JPS63131007A - Three-dimensional coordinates measuring system - Google Patents

Abstract

PURPOSE:To measure three-dimensional coordinates not relying on the arrangement of a measuring system, by calibrating a conversion matrix for calculating the three-dimensional coordinates from the two-dimensional coordinates of the position of an object. CONSTITUTION:A conversion matrix capable of calculating the three-dimensional coordinates of an object to be inspected by converting the same from the image surface coordinates at the reflecting beam spot on the image of the object to be inspected is provided in a calculation formula. At first, the position of the specimen 10 of the object to be inspected on an inspection stand 16 is set and slit beam 6 is allowed to irradiate the specimen 10 to project an image 15 on the image surface of CRT 13. Then, the coordinates (x, y, z) at the point on a cut line wherein the slit beam 6 cuts the object to be inspected and the corresponding coordinates (u, v) at the point on the image surface are precisely and actually measured. This operation is performed for necessary times according to objective accuracy. Each element of the aforementioned conversion matrix is calculated from the data obtained from said operation to determine the constant in the calculation formula.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a three-dimensional coordinate measuring method of an object using slit light.

(Prior art) Conventionally, this type of three-dimensional coordinate measurement method uses the coordinate calculation formula 18
For simplicity, constraints were placed on the arrangement of the measurement system (the slit light irradiated onto the object and the position of the camera).

[Problem that the invention seeks to solve]

In the conventional three-dimensional coordinate measurement method described above, if the arrangement of the measurement system or the way the coordinate system is changed, the calculation formula changes, making it less versatile, and furthermore, accurate position setting and position measurement of the measurement system are required. However, there is a drawback that these setting errors and measurement errors are much larger than the errors in three-dimensional coordinate measurement.

(Means for Solving the Problems) The three-dimensional coordinate measuring method of the present invention determines the three-dimensional coordinates of an object to be measured by irradiating the object with slit light and reflecting it on an image of the object. It has a transformation matrix that can be converted from the image plane coordinates of a light spot, and data obtained by actually measuring the three-dimensional coordinates of the object position and the corresponding image plane coordinates before measuring the object position coordinates. By calculating each element of the conversion matrix index, the values of the constants in the calculation formula are determined, and in subsequent measurements, the three-dimensional coordinates of the object position are calculated using the calculation formula to calculate the corresponding image plane coordinates. This method is calculated from the value.

[Effect]

The principle of the present invention will be explained with reference to the drawings.

Figure 1 shows the position coordinate system of an object irradiated with slit light,
FIG. 2 is an explanatory diagram showing the relationship between a camera position that images an object and an image plane coordinate system of the object.

The coordinate system of the object position is the x-y-z coordinate system, and the image plane coordinate system of the camera (not shown) is @U-V coordinate system, and the measurement target (not shown) irradiated with the slit light is Side 1
The point ρ(X.

y, z) are imaged at a point p (u, v) on the image plane 4 perpendicular to the optical axis 2 of the camera lens. At this time, point p on the image plane 4 is the point where the straight line connecting the point P to be measured and the optical center 3 of the camera lens intersects with the image plane 4. :
? , p can be used to calculate the x, y, and z coordinate values of point P from the U and V coordinates of point P using a conversion formula.

This conversion formula is expressed as follows using a homogeneous matrix form. Here, S is the homogeneous coordinate, ml, ~m43 are each element of the transformation matrix.0 Formula (1) is a general expression for three-dimensional measurement using slit light, and using this transformation formula, Both the x-y-z coordinate system and the UV coordinate system do not need to be orthogonal coordinate systems, and the unit vectors of each coordinate axis can also be arbitrarily set. That is, there is an advantage that a convenient coordinate system can be arbitrarily set for three-dimensional coordinate measurement.

Next, a method of calibrating the transformation matrix in equation (1) will be explained.

Since equation (1) is in a homogeneous form, one of the non-zero elements of the transformation matrix can be set to 1. Usually, the final element mas will not be 0, so m43 is set to 1 and the undetermined elements are mll~ There are 11 mats.

Equation (1) 'i! By expanding and eliminating S, and rearranging each element mo to ma2 of the transformation matrix, we obtain.

From this, one set of measured data (x, y.

Since three equations are obtained by Z, U, V), 1
In order to solve one unknown element m u -m a □, four sets of actual measurement data corresponding to at least four points are required. Generally, when solving using n sets of measured data, the set number of the measured data is added as a subscript to indicate the conversion formula to be solved.

Letting the matrix on the left side of equation (3) be U, the vector of elements of the transformation matrix as M, and the vector on the right side as X, then equation (3) is as follows: tJM=X (4)
The pseudo-inverse matrix of the matrix U is expressed as U
4, the least squares solution of M can be found by. Here, the pseudo reverse 71-links U4 is iJT
Let be the transposed matrix of U, U″′= (U” Ll) −
'LIT......Defined in (6).

If the conversion matrix is calibrated using the above method, then
There is also the advantage that even if the accuracy of the actual measurement data is not good, it is possible to perform highly accurate calibration by increasing the number of actual measurement data.

(Example) Next, an example of the present invention will be described with reference to the drawings.

FIG. 2 is an external view showing an inspection situation by an automobile chassis mounting nut presence/absence inspection apparatus to which an embodiment of the three-dimensional coordinate measuring method of the present invention is applied.

A nut attached to an automobile chassis is welded to a hole in a steel plate using nut projection.

Check the presence or absence of this nut on the nut side (back side) and the large side (front side).
It is required that the image can be recognized from either direction. With normal binary image processing, it is difficult to determine the presence or absence of a nut because the contour of the nut cannot be extracted accurately, but the installation of a nut can be easily determined by measuring the height of the part.

