Regarding

[0002]

2. Description of the Related Art Conventionally, as a multidimensional coordinate measuring machine of this type, for example, measuring items such as the geometrical shape of an object to be measured are
After selecting from "Points, straight lines, circles, rectangles, distance between lines, parallelism between lines, and intersections between lines," and after key input,
The object to be measured is measured at a plurality of measurement points. After the measurement is completed, by operating the calculation start execution key, the calculation unit of the coordinate measuring machine calculates the keyed measurement item based on the measurement value of each measurement point, and outputs the calculation result. Two-dimensional coordinate measuring machines are known.

Further, as a conventional multidimensional coordinate measuring machine, for example, a measured value is obtained by bringing a contact type probe into contact with each measuring point of an object to be measured and a measuring direction at each measuring point. A three-dimensional coordinate measuring machine that recognizes the shape of an object is known.

[0004]

However, in the former prior art, the measurement item of the measured object is "point, straight line,
It is necessary to select from “Circle, Square, Distance between lines and points, Parallelism between lines, and Intersection point between lines” and enter the key. Therefore, there is a problem that an unintended measurement item is input at the time of key input, the shape of the object to be measured may be erroneously recognized due to this input error, and the operation is troublesome.

In the latter prior art described above, in order to automatically recognize the shape of the object to be measured based on the measurement value and the measuring direction of each measuring point of the object to be measured, the contact type probe is set in a direction substantially normal to the contact type probe. Therefore, it was necessary to contact the object to be measured. Therefore, if the direction in which the contact-type probe is brought into contact with the object to be measured is wrong, the shape of the object to be measured may be erroneously automatically recognized. Therefore, when operating a joystick or other operating member to bring the contact-type probe into contact with the object to be measured, great care must be taken, and it is difficult to perform the measurement operation and whether the shape of the object to be measured is correctly recognized. Had to be checked for each measurement, and there was the problem that the efficiency of the measurement work was poor.

The present invention has been made in view of such circumstances, and its problems are multi-dimensional coordinates which are easy to operate, can prevent erroneous recognition of a shape due to an erroneous operation at the time of measurement, and improve work efficiency. It is to provide a measuring machine.

[0007]

In order to solve the above-mentioned problems, a multidimensional coordinate measuring machine according to the invention of claim 1 measures the shape of an object to be measured by relatively moving the object to be measured and a detector. In the multidimensional coordinate measuring machine, the data capturing means for capturing the measurement value of each measurement point of the object to be measured, the normal vector detection means for detecting the first normal vector at each measurement point, and the number of measurement points Shape recognition means for selecting a measurement item according to, and one or more shapes included in the measurement item selected by the shape recognition means, while calculating based on the measurement value and a previously input mathematical expression The shape recognition means for calculating the error of each of the calculated shapes, and the normal vector calculation means for calculating the second normal vector at each measurement point of each of the calculated shapes, Means before choosing If the measurement item includes a plurality of shapes,
An optimum shape is recognized from the plurality of shapes based on at least one of the comparison result of the both normal vectors and the comparison result of the error of each shape.

The measurement item selecting means selects a measurement item including one or more shapes in accordance with the number of measurement points. When the measurement item includes a plurality of shapes, the first and the first of each of the plurality of shapes are selected. The optimum shape is automatically recognized from among the plurality of shapes based on at least one of the comparison result of the normal vector 2 and the comparison result of the error of each shape.

Therefore, the operation of selecting the measurement item and inputting the key becomes unnecessary. This simplifies the measurement operation and prevents erroneous recognition of the shape due to an input error. Also,
Instead of recognizing the shape based on the measurement value and the measurement direction as in the conventional technique, the shape is determined based on at least one of the comparison result of the first and second normal vectors and the comparison result of the error of each shape. Since it is recognized, it is not necessary to move the detector in the normal direction of the measured object when moving the detected element to each measurement point of the measured object. In addition, it is not necessary to check for each measurement whether or not the shape of the measured object is correctly recognized.
After one measurement is completed, the next measurement can be continued.

In a multidimensional coordinate measuring machine according to a second aspect of the present invention, the detector includes the normal vector detecting means,
The normal vector detection means is a camera that obtains image data of the object to be measured, image processing of the image data from the camera, and coordinates of the edge portion of each measurement point and the first normal vector. And an image processing unit that creates an edge normal vector.

By simply moving the object to be measured and the camera relative to each other so that each measuring point of the object to be measured is within the screen of the camera,
The coordinates of the edge portion of each measurement point and the edge normal vector are obtained from the image processing unit.

In a multidimensional coordinate measuring machine according to a third aspect of the present invention, the detector includes the normal vector detecting means,
The normal vector detection means is a scanning probe which is displaceable in three directions orthogonal to each other when it comes into contact with the object to be measured, and a surface normal vector which is the first normal vector from the displacement amount of the scanning probe. And a surface normal vector calculation unit for calculating

When the scanning probe is brought into contact with the measuring point of the object to be measured, the position coordinates of the measuring machine and the displacement amount of the scanning probe at each measuring point are obtained, and the normal vector calculation unit obtains the scanning obtained at each measuring point. The surface normal vector at each measurement point is calculated from the displacement of the probe. Therefore, when measuring the shape of the measured object, it is possible to automatically recognize the measured shape of the measured object only by bringing the scanning probe into contact with each measurement point of the measured object from any direction.

In the multidimensional coordinate measuring machine according to a fourth aspect of the present invention, the shape recognizing means divides all the measuring points into point elements and linear elements, and based on the point elements and the linear elements. The feature is that an optimum shape is recognized from a plurality of shapes.

All the measuring points are divided into point elements and linear elements, and the optimum shape is recognized from a plurality of complex shapes based on the point elements and linear elements. Therefore, all recognizable shapes are measured. There is no need to predetermine the order.

[0016]

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

FIG. 1 is a block diagram showing a schematic structure of a 2.5-dimensional image measuring machine according to the first embodiment of the present invention, and FIG. 2 is a perspective view showing a schematic structure of the measuring machine.

2. The five-dimensional image measuring machine comprises a measuring machine body 1 shown in FIG. 2 and a control unit 2 shown in FIG.

As shown in FIG. 2, the measuring machine main body 1 is provided with a base 3, a Y stage 4 arranged on the base 3 and movable in the Y-axis direction, and a bridge 5 movable in the X-axis direction. The carriage 6 is provided. A carriage 6 is supported by the X-axis guide 5a of the bridge 5 so as to be movable in the X-axis direction. A CCD camera 8 (see FIG. 1) as a detector is attached to the carriage 6.

As a result, the CCD camera 8 can move in the two axial directions of the X axis and the Y axis. The optical system built into the CCD camera 8 is movable in the Z-axis direction for automatic focusing. The 0.5th dimension means that the optical system of the CCD camera 8 can move in the Z-axis direction for automatic focusing.

By relatively moving the CCD camera 8 and the carriage 6 and the object to be measured placed on the Y stage 4 in the X-axis and Y-axis directions, the two-dimensional shape of the object to be measured can be measured. it can.

As shown in FIG. 1, the image data of the object to be measured output from the CCD camera 8 is input to the image processing device. The image processing device 9 image-processes the image data from the CCD camera 8 to coordinate the edge portion of each measurement point of the DUT and the edge normal vector (first normal vector).
Is calculated, and data representing the coordinates of the edge portion of each measurement point and the edge normal vector are output to the control unit 2. The CCD camera 8 and the image processing device 9 constitute normal vector detection means for detecting the edge normal vector at each measurement point of the object to be measured.

FIG. 7A shows two measurement points and the edge normal vector at each measurement point when the measured shape of the object to be measured is a straight line. FIG. 7B shows three measurement points and the edge normal vector at each measurement point when the measurement shape is a circle. FIG. 7C shows five measurement points and the edge normal vector at each measurement point when the measurement shape is a quadrangle. FIG. 7D shows three measurement points and an edge normal vector at the two measurement points in the case where the measurement shape is a composite shape of “distance between line and point”. FIG. 8A shows four measurement points and edge normal vectors at the four measurement points in the case where the measurement shape is a composite shape of “parallelism between lines”. FIG. 8 (b)
4 shows four measurement points and edge normal vectors at the four measurement points in the case where the measurement shape is a composite shape of “line and line perpendicularity”. FIG. 8C shows four measurement points and edge normal vectors at the four measurement points in the case where the measurement shape is a composite shape of “intersection points of lines”.

