JPH01214706A - Calibration method in three-dimensional shape measuring apparatus - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a calibration method for automatically calibrating the installation position of a one-dimensional sensor used in a three-dimensional shape measuring device that measures the shape of a target object.

[Prior art] First, the principle of a one-dimensional position detection sensor unit is shown in Fig. 3 (
This will be explained using A).

L is a cylindrical lens, which has the shape of a cylindrical body pressed uniformly from both lateral directions, and its central axis direction Z is set in the vertical direction. S is
1 provided horizontally at the image formation position by the cylindrical lens L.
It is a dimensional line sensor. Now, assume that imaginary planes A, B, and C extend in the radial direction about the central axis Z. a, b.

C are dots located on the planes A, B, and C, respectively, and a+, b+, and c+ are images of the dots a, b, and c formed by the cylindrical lens L, respectively. When the cylindrical lens L is used in this way, the image for the dot becomes a line (focal line) of finite length, and its direction is
It is determined by the angle that the plane containing the focal line and the point light source makes with the optical axis.

Now, when dot a moves downward as shown by arrow 1 on plane A, image a° at this time moves upward as shown by arrow 1. Moreover, if the said driver (a) moves in the direction of the cylindrical lens L as shown by the arrow, the image a' at that time is at the same position. As can be seen from this, even if the dot a on the plane A moves, the position of the line sensor S in the image a° relative to the dot a in the line direction (X direction) remains unchanged. In other words, if the one-dimensional coordinate X, which is the measured value of the one-dimensional position detection sensor unit, is known, the driver la
It is possible to specify a plane A that includes Therefore, as shown in Fig. 4, three such devices (hereinafter referred to as sensor units) (U, U', U'') are placed at different positions, and each line sensor S, S'', S'' are arranged so that they are not in the same direction, and when the spot light is irradiated on point P of the object J, the respective sensor units U, U', U''
Line sensors s, s', s for the irradiation point P obtained in
If one plane is specified from each of the imaging positions S 11 S ′ II S ″i on the image plane S 11 S ′ II S″i, it is possible to find the intersection of the three planes and calculate the three-dimensional object coordinates of the point P. By sequentially scanning the irradiation point P of such a spot light onto the object J, the entire object can be captured as three-dimensional coordinates.

FIG. 3(B) shows a one-dimensional position detection sensor unit using a convex lens 31. In this case, images of dots a, b, and c are formed on a two-dimensional plane, so , it is necessary to use a one-dimensionally divided area sensor 32 as a position detection sensor unit.

[Problems to be Solved by the Invention] In such a measuring device, the lenses L and L' of the sensor units U, U', and U'' in the object coordinate system.

Calibration work to detect the spatial coordinates of the center of L" and its orientation is essential. However, in order to calibrate each sensor unit, it is difficult to calibrate each sensor unit without using a large-scale measurement system such as positioning it on an optical bench. Furthermore, it is extremely difficult to move the sensor unit while maintaining its position and orientation determined on the optical bench.

This invention enables the position and orientation of the one-dimensional position detection sensor unit in an arbitrary object coordinate system including the measurement target to be determined when three or more one-dimensional position detection sensor units are arbitrarily arranged in the field with respect to the measurement target. Provides a calibration method that automatically determines.

[Means for Solving the Problems] The calibration method in the three-dimensional shape measuring device of the present invention includes a spot light irradiation means that irradiates the surface of the object to be measured with spot light, and one-dimensional positioning means arranged at three or more locations. a detection sensor unit, the three-dimensional shape measuring device for measuring three-dimensional coordinates of the spot light from the imaging position of the spot light on the one-dimensional position detection sensor unit, wherein the coordinate position is known in an object coordinate system. A one-dimensional position detection sensor unit sequentially detects a plurality of reference light sources, and based on the detection results, calibrates the set position and orientation of the one-dimensional position detection sensor unit placed at a free position relative to the object to be measured. It is characterized by performing a tion.

[Embodiment] FIG. 1 shows an embodiment of a three-dimensional shape measuring device to which the calibration method of the present invention is applied.

U is a unit that detects a one-dimensional imaging position of a light spot, and is the one-dimensional position detection sensor unit S described above.

Video amplifier 11. Peak hold circuit 12° A/D converter 13. It consists of a peak position/wave height value detection circuit 14.

14 is a peak position/wave height value detection circuit, which detects a one-dimensional position detection sensor S based on the value from the A/D converter 13.
The position of the peak center of the detection signal and its peak value among the pixels are detected. The reason for finding the peak center in this way is that the image of the irradiation point P formed on the one-dimensional position detection sensor S is blurred due to changes in the distance between the object J and the lens system. This is because an image spanning multiple pixels is generally formed due to the influence of the reflectance of the object J, and the wave height value is detected because the detected data is not reflected light from surrounding lighting, etc., but is based on the actual spot. This is to determine whether it is caused by light or not based on the detection level. The data obtained by this peak position/crest value detection circuit 14 is input to a shape measuring section 100, which will be described next.

