JP2737045B2 - How to measure gaps and steps - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for measuring a gap and a step between two adjacent edges of a work such as an automobile body.

[0002]

2. Description of the Related Art An automobile body has an edge portion of a front fender a and a door b shown in FIG. 1A, an edge portion of both front and rear doors b shown in FIG. , FIG.
There is a part where two edges are adjacent, such as the side panel c and the edge of the roof d shown in (c), and it is necessary to manage the gap and the step between the two adjacent edges in order to secure the assembly accuracy of the vehicle body. There is. Conventionally, the gap and the step are measured by a contact method using a gauge or the like, but this may damage the vehicle body.

[0003] Further, an optical measuring device having a light projector for irradiating a slit light on a work and an imager for capturing a light cut image drawn by the slit light irradiated on the work is used to perform light cutting on a screen of the imager. A method of measuring the shape and position of a workpiece from an image is known, and a method of measuring a step between a roof and a slide roof using the method is disclosed in
It is known from JP 61107. In this method, the step between the roof and the slide roof is measured from the amount of deviation between the line on the roof side and the line on the slide roof side of the light-section image.

[0004]

It is conceivable to measure a gap and a step between two adjacent edge portions using the above-mentioned optical measuring device. In this case, when a light cut image is captured by irradiating slit light so as to straddle two edge portions, a light cut image ending at one edge portion and a light cut image ending at the other edge portion are displayed on the screen. The image appears, and it becomes possible to measure the gap and the step between the edge portions from the position of the end portion on each edge portion side of both light cut images.

[0005] However, the end of the light-section image has a tapered curved shape with a radius corresponding to the bent radius of the edge, and it is possible to specify and measure the position of the edge at one point on the screen. Therefore, it is difficult to accurately measure the gap and the step between the edge portions.

The present invention solves the above-mentioned problems and provides a method capable of accurately measuring the gap and the step between the edge portions in a non-contact manner by using the optical measuring device as described above. That is the purpose.

[0007]

In order to achieve the above object,
The present invention uses a light projector that irradiates a slit light to a work, and an optical measurement device that includes an imager that captures a light cut image drawn by the slit light irradiated to the work, and a light cut image on a screen of the imager. In the method for measuring the gap and the step between two adjacent edge portions of a workpiece from a part, a portion extending in a straight line of each light cut image for each of two light cut images divided between the two edge portions Calculating the equation of a first straight line representing the following equation ;
A step of determining the coordinates of the end points, a step asking you to equations of the second straight line passing through the endpoints of each edge portion of each of the optical section image is orthogonal to the first straight line, the first and second intersection of the two straight lines a step of calculating the coordinates, consists of a step of measuring with said gap and the step from the point of intersection of coordinates is calculated from the intersection of coordinates and the other optical section image is calculated from one of the optical section image, said end
The step of obtaining the coordinates of the point is substantially flat in the longitudinal direction of the light-section image.
The coordinate axis on the screen is a first axis, a screen orthogonal to the first axis
With the upper coordinate axis as the second axis, the edge side of each light section image
A window extending outward in the first axial direction from the end of the
Process and margins outside the light section image in the window
The coordinate value of the center of gravity of the image in the second axis direction is expressed by the second axis direction of the end point.
And the position of the center of gravity in the first axis direction
From the value obtained by doubling the standard value, the outer side of the window in the first axial direction is
The value obtained by subtracting the coordinate value in the first axis direction is the first axis direction of the end point.
Determining the coordinate values of the direction .

[0008]

The intersection of the first and second straight lines is a point corresponding to the corner when the edge is bent at a right angle, and the position of the edge can be accurately specified from the coordinates of the intersection. And
The gap and the step between the two edge portions can be accurately measured as a difference between the coordinate values of the two intersections calculated from the two light-section images. In addition, the noise is applied to the outer edge of the curved end of the light-section image.
If irregularities due to noise or the like appear, the outer edge projections
Although it may be detected as an end point of the image, the present invention
If it is not affected by such irregularities, the actual
The coordinates of a typical end point can be obtained. That is, the wi
And the center of gravity of the margin image in the window
The distance in the first axial direction between the outer side and the light section image is substantially
Should be half the distance in the first axial direction between
And the first coordinates of the substantial end point of the light-section image by the method described above.
Axial coordinate values can be calculated. And the end of the light section image
Even if irregularities appear on the outer edge, the area of these irregularities is very small
Because of this, the position of the center of gravity of the margin image in the window
In this way, the coordinates of the end points can be measured with high accuracy without variation
You.