FIGS. 3A to 3C are explanatory views showing how to obtain the coordinates of the edge portion of each measurement point and the edge normal vector. FIG. 3A shows an object S to be measured S having a circular cross section, for example.
Measurement points P1 and P of the object to be measured S for measuring the shape of
2 shows a state in which at least one of the carriage 6 and the Y stage 4 is driven so that 2 or P3 falls within the imaging range (screen) 8a of the CCD camera 8. Code 8b
Indicates the measurement range (caliper). Measurement points P1 to P
In each measurement of No. 3, the image processing device 9 performs a predetermined calculation based on the output signal from each pixel in the measurement range 8b to obtain the data representing the coordinates of the edge portion of each measurement point P1 to P3 and each Data representing the edge normal vectors V1 to V3 at the measurement points P1 to P3 are output, respectively. The coordinates of the edge portion are based on the reference position within the imaging range 8a as the origin.

Specifically, the coordinates of the edge portion of each of the measurement points P1 to P3 can be detected by a so-called 256 gradation method or a binarization method.

Next, how to obtain the edge normal vector of each measurement point will be described. To obtain the edge normal vector of the point P5 shown in FIG. 3B, first, the coordinates of the edges of the points P1 to P4 and P5 to P8 in the vicinity of the point P5 and the point P5 are determined by the area calipers C1 to C8. Ask. The obtained coordinate data of the edge portion is transferred to a host computer (not shown). This host computer performs spline interpolation (see FIG. 3 (c)) or linear approximation calculation using the coordinates of the edge portions P1 to P8 to calculate the edge normal vector V at the point P5 (see FIG. 3 (c)). ).

A controller 10 (see FIG. 1) for controlling the relative movement of the carriage 6 and the Y stage 4 is provided. The controller 10 includes a carriage 6 and a Y stage 4
Is output to the control unit 2 representing the coordinates (X, Y).

The control unit 2 is composed of a computer having input means such as a keyboard (not shown), an input circuit, an output circuit, and a central processing unit. The control unit 2 selects a measurement item including one or more geometric shapes according to the number of measurement points, a data capturing section 11 that captures the measurement data of each measurement point and the data of the edge normal vector, the shape calculation section 12. The shape recognition unit 13 is provided. Data acquisition unit 11, shape calculation unit 12
The shape recognition unit 13 is configured to be controlled by the central processing unit.

The data acquisition unit 11 is used by the image processing device 9
Data representing the coordinates of the edge portion and the edge normal vector at each measurement point output from the controller 1 and the controller 1.
The data representing the coordinates of the carriage 6 and the Y stage 4 at each measurement point output from 0 is loaded.

The data capturing section 11 has an edge coordinate calculating section 11a and memories 11b and 11c. The edge coordinate calculation unit 11a adds the data indicating the coordinates of the edge portion output from the image processing device 9 and the data indicating the coordinates of the carriage 6 and the Y stage 4 output from the controller 10, and the addition results are each The data is output as data representing the edge coordinates (measurement value) of the measurement point. The memory 11b stores the edge coordinate data of each measurement point output from the edge coordinate calculation unit 11a.
Reference numeral 1c stores the edge normal vector of each measurement point output from the image processing device 9.

The shape calculation section 12 stores one or more shapes (for example, geometric shapes) included in the measurement item selected by the shape recognition section 13 at the edges of each measurement point stored in the memory 11b of the data acquisition section 11. The calculation is performed based on the coordinate data and a plurality of mathematical formulas that respectively represent a plurality of shapes that are input in advance (hereinafter, the calculated shape is referred to as a measurement shape), and the calculation is performed for each shape (for example, a geometric shape). An error of each measurement shape (hereinafter, this error is referred to as a shape error) is calculated.

More specifically, the shape calculation unit 12 uses the "straight line,
Among the geometrical shapes of “circle, ellipse, hyperbola and parabola”, one or more geometrical shapes included in the measurement item selected by the shape recognition unit 13 are used as the edge coordinate data of each measurement point stored in the memory 11b. , And is configured to calculate the shape error of each measured shape calculated for each geometric shape, as well as based on a mathematical expression representing the geometric shape of a straight line, a circle, an ellipse, a hyperbola, and a parabola that are input in advance. There is.

As the shape error, the straightness which is an error of the measurement shape with respect to a straight line, the roundness which is an error of the measurement shape with respect to a circle, and the error of the measurement shape with respect to an ellipse (this error is referred to as an ellipse error in the following description). ), An error in the measurement shape for the hyperbola (this error is referred to as a hyperbola error in the following description), and an error in the measurement shape for a parabola (this error is referred to as a parabolic error in the following description).

In addition to the above formulas, the shape calculation unit 12 also includes a complex shape including a square (distance between line and point (see FIG. 7 (d)).
Line), parallelism between lines (see FIG. 8A), perpendicularity between lines (see FIG. 8B), and intersection of lines (see FIG. 8).
The mathematical expressions for calculating (c)) are respectively input in advance. The shape calculation unit 12 calculates one of the composite shapes selected by the shape recognition unit 13 based on the edge coordinate data of each measurement point stored in the memory 11b and a mathematical formula for calculating the composite shape. Is configured to.

The shape recognition section (measurement item selection means) 13 is
Normal vector calculation unit 13a and vector comparison unit 13b
And a shape error comparison unit 13c.

The normal vector calculation unit 13a determines the method at each measurement point for the measurement shape calculated by the shape calculation unit 12 among a plurality of geometric shapes such as "straight line, circle, ellipse, hyperbola and parabola". It is configured to calculate a line vector. In the first embodiment, the normal vector calculation unit 13a is shown as constituting a part of the shape recognition unit 13, but the calculation unit 13a may be provided in the shape calculation unit 12. That is, it suffices that the normal vector calculation unit 13a is provided in the control unit 2.

The vector comparison unit 13b causes the shape calculation unit 12 to compare the edge normal vector data stored in the memory 11c of the data acquisition unit 11 and the normal vector data output from the normal vector calculation unit 13a. All of the calculated measurement shapes are compared, and as a result of the comparison, data representing whether or not the two vectors are substantially equal is output.

The shape error comparison unit 13c has a first comparison unit and a second comparison unit. The first comparison unit calculates the shape error (straightness, roundness, ellipticity error, hyperbolic error, and parabolic error) calculated by the shape calculation unit 12 and an error constant (straightness constant, circularity error) input in advance. Constant, elliptic error constant, hyperbolic error constant, and parabolic error constant), and outputs the comparison result. The second comparison unit is configured such that the shape error (straightness, roundness, ellipticity error,
Hyperbolic error and parabolic error). From this comparison result, it is possible to recognize the magnitude relationship between the respective shape errors. Specifically, the measurement shape (straight line, circle, ellipse, hyperbola, and parabola) can be selected in order from the one having the smallest shape error.

The error constants (straightness constant, roundness constant, ellipticity error constant, hyperbolic error constant and parabolic error constant), which are the upper limits of straightness, roundness, ellipticity error, hyperbolic error and parabolic error, are , Each can be input by input means such as a keyboard. Shape recognition is not performed for shapes for which no error constant has been input.

The shape recognizing section 13 determines, according to the number of measurement points,
Points, multiple geometric shapes (straight lines, circles, ellipses, hyperbolas, parabolas and squares) and multiple complex shapes (distance between lines, parallelism between lines, perpendicularity between lines and intersection between lines and lines). The measurement items including one or more of the above are selected.