Although FIG. 1 shows only the one-dimensional position detection sensor unit U, it is assumed that the other sensor units U' and U'' have the same configuration and are arranged as shown in FIG. 4.

Next, the shape measuring section 100 will be explained. 21 is 1 each
This is an object surface coordinate calculation circuit for calculating the three-dimensional coordinates of the irradiation point P of the object J from the one-dimensional position data obtained by the dimensional position detection sensor units U, U', and U''.

Now, each line sensor s for point P in the object coordinate system
, s', s'' are expressed as Sα, S'α, S
If the position and orientation of each one-dimensional position detection unit are known, the coordinates of point P (X
α, Yα, 2α) can be calculated.

22 is a calibration calculation circuit for automatically measuring each of the above-mentioned parameters representing the coordinates and orientation of each sensor unit U, U', U'' before starting measurement of the object J; This section describes a method for calibrating the sensor unit related to this.

Now, the known coordinates of a predetermined number of reference light sources P,,P,... are bound to (X,,Y,,Zυ,(X2.Y,z,)...,
At that time, each one-dimensional position detection sensor s, s', s''
Let the imaging position at (s,,s',,s''+),(S t
, S' t, S'', )..., each one-dimensional position detection sensor unit U, U', U is calculated using these values.
We have obtained parameters representing the position and orientation of ``.The specific method will be explained next.

In Fig. 2, P is the position vector of a reference light point whose three-dimensional coordinates are known in the object coordinate system placed in front of the one-dimensional detection sensor, and P is the optical axis and the optical axis from the point P. Indicates the intersection with the perpendicular line drawn against.

Also, the position vector Q of the lens center and the line sensor S
If the distance from Here U (uvr uvr uz)+
V (wx+Wy+Wz) is a unit vector indicating the attitude of the lens L, and h=Q-W, g=Q-
If U, originally, to determine the position and orientation of the one-dimensional position detection sensor unit, L h* f+ uvr uY
+ uvr Wx + wv* It is necessary to know the 10 unknowns Wz, but in the above equation (2), three-dimensional measurement can be performed just by knowing the seven parameters jl + F + ... t7 of the following equation that combines them. Recognize.

Therefore, automatic calibration for determining the attitude parameters t I+ . . . t of the one-dimensional position detection sensor unit can be performed by the following procedure.

In equation (3), a, b, and c are the 3 points of the reference light point source.
The dimensional coordinates S are values measured by a one-dimensional position detection sensor at the reference light point, and both are known. If this equation (3) is rewritten as a multiplication equation, it becomes one equation including seven posture parameters. Therefore, seven or more reference light sources P1(
ai, bi, ci) with i = 1.2. −, N, the above parameters can be obtained by simultaneously solving equations (4).

The homogeneous equation obtained from the known coordinates of N reference light points and their measured values is: Dt=s... (4) From the above, the least squares solution for finding the unknown attitude parameter vector t is: t - (DtD)-ID's (5) where D is the transposed matrix of D.

In this way, the posture parameters tI...t7 in equation (3) are automatically determined from equation (5) by simply measuring the N reference light points in sequence. After all posture parameters are determined, the three-dimensional coordinates (x, y, z
), by transforming equation (2), -s -t,
...(6) By putting the measured values S++82+83 measured by the three sensing units into the above equation and solving the three-dimensional simultaneous equations, it is possible to obtain the X, Y, and Z of the measurement point.

Returning to FIG. 1, 23 is a memory circuit that stores the surface coordinates of the object J calculated sequentially by the object surface coordinate calculation circuit 21, and the data stored here can be stored in an external storage device or Data stored externally can be taken into the storage circuit 23. 24 is a shape configuration calculation circuit for displaying three-dimensional coordinate data on a display tube such as a CRT display tube; for example, three-dimensional graphics display can be performed by using the wire frame method. ing.

25 rotates the scanner 10 to connect the laser light source 9
26 is a laser scanner control circuit for controlling the laser scanner driving device 25; 27 is the laser scanner control circuit 2
This is a scanning path teaching and automatic setting circuit for setting the measurement range and measurement pattern of the measurement target in advance using a joystick etc. Light scanning is possible. Reference numeral 28 denotes a feature extraction circuit that extracts feature points such as edges and holes of the object J from the coordinate data obtained by the object surface coordinate calculation circuit 21. When an edge is extracted by this feature extraction circuit 28, for example, , the scanning is stopped outside this edge, and the scanning is performed at another position.