[0009]

1 shows a front fender a and a door b shown in FIG.
An embodiment for measuring the gap and the step between the edge portions of the above will be described. Referring to FIG. 2, reference numeral 1 denotes a projector for irradiating slit light, and 2 denotes an imager including a CCD camera or the like.
An imager 2 that irradiates slit light from the light projector 1 so as to straddle the front fender a and the door b, and is disposed so that an optical axis is oblique to a light surface of the slit light at a predetermined angle (for example, 45 °). Images the irradiated portion of the slit light.

When the slit light is irradiated as described above, light cut images s1, s2, and s3 are drawn on the front fender a, on the door b, and on the flange on the back side of the front fender a facing the gap between both a and b. 3 is displayed on the screen of the imaging device 2.
As shown in FIG. 7, three light-section images S1, S2, and S3 appear. Each light-section image is represented by Y, which is a coordinate axis in the vertical direction of the screen.
The image is taken so as to be substantially parallel to the axis. An image signal from the image pickup device 2 is sent to an image processing device 3 composed of a computer, and the image processing device 3 measures a gap and a step between the edges of the front fender a and the door b.

At the time of this measurement, first, two light section images having a large image area, that is, a light section image S1 on the front fender a and a light section image S2 on the door b are selected as the light section images to be processed. The coordinates of the end point on the edge side of each of the images S1 and S2, that is, the end point A1 below the light-section image S1 in the Y-axis direction and the end point A2 above the light-section image S2 in the Y-axis direction are obtained.

By the way, the edge of each of the light cut images S1 and S2 on the edge portion side has a tapered curved shape with a radius corresponding to the bent radius of the edge portion, and irregularities due to noise and the like appear on the outer edge of the edge. Then, the convex portion is detected as an end point, and therefore, a device is required to find the coordinates of the normal end point that matches the normal shape of the end portion. This will be described with reference to FIG. First, the outermost points in the Y-axis direction of the light-section images S1 and S2 are detected as temporary end points B1 and B2. When irregularities due to noise appear on the outer edges of the end portions of the light-section images S1 and S2, the convex portions are detected as temporary end points B1 and B2. Next, a window W extending outward from the ends of the light-section images S1 and S2 in the Y-axis direction with reference to the temporary end points B1 and B2.
1, W2 are set, and the centers of gravity G1, G2 of the margin images outside the light-section images S1, S2 in the windows W1, W2 are detected. Even if irregularities appear at the outer edge of the end, the positions of the centers of gravity G1 and G2 hardly change because the area of the blank image is much larger than the irregularities.

Here, the distance between the outer sides of the windows W1 and W2 in the y-axis direction and the normal end points A1 and A2 should be twice the distance between the outer sides and the centers of gravity G1 and G2. The Y-axis coordinate values AY of the normal end points A1, A2 and the window W1,
The relationship of WY-AY = 2 (WY-GY) is established between the Y-axis coordinate value WY of the outer side of W2 and the Y-axis coordinate values GY of the centers of gravity G1 and G2, and the following equation is obtained: AY = 2GY-WY Can be used to determine the Y-axis coordinate values of the normal end points A1 and A2.
The X-axis coordinate values of the end points A1 and A2 are the X-axis coordinate values of the center of gravity G1.

After the coordinates of the end points A1 and A2 of the light-section images S1 and S2 have been obtained as described above, a predetermined two positions are defined on the linearly extending portion of the light-section image S1 with reference to the end point A1. Set windows W3 and W4, and set both windows W3 and W4
, The center of gravity G3, G4 of the light section image S1 is detected, and a straight line passing through both the center of gravity G3, G4, that is, the light section image S
An equation of a straight line L1 representing one linear extending portion is obtained, an equation of a straight line L2 orthogonal to the straight line L1 and passing through the end point A1 is obtained, and coordinates of an intersection Q1 of the two straight lines L1 and L2 are calculated. Similarly, two predetermined windows W5 and W6 are defined in the linearly extending portion of the light section image S2 with reference to the end point A2.
Is set, the center of gravity G5, G6 of the light section image S2 in both windows W5, W6 is detected, and a straight line passing through both center of gravity G5, G6, that is, a straight line L3 representing a linear extension of the light section image S2 is detected. And the equation of a straight line L4 orthogonal to the straight line L3 and passing through the end point A2 is obtained.
The coordinates of the intersection Q2 of No. 4 are calculated.

The intersections Q1 and Q2 correspond to the corners when the front fender a and the door b are bent at right angles at the edges, respectively. The position of each edge can be specified accurately.