Specifically, when the number of measurement points is 1, the shape recognition section 13 recognizes the measured shape of the object to be measured as a "point", selects the measurement item of the point, and stores it in the memory 11b. The shape calculation unit 12 is caused to calculate a point based on the edge coordinates. In the case of two points, the measurement shape is recognized as a “straight line”, a straight line measurement item is selected, and the shape calculation unit 12 is caused to calculate the straight line based on the edge coordinates. If there are 3 or 4 points, select the measurement item of "straight line, circle, composite shape", and if there are 5 or more points, select "straight line,
Select a measurement item such as “circle, ellipse, hyperbola, parabola, and complex shape”.

The number of measurement points is, for example, the memory 11b.
Is obtained by counting the number of edge coordinate data stored in the counter.

Further, the shape recognition unit 13 determines the vector comparison unit 1 when the selected measurement item includes a plurality of shapes such as when the number of measurement points is 3 or 4 and when the number of measurement points is 5 or more.
One of the plurality of shapes is recognized based on at least one of the comparison result by 3b and the comparison result by the shape comparison unit 13c.

The control unit 2 having the above-mentioned configuration is configured to execute the shape recognition processing shown in FIG. 4, the composite shape recognition processing shown in FIG. 5, and the measurement point grouping processing shown in FIG. ing.

Next, the operation of the first embodiment having the above structure will be described with reference to FIGS.

When measuring the shape of the object to be measured, for example, as shown in FIG. 3, the measuring point P1 of the object to be measured is placed within the imaging range 8a of the CCD camera 8 while looking at a monitor screen (not shown). Then, an operation member (not shown) is operated to drive at least one of the carriage 6 and the Y stage 4. next,
The operation member is operated so that the measurement point P2 falls within the imaging range 8a. Next, the operation member is operated so that the measurement point P3 is within the imaging range 8a.

In this way, each measurement point P1 to
When P3 is measured, data representing the coordinates of the edge portion of each measurement point and the edge normal vectors V1 to V3 at each measurement point are output from the image processing device 9, and at the time of measurement of each measurement point. Data representing the coordinates of the carriage 6 and the Y stage 4 and the edge normal vector are output from the controller 10. The data capturing unit 11 captures these data.

The edge coordinate calculation unit 11a of the data fetching unit 11 adds the data representing the coordinates of the edge portion and the data representing the coordinates of the carriage 6 and the Y stage 4 to obtain the edge coordinates (measurement value) of each measurement point. The data is output and these data are stored in the memory 11b.

On the other hand, data representing the edge normal vector at each measurement point output from the image processing device 9 is stored in the memory 11c.

Next, the processing operation for recognizing the shape of the object to be measured executed by the control unit 2 based on the data stored in the memories 11b and 11c will be described with reference to FIGS.
It will be described based on.

First, the "shape recognition process" shown in FIG. 4 will be described.

In step 101, it is determined whether or not the number of measurement points is 1. If the determination result is Yes, the process proceeds to step 102, recognizes that the measured shape of the measured object is a “point”, and ends.

If the determination result in step 101 is No,
In step 103, it is determined whether the number of measurement points is 2. When the number of measurement points is 2, the process proceeds to step 104, recognizes that the measured shape of the object to be measured is “straight line”, and ends.

When the number of measurement points is larger than 2, the routine proceeds to step 105, where it is determined whether the number of measurement points is 3 or 4. If the number of measurement points is 3 or 4, step 106
Then, it is determined whether or not the straightness constant has been input in advance. If the straightness constant is entered, step 10
7, the straight line (measurement shape) is calculated by the edge coordinate data of each measurement point stored in the memory 11b and the mathematical expression representing the straight line that is input in advance, and the straightness of the calculated straight line ( Shape error). The allowable value of the calculated straightness is the "straightness constant".

Next, it is determined whether or not the calculated straightness is less than or equal to the straightness constant (step 108). If the straightness is equal to or less than the straightness constant, the process proceeds to step 109, and the edge normal vector at each measurement point stored in the memory 11c and the normal vector at each measurement point calculated by the normal vector calculation unit 13a. It is determined whether and are substantially parallel.

If both normal vectors are substantially parallel, the process proceeds to step 110, the measurement shape is recognized as a "straight line", and the process ends.

When the determination results of steps 106, 108 and 109 are No, the process proceeds to step 111 and it is determined whether or not the roundness constant has been input in advance. If the roundness constant is entered, proceed to step 112,
A circle (measurement shape) is calculated by the data of the edge coordinates (measurement value) of each measurement point stored in the memory 11b and a previously entered mathematical expression representing the circle, and the calculated circularity of the circle ( Shape error). The permissible value of the calculated roundness is the “roundness constant”.

Next, it is determined whether or not the calculated roundness is less than or equal to the roundness constant (step 113). If the roundness is less than or equal to the roundness constant, the process proceeds to step 114, and it is determined whether the edge normal vector and the normal vector at each measurement point are substantially parallel.

If both normal vectors are substantially parallel, the procedure proceeds to step 115, where the measured shape is recognized as a "circle",
Over.

When the determination results of steps 111, 113 and 114 are No, the process proceeds to step 116, the complex shape recognition process shown in FIG. 5 is executed, and the process ends.

When the number of measurement points is 5 or more, that is, when the determination result in step 105 is No, the process proceeds to step 117 and it is determined whether or not the straightness constant has been input in advance. If the straightness constant has been entered, step 118.
Then, the straight line is calculated by the edge coordinate data of each measurement point and the mathematical expression representing the straight line that is input in advance, and the straightness of the calculated straight line is obtained.

After executing step 118, the process proceeds to step 119. If the determination result in step 117 is No,
That is, when the straightness constant has not been input, the process proceeds to step 119 without executing step 118. In this step 119, it is determined whether or not the roundness constant has been input in advance. When the roundness constant is input, the process proceeds to step 120, and the circle is calculated by the edge coordinate data of each measurement point and the previously entered mathematical expression representing the circle, and the calculated circle is the roundness. Ask for degrees.

After executing step 120, the process proceeds to step 121. If the determination result of step 119 is No,
That is, when the roundness constant is not input, the process proceeds to step 121 without executing step 120. In this step 121, it is determined whether or not the elliptic error constant has been input in advance. When the ellipse error constant is input, the process proceeds to step 122, and the ellipse is calculated by the edge coordinate data of each measurement point and the previously input equation representing the ellipse, and the calculated ellipse (measurement shape) is also calculated. Elliptic error (shape error) of The permissible value of the obtained elliptical error is the “elliptical error constant”.

After executing step 122, the process proceeds to step 123. If the determination result of step 121 is No,
That is, when the elliptic error constant is not input, the process proceeds to step 123 without executing step 122. In step 123, it is determined whether or not the hyperbolic error constant has been input in advance. When the hyperbolic error constant is input, the process proceeds to step 124, and the hyperbola is calculated by the edge coordinate data of each measurement point and the previously input mathematical expression representing the hyperbola, and the calculated hyperbola (measurement shape) is calculated. The hyperbolic error (shape error) of is calculated. The permissible value of the calculated hyperbolic error is the “hyperbolic error constant”.

After executing step 124, the process proceeds to step 125. If the determination result of step 123 is No,
That is, when the hyperbolic error constant is not input, the process proceeds to step 125 without executing step 124. In this step, it is determined whether or not the parabolic error constant has been input in advance. If a parabolic error constant is entered,
Proceeding to step 126, a parabola is calculated by the edge coordinate data of each measurement point and a mathematical expression representing a parabola that has been input in advance, and a parabolic error (shape error) of the calculated parabola (measurement shape) is obtained. The allowable value of the calculated parabolic error is the “parabolic error constant”.

As described above, when the number of measurement points is 5 or more, in steps 117 to 126, of the geometric shapes of “straight line, circle, ellipse, hyperbola, and parabola”, the shape to which the error constant is input is Only the measurement shape is calculated based on the edge coordinate data of each measurement point.

After executing step 126, the process proceeds to step 127. If the determination result of step 125 is No,
That is, if the hyperbolic error constant has not been input, the process proceeds to step 127 without executing step 126. In this step, of the calculated measured shapes, the shape with the smallest shape error is obtained.