Next, a measurement method using the above shape measuring device will be described.

First, three sensor units U placed at arbitrary positions,
In order to calibrate the positions and orientations of U'' and U'' in the object coordinate system, a light source stand M is installed within the measurement area of the sensor unit, which is provided with a plurality of calibration reference point light sources m whose coordinates are known. The number of light sources is determined by the required measurement accuracy, and if the required measurement accuracy is 0.1%, several dozen light sources are used.

Then, after switching the switch SWl to the calibration side, the reference point light source m is turned on in sequence, and each time, each linear sensor s, s' is detected by the three sensor units U, U', U''.
, s", and sends the detection result to the automatic calibration circuit 22. When the measurement for the required number of reference point light sources is completed, by pressing a calculation start button (not shown). , in the automatic calibration calculation circuit 22, the sensor units U, U', U''
The coordinates and orientation of the object are calculated by the method described above, and the calculation results are sent as parameters to the object surface coordinate calculation circuit 21. Now each sensor unit U,
The calibration for the installation positions of U' and U''' is completed.

Next, a procedure for measuring the surface coordinates of the object J will be explained.

First, the changeover switch SWl is switched to the measurement side, and then the changeover switch SW is switched to the operation bus teaching/automatic setting circuit side. Next, the operation bus is taught so that the measurement area including the object J is irradiated with a spot light from the laser light source 9.

The rotation range of the scanner 10 is set from the automatic setting circuit 27. After that, a scanner start button (not shown) is pressed. Thereby, the scanner 10 is driven by the laser scanner driving device 25, and the spot light from the laser light source 9 is directed onto the surface of the object J.

The light is irradiated in units of images and scanned sequentially.

At this time, each irradiation point on the object is measured as one-dimensional coordinate data by the three sensor units U, U', U'', and sent to the object surface coordinate calculation circuit 21 of the shape measurement section 100. The coordinate calculation circuit 21 calculates the three-dimensional coordinates of the irradiation point from the three one-dimensional coordinate data and the parameters according to the calculation formula described above, and the results are sequentially stored in the storage circuit 23.

When the measurement of the entire surface of the object J is completed in this way and all the data is stored in the storage circuit 23, the data can be stored in an external magnetic storage device or the like, or can be stored in the shape configuration calculation circuit. 24, the data can be converted into predetermined data and displayed three-dimensionally on an external display.

Next, a case will be described in which the coordinates and hole diameter of a hole on the object J are measured.

When the switch SW is switched to the feature extraction circuit side,
From the detection data from the object surface coordinate calculation circuit 21, the algorithm in the feature extraction circuit 28 extracts its shape features, and by feeding back the extracted data to the laser scanner control circuit 26, the shape of edges and holes is extracted. can be conveniently measured.

[Effects of the Invention] This invention detects a plurality of reference light sources whose coordinates are known when calibrating a sensor unit of a three-dimensional shape measuring device, and automatically calculates the position and orientation of the sensor unit based on the detection results. As a result, the complicated calibration work has become easier, and as a result,
The sensor unit can be installed at any suitable location for each measurement.

[Brief explanation of the drawing]

FIG. 1 is a control block diagram of a three-dimensional shape measuring device to which this invention is applied, FIG. 2 is a diagram for explaining the invention in detail, and FIG. 3 (A) is for explaining the function of a cylindrical lens. Figure 3 (B) is a schematic diagram of a one-dimensional position detection sensor unit using a convex lens and a one-dimensional divided area sensor, and Figure 4 is a schematic system diagram of a three-dimensional shape measuring device. . U...Sensor unit, S...One-dimensional position detection sensor, L...Cylindrical lens, 100...Shape measurement unit, 2
2...Automatic calibration calculation circuit, M...Light source stand, m...Reference light source. Patent applicant: Kurashiki Boseki Co., Ltd. Agent: Patent attorney: Mr. Maeda, and two others:

Claims (1)

[Claims] (1) A spot light irradiation means for irradiating the surface of the object to be measured with spot light, and one-dimensional position detection sensor units arranged at three or more locations, and the spot light on the one-dimensional position detection sensor unit is provided. In a three-dimensional shape measuring device for measuring the three-dimensional coordinates of a spot light from an imaging position, a one-dimensional position detection sensor unit sequentially detects multiple reference light sources whose coordinate positions are known in an object coordinate system. A calibration method for a three-dimensional shape measuring device, characterized in that, based on the results, the set position and orientation of the one-dimensional position detection sensor unit placed at a free position with respect to the object to be measured are calibrated.