The spatial coordinate values of the points q1 and q2 on the slit light plane corresponding to the intersections Q1 and Q2 are calculated from the coordinate values on the screen of the intersections Q1 and Q2. This calculation principle will be described with reference to FIG. As shown in FIG. 2, the x-axis and the y-axis on the slit optical surface and the z perpendicular to the slit optical surface are taken with the origin being the intersection of the optical axis 0 of the imager 2 with the slit optical surface.
Consider a spatial coordinate system composed of axes. 5 is a projection plane of the screen of the imaging unit 2 onto the spatial coordinate system, and the coordinates of each point on the slit light plane on the screen are relative to the projection plane CP of the line of sight for each point viewed from the imaging device 2. The coordinates of the intersection. Then, a point on the screen corresponding to the origin of the spatial coordinate system,
That is, when the X axis and the Y axis are taken on the screen (projection plane CP) with the center point of the screen as the origin, the coordinates (Q1X, Q1Y) on the screen of the intersection Q1 and the slit light plane corresponding to the intersection Q1 The following relationship is established between the point q1 and the space coordinates (q1x, q1y).

That is, the inclination angle of the optical axis O of the image pickup device 2 with respect to the slit light plane is α (α is set to a predetermined value by calibration), and the inclination angle of the line of sight to the optical axis O with respect to the point q1 is β. Q1X = q1x · sinα−q1x · cosα · tanβ (1), where N is the distance between the imager 2 and the slit light surface (N is set to a predetermined value by calibration) tanβ = Q1X / N (2), and substituting equation (2) into (1) for q1x, q1x = Q1X / {sinα−cosα × (Q1X / N)}. The ratio between q1y and Q1Y is equal to the ratio between the distance N ′ and N between the image pickup device 2 and the leg lowered from the point q1 to the optical axis O, where N ′ = N + q1x · cosα. From this, q1y = Q1Y {1+ (q1x · cosα) / N}, and thus the spatial coordinates of the point q1 on the slit light surface corresponding to the intersection Q1 can be calculated from the coordinates of the intersection Q1 on the screen.

The spatial coordinates (q2x · q2y) of the point q2 on the slit light surface corresponding to the intersection Q2 are also calculated from the coordinates (Q2X, Q2Y) of the intersection Q2 on the screen in the same manner as above: q2x = Q2X / ｛ sinα−cosα · (Q2X / N)｝ q2y = Q2Y {1+ (q2x · cosα) / N}.

Then, the gap between the edge portions can be accurately measured from the difference between the y-axis coordinate values of the points q1 and q2, and the step between the edge portions can be accurately measured from the difference between the x-axis coordinate values of the points q1 and q2.

The measurement of the gap and the step between the edge portions of the front fender a and the door b has been described above.
The gaps and steps between the edges of the doors b and b shown in FIG. 1B and between the edges of the side panel c and the roof d shown in FIG. 1C can be measured in the same manner as described above.

[0021]

As is apparent from the above description, according to the present invention, the gap and the step between two adjacent edge portions can be measured optically with high accuracy. There is no risk of being scratched.

[Brief description of the drawings]

FIG. 1 is a sectional view of each part of an automobile body (a) (b) (c)

FIG. 2 is a perspective view showing an outline of an optical measuring device.

FIG. 3 is a diagram showing image processing on a screen according to the present invention.

FIG. 4 is a diagram showing image processing for obtaining an end point of a light section image;

FIG. 5 is a diagram showing a relationship between a point on a screen and a corresponding point on a slit light surface.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Projector 2 Imager S1, S2 Light cut image A1, A2 End point L1, L3 First straight line L2, L4 Second straight line Q1, Q2 Intersection

Claims (1)

(57) [Claims] 1. An optical measuring apparatus comprising: a projector for irradiating a slit light on a work; and an imager for capturing a light-cut image drawn by the slit light irradiated on the work, the light cutting on a screen of the imager. In a method of measuring a gap and a step between two adjacent edge portions of a workpiece from an image, each of the two light-section images divided between the two edge portions extends in a straight line of each light-section image. A step of obtaining a first straight line equation representing a portion, and a step of obtaining coordinates of an end point on each edge portion side of each light-section image.
When a step asking you to equations of the second straight line passing through the endpoints of each edge portion of each of the optical section image is orthogonal to the first straight line, the step of calculating the first and second intersection coordinates of the two straight lines If consists of a step of measuring with said gap and the step from the point of intersection of coordinates is calculated from the intersection of coordinates and the other optical section image is calculated from one of the optical section image, the step of determining the coordinates of the endpoints , Light cutting image longitudinal direction
The coordinate axis on the screen substantially parallel to the first axis is orthogonal to the first axis.
The edge of each light section image is set with the coordinate axis on the screen
A window extending outward in the first axial direction from the end on the side of the
And the outside of the light section image in the window
The coordinate value of the center of gravity of the margin image in the second axis direction is
A step of obtaining coordinate values in the axial direction, and a first axis direction of the center of gravity.
Outside the first axis direction of the window from the value obtained by doubling the coordinate value of the direction
The value obtained by subtracting the coordinate value of the side in the first axis direction is used as the value of the end point.
Determining a gap and a step in a uniaxial direction .