Next, the process proceeds to step 128 and step 1
It is determined whether or not the shape error of the shape obtained in 27 is less than or equal to an error constant (for example, if the obtained shape is a straight line, it is determined whether or not the straightness is less than or equal to the straightness constant). If the shape error is equal to or less than the error constant, the process proceeds to step 129, and it is determined whether the edge normal vector and the normal vector at each measurement point of the shape obtained in step 127 are substantially parallel.

If both normal vectors are substantially parallel, the process proceeds to step 130, and the shape obtained in step 127 is recognized as the measured shape of the object to be measured, and the processing ends.

If the determination result of step 129 is No,
That is, when the two normal vectors are not substantially parallel, the process proceeds to step 131, and it is determined whether or not the measurement item is completed.
That is, it is determined whether or not the shapes calculated in steps 117 to 126 remain in addition to the shapes obtained in step 127. If the measurement items remain, that is, if the measurement item is not completed, the process proceeds to step 132.

In this step 132, step 117
The shape with the next smallest shape error among the shapes calculated in steps to 126 is obtained.

Next, returning to step 128, step 1
It is determined whether the shape error of the shape obtained in 32 is equal to or less than the error constant. If the shape error is less than or equal to the error constant,
In step 129, it is determined whether the edge normal vector and the normal vector at each measurement point of the shape obtained in step 132 are substantially parallel. If the result of this determination is No, the process proceeds to step 131 and it is determined whether or not the measurement item has ended. Step 12 until the measurement items are completed
8, 129, 131 and 132 are repeated.

If neither of the shapes calculated in steps 117 to 126 is substantially parallel to each other, the determination result in step 131 is No, and the process proceeds to step 133 and "Composite shape recognition" shown in FIG. "Process" is executed and the process ends.

Next, the "composite shape recognition process" shown in FIG. 5 will be described.

First, in step 142 of FIG. 5, the "measurement point grouping process" shown in FIG. 6 is executed.

In step 171 of FIG. 6, all the plurality of measurement points are divided into groups of edge normal vectors whose directions are substantially the same. That is, a point sequence including all measurement points for which edge coordinate data is stored in the memory 11b of the data acquisition unit 11 (for example, if the measurement shape of the measured object is a quadrangle as shown in FIG. For a point sequence consisting of three measurement points), the measurement points whose edge normal vector directions are substantially the same are divided into the same group.

Specifically, when the measured shape of the object to be measured is a quadrangle as shown in FIG. 9, there are six measurement points in total.
For ease of explanation, the six measurement points are labeled with measurement numbers 1 to 6 (see Table 1 and FIG. 9). These 6
When all the one measurement points are divided into groups of edge normal vectors having substantially the same direction, four groups of group 1 to group 4 shown in Table 1 are formed. That is, the number of groups is 4, the number of measurement points of the first group 1 is 3, and the number of measurement points of the second group 2, the third group 3 and the fourth group 4 is 1, respectively.

[0078]

[Table 1] The description after step 172 of FIG. 6 will be given by taking the quadrangle shown in FIG. 9 as an example.

After executing step 171, step 1
Proceeding to 72, a counter (not shown) for counting the number L of groups is reset to L = 0.

Next, at step 173, the count value of the number L of groups is incremented by 1 to set L = 1.

Next, at step 174, it is determined whether L exceeds the number of groups. At this time, L is 1 and does not exceed the number of groups (4). Therefore, the determination result of step 174 is No, and the process proceeds to step 175.

At step 175, the number of measurement points of the L-th group is loaded. At this time, L is 1, and since the first group 1 includes three measurement points of measurement numbers 1 to 3 (see FIG. 9 and Table 1), the number of measurement points 3 is given.

Next, in step 176, it is determined whether or not the number of measurement points is 1. At this time, since the number of measurement points is 3, the determination result of step 176 is No, and the process proceeds to step 178.

In step 178, the number M of measurement points in the L-th group is set to M = 0. That is, a counter (not shown) that counts the number of measurement points M is reset to M = 0.

Next, at step 179, the count value of the number of measurement points M is incremented by 1 to set M = 1.

Next, in step 180, it is determined whether M (M = 1) exceeds the number of measurement points in the group (here, the first group 1 shown in Table 1). At this time, since the number of measurement points in the first group 1 is 3, the determination result in step 180 is No, and the process proceeds to step 181.

In step 181, it is determined whether or not the Mth measurement point has been registered. At this time, the Mth measurement point is the first measurement point 1 shown in FIG.
Has not been registered yet, the determination result in step 181 is No, and the process proceeds to step 182.

In step 182, the coordinates (edge coordinates stored in the memory 11b) of the Mth measurement point (here, the first measurement point 1 shown in FIG. 9) are loaded.

Next, in step 183, the number of measurement points N
To N = 1. That is, the count value of a counter (not shown) for counting the number of measurement points N is set to 1.

Next, in step 184, the number of measurement points N
The count value of the counter that counts is incremented by 1 to N = 2.

Next, in step 185, it is determined whether N (N = 2) has exceeded the number of measurement points in the group. At this time, since the number of measurement points of the first group 1 is 3,
The determination result of step 185 is No, and step 18
Proceed to 6.

In step 186, it is determined whether or not the Nth measurement point has been registered. At this time, N = 2 and the second measurement point 2 (see FIG. 9) is not registered yet, so the determination result of step 186 is No, and the process proceeds to step 187.

In step 187, it is determined whether N = M. At this time, since N = 2 and M = 1, the determination result is No, and the process proceeds to step 188.

In this step 188, the coordinates (edge coordinates) of the Nth measurement point (here, the second measurement point 2 shown in FIG. 9) are loaded.

Next, in step 189, it is determined whether or not the Mth measurement point (here, the first measurement point 1) has been registered as a linear element. At this time, the first measurement point 1
Is not registered as a straight line element, the determination result is No and the process proceeds to step 190.

In this step 190, a straight line is calculated from the Mth measurement point (here, the first measurement point 1) and the Nth measurement point (here, the second measurement point 2).

Then, the process proceeds to step 191, and step 1
The normal vector at each measurement point 1 and 2 perpendicular to the straight line calculated at 90 is calculated, and each calculated normal vector and the memory 11
The edge normal vectors at the measurement points 1 and 2 stored in c are compared with each other to determine whether the normal vectors at the measurement points 1 and 2 are substantially parallel. If both normal vectors are substantially parallel at the measurement points 1 and 2, the process proceeds to step 192.

In this step 192, the M-th measurement point (the first measurement point, that is, the measurement point of measurement number 1 shown in FIG. 9) and the N-th measurement point (the second measurement point, that is, the measurement number shown in FIG. 9). 2 measurement points) are registered as linear elements.

After this registration, the process returns to step 184. Also,
When the determination result of step 191 is No, that is, when the both normal vectors are not substantially parallel at the measurement points 1 and 2, the process returns to step 184. At this time, step 184
Then, the count value of the counter for counting the number of measurement points N is incremented by 1 to set N = 3.

Next, in step 185, it is determined whether N (N = 3) has exceeded the number of measurement points in the group. At this time, since N does not exceed the number of measurement points of the first group 1, the determination result is No and the process proceeds to step 186.

At this time, the Nth measurement point is the third measurement point 3 (see FIG. 9), and since this measurement point 3 has not been registered yet, the determination result in step 186 is No, and step 187 Proceed to.

At this time, since N = 3 and M = 1, the determination result of step 187 is No, and the process proceeds to step 188.

In step 188, the coordinates (edge coordinates) of the Nth measurement point (here, the third measurement point 3) are loaded.

Then, the process proceeds to step 189. At this time,
Since the Mth measurement point (first measurement point 1) has already been registered as a linear element, the determination result of step 189 is Yes.
s, and the process proceeds to step 193.

In this step 193, step 192
The distance (the distance between the line and the point shown in FIG. 7D) between the straight line element registered in step 3 and the Nth measurement point (here, the third measurement point 3) is obtained.