This device uses an infrared semiconductor laser device 5 as a light source, and uses a slit light 6 whose beam is expanded by a cylindrical lens, so that this slit light 6 has an angle of 30° with respect to the optical axis of a camera 7. A camera 7 and a laser device 5 (wavelength: 780 nm) are fixed to a robot 8, and the optical axis of the camera is perpendicular to the plate 9 to be inspected. Focal length for camera 7! A 50mm lens and a 10mm close-up ring are attached, and a red filter (R-69) is attached to avoid the effects of ambient light. l1lJ'i#t can be almost completely removed. The plate 9 is placed on a black inspection table (not shown) with a depth of 100 mm so as not to receive reflected light from the back side of the hole. The output of the camera 7 is processed by an image processing device (not shown) and displayed on an image plane.

During the inspection, the entire surface of the plate 9 can be inspected by moving the laser device 5 and camera 7 using the robot 8.

FIG. 3 is an external view showing an experimental setup M to which an embodiment of the three-dimensional coordinate measurement method of the present invention is applied.

An adjustment table 17 is placed on the examination table 16 along the examination table coordinate axis (x).

A sample 10 to be inspected is placed on an adjustment table 17, and the position of the sample 10 can be adjusted forward, backward, left and right using screws 182 and 182. After the laser beam emitted from the laser device 5 is converted into a slit beam 6 and reflected by the rotating mirror 11,
is projected onto the sample 10.

The camera 7 images the sample 10 irradiated with the slit light 6, and the image 15 is projected onto the image plane of the CRTI 3 by the image processing device 12 and console 14. CR
The image plane T13 constitutes a two-dimensional coordinate system (UV coordinate system).

FIG. 4 is a flowchart showing the procedure for measuring the position coordinates of the target object in the above two examples.

First, after setting the position M of the plate 9 or sample 10 to be inspected on the inspection table 16, the slit light 6 is irradiated to project 15 images on the image plane. In this way, the coordinates (x,
y, z) and the coordinates (U, V) of the corresponding point on the image plane (step 20). Repeat this operation as many times as necessary depending on the desired accuracy (step 21). The element vector MV of the transformation matrix is calculated using the above equation (5) using the measured data (step 22).
.

The above operations have completed the calibration and obtained the necessary general conversion formula. From now on, the inspection object can be irradiated with the slit light 6, and the coordinates (U, V) of any point on the irradiation line on the image plane
) (step 23), the position coordinates (x, y
, z) using the determined conversion formula (step 24). Coordinate calculations are performed the necessary number of times (step 25), and when completed, the program is terminated.

Note that in all of the above-described embodiments, a camera having a two-dimensional image sensor is used, but when a camera having a -dimensional image sensor is used, the calculation operation becomes much simpler. In this case, since the image is one-dimensional, it can be assumed that the image elements are lined up on the U axis in Figure 1. In this case, ■ Since the coordinate value is always O, the transformation of equation (1) The expression can be rewritten as Here too, the final element m42 of the transformation matrix can be set to 1, so the equation to be solved as in the two-dimensional case is as follows. Each element mll to ma+ of the transformation matrix can be easily obtained using a pseudo-inverse matrix.

In this case, since the number of undetermined elements m+, ~m a + is seven, calibration can be performed using actual measurement data from at least three points.

〔Effect of the invention〕

As explained above, the present invention has a transformation matrix for calculating the three-dimensional coordinates of the object position to be measured from the two-dimensional coordinates of the corresponding object position on the image screen. By calibrating the values of each element of the index using data obtained by actually measuring the object position a necessary number of times before starting measurement, a general formula for obtaining the object position coordinates from the coordinates on the image plane can be obtained. , it becomes possible to measure three-dimensional coordinates without depending on the snow distribution of the measurement system, and there is no need for rough positioning or position measurement of the measurement system, and the causes of measurement errors caused by these are completely eliminated, allowing for high-precision measurement. This has the effect of making it possible.

[Brief explanation of the drawing]

Figure 1 shows the position coordinate system of an object irradiated with slit light,
An explanatory diagram showing the camera position for imaging an object and its relationship with the image plane coordinate system of the object. FIG. Fig. 3 is an external view showing the test equipment M to which a similar embodiment is applied;
The figure is a flowchart showing the operation of the apparatus of FIGS. 2 and 3. 1... Slit optical surface, 2... Lens optical axis,
3... Lens optical center, 4... Image plane, 5
... Laser snow making, 6 ... Slit light,
7... Camera, 8... Robot,
9... Plate, 10... Sample, 1
1... Rotating mirror, +2 -... Image processing device, 13 - CRT, 14... Console, 15... Image, 16 -
... Examination table, 17... Adjustment table, +81.1
82...Screw, x, y, z, u, v''-coordinates.

Claims (1)

[Claims] An object is irradiated with slit light, an image of the irradiated object is captured on an image plane, and the position of the object is determined from the image plane coordinates of a light point reflected by the slit light on the image of the object. It is a three-dimensional coordinate measurement method that calculates coordinates using a calculation formula, and the calculation formula includes a transformation matrix that can be calculated by converting the three-dimensional coordinates of the object from the image plane coordinates of the reflected light point on the image of the object. and, before starting the measurement of the position coordinates of the object, each element of the transformation matrix is calculated in advance from data obtained by actually measuring the three-dimensional coordinates of the object position and the corresponding image plane coordinates, and a calculation formula is calculated. Determine the value of the constant inside, and in subsequent measurements,
A three-dimensional coordinate measurement method that calculates the three-dimensional coordinates of an object position from the corresponding image plane coordinate values using the calculation formula.