RU2280838C2 - Method of contact-free measurement of objects having defocused borders onto image - Google Patents

Abstract

FIELD: measurement technology.

SUBSTANCE: method of contact-free measurement of objects which have defocused borders onto image is based upon registration of object's image in memorizing unit, on presetting of rectangular areas of images fro subsequent detection of object's shape, on performing of differential-integral transforms, upon finding of coordinates of shape and calculating of sizes. In addition due to calibration of measurement system, the width of shape's lines is found depending on shift of object. While measuring, width of shapes of lines is found, and using the found dependence, the width of lines is transformed into distance from borders of object to defocusing plane, which distance is taken into account when calculating sizes of objects. Set of blanks with different thickness can be measured without re-focusing of system. Three-dimensional object sizes can be measured when faces of objects are disposed at some distance from focusing plane and are disposed not in parallel to it.

EFFECT: improved efficiency of operation.

7 dwg

Description

The invention relates to optoelectronic methods for determining the shape and geometric dimensions of objects using multi-element photodetectors. The purpose of the invention is to increase the accuracy of measurements under conditions of random displacement of the measured object relative to the plane onto which the camera lens (the plane of objects) is focused.

A known optical-electronic system for controlling the size of openings in sieves is described in the journal "Autometry" (Siberian Branch of the Russian Academy of Sciences, Institute of Automation and Electrometry, 2003 No. 5, Volume 39, pages 53-61).

The method of measuring holes, implemented in the specified system, consists in sequentially registering holes with a video camera and two-stage processing of the obtained images. At the first stage, the image zones in which the edges of the measured hole are located are automatically determined. At the second stage, brightness transformations are performed in the zones, which make it possible to distinguish the contours of the hole. After that, the two-dimensional coordinates of individual contour points are determined by the center of mass method. The diameter of the hole is calculated as the distance between two diametrically opposite points of the contour.

The method allows with high accuracy and speed to measure round holes stamped in a thin steel sheet of sieves, but only if the surface of the sieve coincides with the focus plane.

During operation, the working surface of the sieve extends and bends, so the edge contour of a particular hole may not coincide with the plane of the sieve shell, i.e. turns out to be out of focus of the camcorder. To accurately measure such holes, you need to focus the camcorder and make the appropriate corrections, since the image is scaled during focusing. However, it is impossible to focus all the points of the edge of the hole on the slopes of the working surface of the sieve, which leads to ambiguity in focusing and measurement errors. In addition, the registration of holes from a certain angle causes the appearance of promising distortions and, as a result, leads to an increase in the measurement error.

By the totality of the features closest to the proposed one is the method described in US patent No. 6621928 "Image edge detection method, inspection system and recording medium", G 06 F 15/42, publ. September 16, 2003

This control method consists in detecting contrast transitions in the image and determining the distances between them. The sequence of operations during the monitoring of objects includes the following steps: saving the image of the controlled object in the storage device; selection of a rectangular area to search for edges; integration of the brightness of the pixels in the selected area along the first direction (parallel to the edge); performing differentiation of the integral function along a direction perpendicular to the edge; determination of the coordinates of the edges; calculation of the external size of the controlled object equal to the distance between two opposite edges of the object.

A system that implements the method described above controls the workpieces moving on the conveyor belt in the field of view of the camera. This method has limited functional capabilities, as it is not suitable for measuring objects that for some reason deviate from the plane onto which the camera lens is focused, for example, standing obliquely or having different thicknesses.

The disadvantages of the method adopted for the prototype include the fact that when determining the size of the workpiece, the two-dimensional coordinates (x, y) of its edges in the image are converted into the plane of the objects of the lens with a fixed proportionality coefficient. In the case of displacement of the workpiece relative to the plane of objects, the optical increase will not correspond to the calculated one and therefore the proportionality coefficient must be changed.

The invention provides a solution to the following problems:

- improving the accuracy of non-contact measurements of objects inadvertently shifted relative to the plane of the camera objects;

- expansion of functionality due to measurements of three-dimensional objects in one image.