Then, the process proceeds to step 194 and step 1
It is determined whether or not the distance obtained in 93 is less than or equal to a predetermined value. In the example shown in FIG. 9, the Nth measurement point (third measurement point 3) is on the substantially same straight line as the measurement points 1 and 2, and the distance obtained in step 193 is substantially zero.
The determination result of 4 is Yes, and the process proceeds to step 195.

In this step 195, the Nth measurement point (third measurement point 3) is registered as a point on the same line as the linear element registered for the measurement points 1 and 2. That is, N
The third measurement point (third measurement point 3) is registered as the same linear element as the measurement points 1 and 2.

Accordingly, the first group 1 (Table 1
It means that the three measurement points 1 to 3 in () are registered as one linear element.

After executing step 195, the process returns to step 184. If the determination result of step 194 is No, that is, if the distance obtained in step 193 is larger than the predetermined value, the process returns to step 184. At this time, in step 184, the count value of the counter for counting the number of measurement points N is incremented by 1 to set N = 4.

Next, in step 185, it is determined whether N (N = 4) has exceeded the number of measurement points in the group. At this time, since N exceeds the number of measurement points 3 of the first group 1, the determination result is Yes, and step 196
Proceed to.

In this step 196, it is determined whether or not the Mth measurement point (here, the first measurement point 1) has been registered. At this time, since the first measurement point 1 has been registered, the determination result is Yes, and the process returns to step 179.

At this time, in step 179, the number of measurement points M
The count value of is incremented by 1 to M = 2.

Then, the process proceeds to step 180. At this time,
Since M is 2, which does not exceed the number of measurement points 3 in the first group 1, the determination result in step 180 is No,
Go to step 181.

At this time, since the M-th measurement point (second measurement point 2) has been registered, the determination result of step 181 becomes Yes, and the process returns to step 179.

At this time, in step 179, the number of measurement points M
The count value of is incremented by 1 to M = 3.

Then, the process proceeds to step 180. At this time,
Since M is 3 and the number of measurement points in the first group 1 has not exceeded 3, the determination result in step 180 is No,
Go to step 181.

At this time, since the Mth measurement point (third measurement point 3) has already been registered, the determination result of step 181 becomes Yes, and the process returns to step 179.

At this time, in step 179, the number of measurement points M
The count value of is incremented by 1 to M = 4.

Then, the process proceeds to step 180. At this time,
Since M is 4, which exceeds the number of measurement points 4 in the first group 1, the determination result in step 180 is Yes,
Return to step 173.

At step 173, the count value of the number L of groups is incremented by 1 to L = 2.

Next, in step 174, it is determined whether L exceeds the number of groups. At this time, L is 2 and does not exceed the number of groups (4), so the determination result of step 174 is No, and the process proceeds to step 175.

At step 175, the number of measurement points of the L-th group is loaded. At this time, L is 2 and the second group 2 includes one measurement point with the measurement number 4 (see FIG. 9 and Table 1), so the number of measurement points 1 is given.

Next, in step 176, it is determined whether or not the number of measurement points is 1. At this time, since the number of measurement points is 1, the determination result of step 176 becomes Yes,
Go to step 177.

In this step 177, the measurement point of the second group 2 (measurement point of measurement number 4) is registered as a point element, and the process returns to step 173.

At step 173, the count value of the number L of groups is incremented by 1 to L = 3.

Next, in step 174, it is determined whether L exceeds the number of groups. At this time, L is 3 and does not exceed the number of groups (4). Therefore, the determination result of step 174 is No, and the process proceeds to step 175.

At step 175, the number of measurement points of the L-th group is loaded. At this time, L is 3, and the third group 3 includes one measurement point with the measurement number 5 (see FIG. 9 and Table 1), so the number of measurement points 1 is given.

Next, in step 176, it is determined whether or not the number of measurement points is 1. At this time, since the number of measurement points is 1, the determination result of step 176 becomes Yes,
Go to step 177.

In step 177, the measurement point of the third group 3 (measurement point of measurement number 5) is registered as a point element, and the process returns to step 173.

At step 173, the count value of the number L of groups is incremented by 1 to L = 4.

Next, in step 174, it is determined whether L exceeds the number of groups. At this time, L is 4 and does not exceed the number of groups (4). Therefore, the determination result of step 174 is No, and the process proceeds to step 175.

At step 175, the number of measurement points of the L-th group is loaded. At this time, L is 4, and the fourth group 4 includes one measurement point with the measurement number 6 (see FIG. 9 and Table 1), so the number of measurement points 1 is given.

Next, in step 176, it is determined whether or not the number of measurement points is 1. At this time, since the number of measurement points is 1, the determination result of step 176 becomes Yes,
Go to step 177.

In this step 177, the measurement point of the fourth group 4 (measurement point of measurement number 6) is registered as a point element, and the process returns to step 173.

At step 173, the count value of the number L of groups is incremented by 1 to L = 5.

Next, in step 174, it is determined whether L exceeds the number of groups. At this time, since L is 5 and exceeds the number of groups (4), the determination result of step 174 is Yes, and the process ends.

By executing the above-mentioned "measurement point grouping process" shown in FIG. 6, all the plurality of measurement points, that is, all the edge coordinate data stored in the memory 11b of the data fetching section 11 are stored. A point sequence including measurement points (for example, a point sequence consisting of six measurement points when the measured shape of the object to be measured is a quadrangle as shown in FIG. 9) can be divided into straight line elements and point elements. In the case of a quadrangle as shown in FIG. 9, a point sequence consisting of six measurement points
It can be divided into one straight line element and three point elements.

In the "grouping process of measuring points" shown in FIG. 6, the first group 1 shown in FIG. 9 and Table 1 has three measuring points 1 to 3 and, for example, these measuring points. If it includes another measurement point greatly deviated from, it is determined in step 196 whether the Mth measurement point (here, the fourth measurement point) has been registered. At this time, the fourth measurement point is not registered, so
The determination result is No, the process proceeds to step 197, and the fourth measurement point is registered as a point element. as a result,
A point sequence consisting of seven measurement points is divided into one straight line element and three point elements. In this case, an error is processed in step 153 of FIG. 5 described later.

After the "measurement point grouping process" shown in FIG. 6 is executed, the process proceeds to step 142 of the "composite shape recognition process" shown in FIG.

In step 142, it is determined whether the number of registered linear elements is 0. If the number of straight line elements is 0, the process proceeds to step 143 and the number of registered point elements is 1
It is determined whether or not.

If the number of point elements is 1, the process proceeds to step 144, the measured shape of the object to be measured is recognized as a "point", and the memory 1
The point is calculated based on the edge coordinates stored in 1b, and the process ends.

If the number of point elements is larger than 1, the process proceeds to step 145, and it is determined whether the number of point elements is 3 or more.

When the number of point elements is 3 or more, the measured shape is recognized as a "circle", the circle is calculated based on the edge coordinates, and the processing is ended.

If the number of point elements is smaller than 3, the process proceeds to step 147, is treated as an error, and ends.

As described above, when the number of linear elements is 0, the measurement shape is determined by the number of point elements.

If the number of linear elements is not 0, step 1
Proceeding to 48, it is determined whether or not the number of linear elements is one. If the number of straight line elements is 1, the process proceeds to step 149, and it is determined whether or not the number of point elements is 1.

If the number of point elements is 1, the process proceeds to step 150, the measured shape is recognized as a composite shape of "distance between line and point", the shape is calculated based on the edge coordinates, and the processing is ended.

If the number of point elements is larger than 1, the routine proceeds to step 151, where it is determined whether or not the number of point elements is 3.

When the number of point elements is 3, the process proceeds to step 152, the measured shape is recognized as a "quadrangle", the shape is calculated based on the edge coordinates, and the processing is ended.

If the number of point elements is larger than 3, the process proceeds to step 153, is treated as an error, and ends.

As described above, when the number of linear elements is 1, the measurement shape is determined by the number of point elements.

If the number of linear elements is larger than 1, the process proceeds to step 154 and it is determined whether the number of linear elements is 2.
If the number of straight line elements is greater than 2, the process proceeds to step 155, is treated as an error, and ends.

