JPH0652167B2 - Image processing method - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for calibrating an image processing apparatus, and more particularly to a calibration method for measuring dimensions or absolute position from a reference plane with the image processing apparatus. Regarding the method.

[Background of the Invention]

Since the image processing apparatus is non-contact and can enlarge and process the shape, it has been widely used for detection and positioning of minute parts. In particular, it has a great effect in detecting a predetermined position of a non-contactable component.

[Conventional technology and problems]

However, due to changes in the focus position due to movement of the camera due to changes in temperature, difficulty in creating a lens with accurate magnification, and variations in pixels including the boundary line when binarizing,
For accurate dimensional measurement and absolute position measurement, calibration must be performed from time to time to correct errors. However, when binarizing, a pixel including a boundary line becomes white or black depending on the value of the threshold, so that accuracy higher than the minimum unit of pixel cannot be obtained.

FIG. 1 and FIG. 2 are diagrams for explaining this, and a small square surrounded by vertical and horizontal fine lines indicates the minimum unit of each pixel.

In FIG. 1, when the apex P shown in FIG. 1 (a) is obtained, the apex P is generally obtained as the intersection of the line segments AP and BP. However, by changing the threshold level, the position of the apex P of the image sequentially changes to P ′, P ″, P as shown in FIGS. 1 (b), (c) and (d).

When the line segments AP and BP are inclined, it becomes more complicated, and the positions of the vertices P are P ′, P ″, and P as shown in FIGS. 2 (a), (b), (c), and (d). In some cases, it may be difficult to judge which of P ′ and Q ′ should be used as the apex.

[Object of the Invention]

The present invention is intended to provide a method of calibrating an image processing apparatus which eliminates the above drawbacks and enables calibration with extremely high accuracy.

[Structure of Invention]

The present invention provides an image processing method in which the position of the center of gravity of a figure having a predetermined shape is calculated, and the position of the center of gravity is used as a reference point for calibration to calibrate with a sufficiently higher accuracy than the minimum unit of pixels. In particular, this is an image processing method in which not only the movement of the reference point but also the inclination can be calibrated by employing two figures drawn at intervals.

Example of Invention

 Hereinafter, the present invention will be described based on examples.

FIG. 3 shows an example of the shape pattern used in the calibration method of the present invention, in which two circles 1a and 1b are drawn. The two circles 1a and 1b are actually black circles, but they are not filled for the sake of explanation. Vertical and horizontal lines drawn so as to overlap two circles indicate pixels for binarizing the color of each dot into black or white as in FIGS. 1 and 2.

In the figure, when the position of the center of gravity of the two circles is the point P and the position of the figure drawn by the imaginary line is the reference, if the actually obtained image is as shown by the solid line, It can be seen that the position P of the center of gravity of the figure has moved to P '. In the actual measurement, the positions of the centers of gravity of the two circles are calculated, and these centers of gravity are designated as P ′, which will be examined.

FIG. 4 shows the binarized figure and the position of the center of gravity of the circle on the right side of FIG. 3 when the threshold level is changed. The center point O drawn by a thin solid line is the reference. The center point of the circle is a center point drawn by a thick short solid line is the barycentric position O ′, O ″, O of the binarized figure. Even at the level, the center position of the circle and the center of gravity of the binarized figure hardly move.

It is natural that the error between the actual position of the center of gravity of the figure and the position of the center of gravity of the binarized figure becomes smaller as the figure gets larger. The difference between the position of the center of gravity of the figure and the position of the center of gravity of the binarized figure is so small that it can be ignored.

Therefore, the barycentric position P'calculated from these two circles becomes extremely accurate and has sufficient accuracy.

Further, according to this method, the inclination θ of the line connecting the centers of gravity of the two figures (circles) is also detected with respect to the inclination of the camera caused by the temperature change, the movement amount of the camera from the reference point, and the error of the magnification.
And the distance L between centers changes to L '.

For large objects that do not fit in the field of view of the lens,
It is generally practiced to use two or more TV cameras, but the same can be applied to such a case, and it is possible to use two or more TV cameras in advance to adjust the center distance and the inclination angle. By detecting the known figures at the same time, the distance and angle between the cameras can be similarly calibrated according to the position and inclination of the figures obtained by each TV camera.

In the above description, two circles drawn at intervals have been described, but it goes without saying that a plurality of figures of arbitrary shapes can be implemented in exactly the same manner. However, if the figure becomes complicated, the calculation of the position of the center of gravity, which is obtained in advance, becomes complicated, and if a large number of figures or pairs of different figures are calculated, the calculation of the position of the center of gravity and the angle of inclination becomes complicated. It is best to use circles of the same diameter.

〔The invention's effect〕

As described above in detail, the present invention defines the position by the position of the center of gravity of the figure, so that the position can be measured with extremely high accuracy compared to the size of each pixel,
The position can be calibrated with sufficiently high accuracy regardless of the size of the pixel, and further, the error of the distance or the angle of inclination can be calibrated with high accuracy.

[Brief description of drawings]

1 and 2 are explanatory views of the conventional method, FIG. 3 is an explanatory view showing the method of the present invention, and FIG. 4 is an explanatory view showing the movement of the center of gravity when the shreshhold level is changed. 1a, 1b ... Circle, P, P '... Position of center of gravity (reference point), L, L' ... Distance between centers of gravity, .theta .... Angle of inclination.

Claims (3)

[Claims] 1. A calibration method for calibrating a measuring device for measuring an absolute position or a dimension by an image processing device, wherein as a reference for calibration, a closed graphic sufficiently larger than a pixel is used, and the graphic is binarized. An image processing method characterized by calibrating an origin or a reference line of a measuring device with the position of the center of gravity of the image as a reference point. 2. A figure is a plurality of figures, and the position of the center of gravity of each binarized image of the plurality of figures is determined,
The image processing method according to claim 1, wherein the position of the center of gravity of the whole is obtained and used as a reference point. 3. The image processing method according to claim 1, wherein the figure is two circles drawn at a predetermined distance.