These tasks are solved by developing a method of non-contact measurement of objects that have defocused boundaries in the image, including registering the image of the object in the storage device, setting rectangular image areas for subsequent detection of the object’s borders, performing differential-integral transformations, determining the coordinates of the boundaries and calculating sizes, as well as determination procedures the width of the contour lines and converting the width of the lines to the distance from the corresponding boundaries of the object to the square -plane objects, the distance to the object plane into account when calculating the size of the object.

The technical result achieved during implementation is the ability to measure workpieces of different thicknesses without additional lens focusing. In addition, this method makes it possible to measure three-dimensional objects having faces located at a certain distance from the plane of objects and not parallel to it.

The achievement of the technical result is possible due to the fact that there is a direct relationship between the deviation of the subject from the surface on which the lens is focused and the amount of image defocusing. It is known that when an object deviates from the plane of objects, its contours in the image turn out to be unsharp (“blurry”), and the further the object is from the plane of objects, the wider its contour line. In the general case, an uneven contour width is possible in the image, which indicates the inclination of the object’s border to the plane of best focus.

In the claimed invention, the thickness of the lines in the image corresponding to the edges or edges of the object is used as a parameter to determine the distance from the corresponding edge to the focus plane. In order to convert the line width to the third coordinate, you need to know the conversion function. This function — the function of the dependence of the amount of defocusing of contour lines on the distance between the object and the plane of objects — must be determined at the stage of calibration of the optoelectronic system.

So, the proposed method of non-contact measurement of objects contains the following steps (figure 1):

1 - registration of the image of the controlled object in the storage device;

2 - display on the monitor screen;

3 - selection of areas (zones) for the search for boundaries;

4 - mathematical transformations of images in zones and outline selection;

5 - determination of the two-dimensional coordinates of the contour points in each of the zones;

6 - determination of the width of the contour lines;

7 - determination of the distance from the object to the focus plane;

8 - calculation of the size of the controlled object.

Common with the prototype signs: registration of the image of the controlled object in the storage device; display on the monitor screen; selection of areas (zones) to search for boundaries; mathematical transformations of images in zones and the allocation of borders; definition of coordinates of borders; calculation of the size of the controlled object.

New features are the additionally introduced steps for determining the width of the contour line in the image and the distance from the object to the plane of objects, as well as changes in the procedure for calculating the size associated with the height of the boundaries of the object above the focus plane.

Figure 1 schematically shows the sequence of actions in the implementation of the proposed method.

Figure 2 shows the implementation of the method on the example of an automatic sieve control device.

Figure 3 presents the image of the holes of the sieve obtained at various stages of the proposed method.

Example.

The controlled hole of the sieve 1 is inserted into the field of view of the video camera 4 using the scanning table 2, which is controlled by the controller 3. The bottom hole is illuminated by the illuminator 5. At the moment the hole is completely in the field of view, the video camera records the image.

The image of the hole recorded by the camera is converted into a digital code and recorded in the storage device of the computer 6.

Next, select zones to search for hole contours. To do this, first determine the area S and the coordinates (x ₀ , y ₀ ) of the center of mass of the figure formed by pixels whose brightness exceeds a threshold. The threshold can be taken equal to half the brightness of the background and object. All preliminary calculations (figure 1, steps 3-8) are performed in the coordinates of the image. The center of the hole in the first approximation is the center of mass of a bright figure:

where Q _i is the brightness of the pixel; x _i , y _i are its coordinates; n is the number of pixels exceeding the threshold.

Исходя из полученных предварительных оценок центра и радиуса отверстия, выбирают зоны для точного определения края. Зоны располагают так, чтобы их центр находился на расстоянии, равном радиусу R от центра отверстия, при этом линия предполагаемого контура будет находиться примерно в середине зоны (фиг.3а). Количество зон определяется количеством измеряемых радиусов. Размеры зон должны превышать допустимый разброс радиуса отверстия.The area is determined by simply counting the number of bright pixels: S = n. Since the shape of the hole is close to a circle, its average radius can be considered equal

Based on the obtained preliminary estimates of the center and radius of the hole, select zones for the exact definition of the edge. The zones are arranged so that their center is at a distance equal to the radius R from the center of the hole, while the line of the proposed contour will be approximately in the middle of the zone (Fig. 3a). The number of zones is determined by the number of measured radii. The dimensions of the zones must exceed the permissible spread of the radius of the hole.