If the number of linear elements is 2, step 156
Then, it is judged whether the number of point elements is 0 or not. If the number of point elements is not 0, the process proceeds to step 157, is treated as an error, and ends.

If the number of point elements is 0, the process proceeds to step 158 to calculate the angle formed by the two straight lines.

Then, the process proceeds to step 159 and step 1
It is determined whether the angle θ calculated in 58 is within the range of 45 degrees <θ <135 degrees. If the determination result is Yes, the process proceeds to step 160, and the measured shape is recognized as a composite shape of “line and squareness of line” (see FIG. 8B), and the coordinates of the intersection of two straight lines are calculated. Is over.

If the determination result of step 159 is No,
Proceed to step 161, and set the measurement shape to "parallelism between lines".
The composite shape (see FIG. 8A) is recognized, the coordinates of the intersection of the two straight lines are calculated, and the process ends.

Thus, when the number of linear elements is 2, 2
The measured shape is recognized as a composite shape of "line-to-line perpendicularity" or "line-to-line parallelism" depending on the angle formed by two straight lines.

As described above, according to the first embodiment, when measuring the shape of the measured object, each measurement point of the measured object is measured by the CCD camera 8 while looking at the monitor screen (not shown). By operating at least one of the carriage 6 and the Y stage 4 by operating an operation member (not shown) so as to enter the imaging range 8a, the shape of the measured object can be automatically recognized.

Specifically, when the number of measurement points is 1 or 2, it is automatically recognized as "point" or "straight line", and when the number of measurement points is 3 or more, the object to be measured is based on the edge normal vector and the shape error. The shape of can be recognized automatically. Therefore, the operator does not have to select the measurement items such as the geometrical shape before the measurement and input them by key, and the relative movement of the detector in the normal direction of the measured object is not necessary, and the shape of the measured object can be It is not necessary to confirm whether or not it is correctly recognized for each measurement, and after one measurement is completed, the next measurement can be continuously performed. Therefore, the shape of the object to be measured can be surely automatically recognized by a simple operation, and the efficiency of the measurement work can be improved.

Further, according to the first embodiment, since the optimum shape is recognized based on the edge normal vector of each measurement point that best represents the shape of the object to be measured, a highly reliable recognition result can be obtained. Obtainable.

According to the first embodiment described above,
The geometric shapes of "points, straight lines, circles, ellipses, hyperbolas and parabolas" can be automatically recognized, and complex shapes including squares (distance between lines, parallelism between lines, perpendicularity between lines and lines and The intersection of the lines) can be automatically recognized.

Further, according to the first embodiment, when the "geometric shape recognition processing" shown in FIG. 4 is executed and none of the plurality of geometric shapes corresponds to the shape of the object to be measured (measurement shape). 5) is executed), the “composite shape recognition process” shown in FIG. 5 can be executed to recognize one of the plurality of composite shapes as the measured shape. In this "composite shape recognition process", the point sequence including all the measurement points is divided into straight line elements and point elements, and one of a plurality of complex shapes based on the number of straight line elements and the number of point elements. Is recognized as the measurement shape, it is not necessary to predetermine the order of measuring the measurement points for all the recognizable complex shapes. Therefore, the operator can measure the measurement points in any order without worrying about the order in which the measurement points are measured, and the measurement operation is also simplified in this respect.

In the first embodiment described above, the types of compound shapes that can be recognized can be increased by predetermining the measurement order of each measurement point corresponding to each of a plurality of compound shapes.

Further, in the first embodiment, the point sequence including all the measurement points is divided into a straight line element and a point element, and one of a plurality of complex shapes is automatically recognized by the number of both elements. However, it is possible to automatically recognize a more complex compound shape (for example, a regular polygon) by making a determination that also takes into consideration the angle between two straight lines, the intersection, and the positional relationship between straight lines. Will also be possible.

Next, a three-dimensional coordinate measuring machine according to a second embodiment of the present invention will be described with reference to the drawings.

FIG. 10 is a block diagram showing a schematic structure of the three-dimensional coordinate measuring machine according to the second embodiment, and FIG. 11 is a perspective view showing a schematic structure of the measuring machine.

As shown in FIGS. 10 and 11, the three-dimensional coordinate measuring machine comprises a measuring machine body 1A and a console 1B. At the tip of the Z-axis 20 of the main body 1A of the measuring device,
When contacting an object to be measured placed on the measuring device body 1
A scanning probe 21 is mounted that is displaceable in three directions (X-axis, Y-axis, and Z-axis directions) orthogonal to A. The operation console 1B is provided with an operation member (joystick) 22 for operating the movement of the scanning probe 21.

In this three-dimensional coordinate measuring machine, the stylus 24 of the scanning probe 21 is brought into contact with each measuring point of the object to be measured by operating the operating member 22 to give a movement command to the measuring machine body 1A. It is possible to measure the three-dimensional shape of an object.

As shown in FIG. 12, the scanning probe 21 includes a mounting portion 23 for the measuring instrument body 1A, a stylus 24, and a support mechanism 25 for supporting the stylus 24 on the mounting portion 23 so as to be displaceable in the above three directions. To 27 and X-axis, Y-axis and Z-axis encoders 28 to 30 that detect the displacement amount of the stylus 24 in each direction.

FIG. 13 is an explanatory view showing a state in which the stylus 24 is in contact with the outer peripheral surface of the cylinder (object to be measured). When the stylus 24 is brought into contact with the outer peripheral surface of the cylinder S1 from any direction indicated by a broken line arrow A in the figure, the scanning probe 21 is displaced with respect to the measuring machine body 1 in a normal direction substantially perpendicular to the contact surface. To do. This displacement, that is, the X axis, Y axis, and Z
Data (measurement data) representing displacement amounts Δx, Δy, and Δz in the axial direction are X-axis, Y-axis, and Z-axis encoders 2, respectively.
Output from 8 to 30 to the control unit 2A.

The measuring instrument body 1A is configured such that the stylus 24 is attached to each measurement point of the object to be measured (for example, measurement points P1 to P shown in FIG.
3) The position coordinates of the measuring machine main body 1A at each measurement point obtained when brought into contact with 3) and the displacement amounts Δx, Δy, and Δz of the scanning probe 21 are added, and the measurement value at each measurement point (sillus A measurement value calculation unit (not shown) for calculating the center position coordinates 24) is provided. From this measurement value calculation unit,
The data (measurement data) of the measurement value at each measurement point is output to the control unit 2A.

The control unit 2A shown in FIG. 10 is composed of a computer having input means such as a keyboard (not shown), an input circuit, an output circuit, a central processing unit, and the like.

The control unit 2A has a data fetching section 3
1, a shape calculation unit 32, and a shape recognition unit 33 that selects a measurement item according to the number of measurement points. The data acquisition unit 31, the shape calculation unit 32, and the shape recognition unit 33 are each configured to be controlled by the central processing unit.

The data fetching section 31 fetches the data indicating the displacement amount of the scanning probe 21 at each measurement point and the data of the measured value at each measurement point from the measuring instrument body 1A and stores these data in the memory. Has become.

In this embodiment, the position coordinate of the measuring machine body 1A at each measurement point and the displacement amount of the scanning probe 21 at each measurement point are added to calculate the measurement value at each measurement point. Although the measurement value calculation unit (not shown) is provided in the measuring machine main body 1A, this calculation unit may be provided in the control unit 2A, for example, in the data acquisition unit 31. In this case, the data capturing unit 31 captures data representing the displacement amount of the scanning probe 21 at each measurement point and the position coordinates of the measuring machine body 1A.

The shape calculation unit 32 calculates one or more geometric shapes included in the measurement item selected by the shape recognition unit 33,
The calculation is performed based on the data of the measurement value of each measurement point stored in the data capturing unit 11 and a plurality of mathematical expressions that respectively represent a plurality of geometric shapes that are input in advance. It is configured to calculate an error (hereinafter, this error is referred to as a shape error).

Specifically, the shape calculation unit 32 is included in the measurement items selected by the shape recognition unit 33 among the shapes of “measured shapes of points, straight lines, circles, planes, spheres, cylinders and cones”. One or more shapes are respectively represented by the measurement value data of each measurement point stored in the data acquisition unit 31 and the geometric shapes of “points, straight lines, circles, planes, spheres, cylinders, and cones” input in advance. It is configured to calculate the shape error of each measurement shape with respect to each geometric shape, as well as to calculate based on the mathematical formula.

As the shape error, the straightness which is the error of the measured shape with respect to a straight line, the circularity which is the error of the measured shape with respect to a circle, the flatness which is the error of the measured shape with respect to a plane, and the The sphericity which is the error, the cylindricity which is the error of the measured shape with respect to the cylinder, and the conicity which is the error of the measured shape with respect to the cone are calculated.

The shape recognition section (measurement item selection means) 33 is
Normal vector calculation unit 33a and surface normal vector calculation unit 3
3b and memories 33c and 33d.

The normal vector calculation unit 33a calculates the geometrical shape of "point, straight line, circle, plane, sphere, cylinder and cone" among
The normal vector at each measurement point is calculated for the measurement shape calculated by the shape calculation unit 32. The data of these calculated normal vectors is stored in the memory 33c.
Is stored. Reference numerals V10 to V30 in FIG. 13 indicate normal vectors calculated based on the “circle” obtained from the measurement points P1 to P3. The normal vector calculation unit 33a
May be in the shape calculator 32. That is, it suffices if the arithmetic unit 33a is provided in the control unit 2A.

The surface normal vector calculation unit 33b is used by the scanning probe 2 at each measurement point stored in the data acquisition unit 31.
The surface normal vector at each measurement point is calculated from the displacement amount data of 1. The data of the calculated surface normal vector is stored in the memory 33d. Reference numerals V1 to V3 in FIG. 13 are surface normal vectors obtained from the data of the displacement amount at each measurement point.

FIG. 17 shows a series of measurement points and surface normal vectors in the measurement shape. FIG. 17A shows two measurement points and a surface normal vector at each measurement point when the measurement shape of the measured object is a straight line. FIG. 17 (b)
Shows three measurement points and a surface normal vector at each measurement point when the measurement shape is a circle. FIG. 17 (c)
Shows three measurement points and a surface normal vector at each measurement point when the measurement shape is a plane. FIG. 17 (d)
Shows four measurement points and a surface normal vector at each measurement point when the measurement shape is a sphere. FIG. 17 (e)
Shows six measurement points and a surface normal vector at each measurement point when the measurement shape is a cylinder. FIG. 17 (f)
Shows six measurement points and a surface normal vector at each measurement point when the measurement shape is a cone.

The shape recognizing unit 33 also stores the normal vector data stored in the memory 33c (for example, the normal vector V10, V20, and V30 data shown in FIG. 13),
The data of the surface normal vector stored in the memory 33d (for example, the data of the surface normal vectors V1 to V3 shown in FIG. 13) is compared with respect to all the measurement shapes calculated by the shape calculation unit 32, and both vectors are compared. A vector comparison unit (not shown) for calculating an angle (angle difference) formed by is provided.

Further, the shape recognizing unit 33 includes the shape calculating unit 3
For each measurement shape calculated in 2, the maximum angle detection unit (not shown) that detects the angle (maximum angle) at which the absolute value of the angle difference is maximum is compared with the maximum angle of each calculated measurement shape. However, a maximum angle comparison unit (not shown) capable of selecting a measurement shape having the smallest maximum angle is provided.

The control unit 2A having the above-mentioned configuration is configured to execute the shape recognition process shown in FIG.

Next, the operation of the second embodiment having the above structure will be described with reference to FIG.

When measuring the shape of the object to be measured, the stylus 2 of the scanning probe 21 is operated by, for example, operating the operating member 22 to give a movement command to the measuring instrument body 1A.
4 is brought into contact with each measurement point of the object to be measured. At this time, the data indicating the displacement amount of the scanning probe 21 at each measurement point and the data of the measurement value at each measurement point are output from the measuring machine body 1A to the control unit 2A.

At this time, the control unit 2A executes the shape recognition processing shown in FIG. First, in step 201,
The measurement value data at each measurement point and the displacement amount data of the scanning probe 21 at each measurement point output from the main body 1A of the measuring machine are read.

Next, in step 202, it is judged whether or not there is a measurement end command. The measurement end command is given, for example, when the push button for instructing the measurement end is pressed. Until there is a measurement end command, steps 201 and 2
02 is repeated and the reading of both data is repeated.

When the measurement end command is issued, the determination result of step 202 becomes Yes, the process proceeds to step 203, and the process branches for each measurement point. That is, a process of selecting a measurement item according to the number of measurement points is performed.

First, when the number of measurement points is 1, step 20
4, the recognized shape of the object to be measured is recognized as a "point", and the process ends.

If the number of measurement points is 2, the process proceeds to step 205, the measurement shape of the object to be measured is recognized as a "straight line", and the process ends.

If the number of measurement points is 3, the process proceeds to step 206, "straight line, circle, plane" is selected as the measurement item, and the process proceeds to step 209. When the number of measurement points is 4 and 5, the process proceeds to step 207, “straight line, circle, plane, sphere” is selected as the measurement item, and the process proceeds to step 209.

If the number of measurement points is 6 or more, step 208
Go to step 209, select “straight line, circle, plane, sphere, cylinder, cone” as the measurement item, and go to step 209.

In this step 209, steps 206, 207 and 20 are performed based on the data of the measured values at the respective measurement points.
The shape included in the measurement item selected in any one of 8
The calculation is performed using a mathematical expression that represents a geometric shape that is input in advance, and the result is stored (stored). For example, when the number of measurement points is 3, the shape of “straight line, circle, plane” included in the measurement item selected in step 206 is calculated by a mathematical expression representing a previously input geometric shape, and the result is stored. To do.

Next, in step 210, for all shapes included in the selected measurement item (for example, if the number of measurement points is 3, all shapes of "straight line, circle, plane"),
The normal vector at each measurement point is calculated from the calculation result of each shape, and the result is stored.

Next, in step 211, for all the shapes included in the selected measurement item, the surface normal vector at each measurement point is calculated from the displacement amount of the scanning probe 21 at each measurement point, and the result is calculated. To store.

Next, in step 212, the angular difference between the normal vector and the surface normal vector at each measurement point is calculated for all the shapes included in the selected measurement item, and the result is stored.

Then, the process proceeds to step 213 and step 2
Correction is performed so that the angle differences calculated in 12 do not exceed ± 90 degrees.

Next, in step 214, the maximum angle at which the absolute value of the angle difference is maximum is found for each shape included in the selected measurement item, and the result is stored.

Next, in step 215, of all the shapes included in the selected measurement item, the shape with the smallest maximum angle is recognized as the measured shape of the object to be measured.

Next, in step 216, it is determined whether or not the selected measurement item includes "circle, cylinder" and the shape (recognition shape) recognized in step 215 is a circle or a cylinder.

If the number of measurement points is less than 6, the selected measurement item does not include "circle, cylinder", and the determination result in step 216 is No. That is, the shape recognized in step 215 is directly recognized as the measured shape of the object to be measured, and the process ends.

When the number of measurement points is 6 or more (that is, the selected measurement item includes “circle, cylinder”) and the shape (recognition shape) recognized in step 215 is a circle or a cylinder. , The determination result of step 216 is Yes.
And proceed to step 217.

In this step 217, the shape error of the circle (roundness) is compared with the shape error of the cylinder (cylindricity), and the shape with the smaller error is recognized as the measured shape of the object to be measured.

As described above, the reason why step 217 is performed is that the normal vector of the circle and the normal vector of the cylinder are the same, so it is necessary to select the measurement shape of the object to be measured from either the circle or the cylinder. Because there is.