Using one of the many possible algorithms, perform integral-differential transformations that allow you to select the edge. As a differential transformation that allows you to select the edges of the hole, it is recommended (Obidin Yu.V., Paterikin V.I. Control of through holes of small diameter in the diamond industry. Theory, methods and means of measurement, control and diagnostics. Materials of the II international scientific and practical conference Novocherkassk. September 21, 2001. Part 1, p. 34-41) use a transformation of the form

Here Q _P is the threshold value.

After such a conversion, the image of the contour takes the form shown in figb.

The exact determination of the coordinates (x, y) of one of the contour points is carried out, for example, through the coordinates of the center of gravity of the contour fragment located in the zone (Figs. 3c, 3d). In this case, it is necessary to take into account the curvature of the boundary of the circular hole.

The width of the contour (w) shown in Fig. 3d, in the direction perpendicular to the border, can be determined, for example, at half the maximum brightness value.

In the next step, the width of the contour lines is converted to the z coordinate. The minimum width of the lines corresponds to the coincidence of the edges of the hole with the focus plane, the broadening of the boundary lines indicates the deviation of the edges from the focus plane. When tilting the surface with the hole, the width of the contour will be uneven (Fig.3b).

The conversion is performed according to the calibration table with linear interpolation between the tabular values:

where w is the measured width of the contour; w _i (z _i ), w _{i + 1} (z _{i + 1} ) - the values of the width of the contour in the images of objects offset by the distance z _i and z _{i + 1} from the focus plane.

The calibration table mentioned above must be completed at the stage of setting up the system in which the inventive method is implemented. In practice, calibration can be performed using the photo template of the hole obtained on a glass substrate coated with chrome. The template is placed parallel to the focus plane of the camera and manually moved using micrometric movement in the direction normal to the plane of objects. In each position, the image of the hole is recorded by the video camera and stored in the computer's memory. You can measure the width of the contour in an image using the Photoshop software package or create a special software procedure that measures the width of the contour. The measurement results w _i (z _i ) are entered in the table. The table can be stored in a separate calibration file.

Thus, after performing the above steps, each processed fragment of the edge contains three-dimensional Cartesian coordinates (x, y, z) of one of the points of the contour of the object. At the final stage, it is necessary to transform the coordinates x _i , y _i in the plane of the images to the coordinates in the plane of objects. To determine the size of the object, for example, the diameter, it is necessary to select two fragments at the opposite edges of the hole.

где x₁, y₁, z₁ - координаты одного края отверстия, x₂, y₂, z₂ - координаты противоположного края.The diameter of the hole, equal to the distance between diametrically opposite points of its contour, is calculated in accordance with the expression

where x ₁ , y ₁ , z ₁ are the coordinates of one edge of the hole, x ₂ , y ₂ , z ₂ are the coordinates of the opposite edge.

Moving the processing zone, you can find all the points that belong to the boundary of the object, and restore its shape, including in three-dimensional space.

Thus, when implementing the invention, an increase in the accuracy of measurements is achieved under conditions of unforeseen deviations of the object from the focal plane of the camera.

At the same time, it is also possible to perform measurements without refocusing the system of workpieces of different thicknesses, as well as three-dimensional objects with faces located at a certain distance from the focus plane and not parallel to it.

The introduction of additional steps in image processing expands the possibilities of automating the quality control of industrial products while reducing the requirements for the accuracy of installation of controlled products.

Claims (1)

The method of non-contact measurement of objects having defocused boundaries in the image, including registering the image of the object in the storage device, defining rectangular areas of the image for subsequent detection of the contour of the object, performing differential-integral transformations, determining the coordinates of the contour and calculating dimensions, characterized in that it is additionally calibrated by measuring the systems find the dependence of the width of the contour lines on the displacement of the object, and when measuring, determine the width l contour lines and, using the obtained dependence, convert the line width to the distance from the boundaries of the object to the focus plane, which is taken into account when calculating the size of the object.

Images