As described above, according to the second embodiment, the stylus 24 of the scanning probe 21 is brought into contact with each measurement point of the object to be measured, so that the position coordinates of the measuring instrument body 1A at each measurement point and Since the displacement amount of the scanning probe 21 is obtained and the surface normal vector at each measurement point is calculated from the displacement amount of the probe at each measurement point, the operation member 22 is operated when measuring the shape of the measured object. Then, the measurement shape of the object to be measured can be automatically recognized only by bringing the stylus 24 of the scanning probe 21 into contact with each measuring point of the object to be measured from any direction. Therefore, it is not necessary for the operator to select a measurement item such as a geometric shape and perform a key input before measurement, and it is not necessary to confirm whether or not the shape of the DUT is correctly recognized for each measurement. After the completion of one measurement, the next measurement can be continued. Therefore, the shape of the object to be measured can be surely automatically recognized by a simple operation, and the efficiency of the measurement work can be improved.

Further, according to the second embodiment, since the optimum shape is recognized based on the surface normal vector at each measurement point that most represents the shape of the object to be measured, a highly reliable recognition result can be obtained. Obtainable.

Next, a modification of the second embodiment will be described with reference to FIG.

FIG. 15 is a flowchart showing a part of the operation of this modified example, and the illustration of steps 201 to 214 which are the same as the flowchart shown in FIG. 14 is omitted.

In this modification, step 214 in FIG.
To step 220. In this step 220,
Of all the shapes included in the selected measurement item, the shape having the smallest maximum angle is defined as the recognition shape 1.

Next, in step 221, the shape having the smallest maximum angle among all the shapes included in the selected measurement item is set as the recognized shape 2.

Next, in step 222, it is determined whether the recognized shapes 1 and 2 match. If they match, the determination result of step 222 becomes Yes, and the process ends. That is, the shape recognized in step 2220 (recognition shape 1) is recognized as it is as the measured shape of the object to be measured, and the process ends.

If the recognized shapes 1 and 2 do not match, the determination result in step 222 is No, and the flow advances to step 223.

At step 223, the recognized shape 1
Is a “plane”, the plane is adopted (recognized) as the measurement shape of the DUT, and the process ends. On the other hand, when the recognized shape 1 is a shape other than the “planar surface”, the recognized shape 2 is adopted (recognized) as the measured shape of the object to be measured, and the process ends.

Next, another modification of the second embodiment will be described with reference to FIG.

FIG. 16 is a flow chart showing a part of the operation of this modification, and like FIG. 15, the steps 201 to 214 which are the same as the flow chart shown in FIG. 14 are omitted.

This modification differs from the modification shown in FIG. 15 only in step 224. In this step 224,
If the recognized shapes 1 and 2 do not match, the operator is left to select which shape to select, and the shape selected by the operator is adopted (recognized) as the measured shape of the object to be measured, and the processing ends. .

[0219]

As described above, according to the multi-dimensional coordinate measuring machine of the first aspect of the invention, the operation is simple, and the erroneous recognition of the shape due to the erroneous operation at the time of measurement can be prevented, and Work efficiency can be improved. Further, since the optimum shape is recognized based on the normal vector that most represents the shape of the object to be measured, it is possible to obtain a highly reliable recognition result.

According to the multidimensional coordinate measuring machine of the second aspect of the present invention, when the object to be measured is measured, the object to be measured and the camera are arranged such that the respective measurement points of the object to be measured are within the screen of the camera. Since it is only necessary to move and relative to each other, the measurement operation is very simple.

According to the multidimensional coordinate measuring machine of the third aspect of the present invention, when measuring the shape of the object to be measured, the scanning probe is brought into contact with each measurement point of the object to be measured from any direction. Only by this, the measurement shape of the object to be measured can be automatically recognized, and the measurement operation is very simple.

According to the multidimensional coordinate measuring machine of the fourth aspect of the present invention, the measurement points can be measured in any order without worrying about the measurement order of the measurement points. Become.

[Brief description of the drawings]

FIG. 1 is a schematic cross-sectional view of a first embodiment of the present invention 2.5.
It is a block diagram showing a schematic structure of a two-dimensional image measuring machine.

FIG. 2 is a perspective view showing a schematic configuration of the 2.5-dimensional image measuring machine shown in FIG.

FIG. 3 is an explanatory view of the principle of the 2.5-dimensional image measuring machine shown in FIG.

4 is a flowchart showing the operation of the 2.5-dimensional image measuring machine shown in FIG.

5 is a flowchart showing the operation of the 2.5-dimensional image measuring machine shown in FIG.

FIG. 6 is a flowchart showing the operation of the 2.5-dimensional image measuring machine shown in FIG.

FIG. 7 is an explanatory diagram showing measurement points in each measurement shape and edge normal vectors at each measurement point.

FIG. 8 is an explanatory diagram showing measurement points in each measurement shape and edge normal vectors at each measurement point.

FIG. 9 is an explanatory diagram illustrating a quadrangle as an example.

FIG. 10 is a block diagram showing a schematic configuration of a three-dimensional coordinate measuring machine according to a second embodiment of the present invention.

11 is a perspective view showing a schematic configuration of the three-dimensional coordinate measuring machine shown in FIG.

12 is a side view showing a scanning probe used in the three-dimensional coordinate measuring machine shown in FIG.

13 is an explanatory view of the principle of the three-dimensional coordinate measuring machine shown in FIG.

14 is a flowchart showing the operation of the three-dimensional coordinate measuring machine shown in FIG.

15 is a flowchart showing a part of the operation of a modified example of the three-dimensional coordinate measuring machine shown in FIG.

16 is a flowchart showing a part of the operation of another modification of the three-dimensional coordinate measuring machine shown in FIG.

FIG. 17 is an explanatory diagram showing measurement points and rows of surface normal vectors in each measurement shape.

[Explanation of symbols]

8 CCD Camera (Camera: Normal Vector Detection Means) 9 Image Processing Device (Image Processing Unit: Normal Vector Detection Means) 11, 31 Data Acquisition Unit (Data Acquisition Means) 12, 32 Shape Calculation Unit (Shape Calculation Means) 13 , 33 shape recognition unit (measurement item selection unit) 13a, 33a normal vector calculation unit (normal vector calculation unit) 21 scanning probe (normal vector detection unit) 33b surface normal vector calculation unit

Claims (4)

[Claims] 1. A multidimensional coordinate measuring machine for measuring the shape of an object to be measured by moving the object to be measured and a detector relative to each other, and data capturing means for taking in measured values at respective measurement points of the object to be measured, Normal vector detecting means for detecting the first normal vector at each measurement point, shape recognition means for selecting a measurement item according to the number of measurement points, and included in the measurement items selected by the shape recognition means Shape calculation means for calculating one or more shapes to be calculated based on the measured value and a previously input mathematical expression, and calculating an error of each calculated shape; and each measurement of each calculated shape. A normal vector calculation means for calculating a second normal vector at a point, and the shape recognition means, when the selected measurement item includes a plurality of shapes, compares the comparison result of both normal vectors with the above. Error of each shape A multidimensional coordinate measuring machine characterized by recognizing an optimum shape from among the plurality of shapes based on at least one of the comparison results. 2. The detector includes the normal vector detecting means, and the normal vector detecting means is a camera for obtaining image data of the object to be measured, and image processing is performed on the image data from the camera to perform the image processing. The multidimensional coordinate measuring machine according to claim 1, further comprising: an image processing unit that creates coordinates of an edge portion of each measurement point and an edge normal vector that is the first normal vector. 3. The detector includes the normal vector detecting means, and the normal vector detecting means is capable of displacing in three directions orthogonal to each other when the normal vector detecting means is in contact with the object to be measured, and the scanning probe. The multidimensional coordinate measuring machine according to claim 1, further comprising a surface normal vector calculation unit that calculates a surface normal vector that is the first normal vector from a displacement amount of a probe. 4. The shape recognition means divides all the measurement points into point elements and straight line elements, and recognizes an optimum shape from a plurality of shapes based on the point elements and the straight line elements. The multidimensional coordinate measuring machine according to any one of claims 1 to 3, characterized in that.