JP2013113696A - Displacement measuring method and displacement measuring apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a displacement measuring method and a displacement measuring apparatus in which accuracy of displacement measurement is improved by performing centroid position calculation without being influenced by a pixel having particular luminance.SOLUTION: A displacement measuring apparatus comprises a slit light irradiation device 2 which irradiates a measurement subject 1 with slit light 3, a camera 5 which images a photo cutting line 4 generated by irradiating the measurement subject 1 with the slit light 3, and an image processing device 7 which determines displacement of the measurement subject 1 from a photo cutting line image captured by the camera 5. In the image processing device 7, luminance of pixels in a centroid position calculation direction of the photo cutting line image captured by the camera 5 is approximated to ideal luminance distribution according to the least square method and further approximated again to the ideal luminance distribution according to the weighted least square method using a differential between a luminance value and an approximation value, and a centroid position of the photo cutting line image is calculated from an expectation value of that distribution, thereby determining the displacement of the measurement subject 1.

Description

  The present invention relates to a displacement measuring method and a displacement measuring apparatus for measuring a displacement of a measurement object.

  As a method for measuring the displacement of the measurement object, there is a light cutting method. In this method, the optical cutting line formed by irradiating the measurement target with slit light is photographed by a camera, and the position of the center of gravity is calculated from the luminance data of the optical cutting line image to obtain the displacement of the measurement target.

  A three-dimensional shape recognition apparatus using this method is disclosed in Japanese Patent Laid-Open No. 3-249509 (Patent Document 1). FIG. 7 schematically shows the configuration of the three-dimensional shape recognition apparatus disclosed in Patent Document 1, a slit light generator 70 that irradiates the measurement object with slit light, and the slit light on the measurement object. A camera 71 that captures the formed optical image, an image capturing device 72 that captures image data from the camera 71 in units of pixel cells, a centroid calculator 73 that calculates the centroid of the image data, and coordinates of the image data aggregated in a linear form And a light intensity averager 75 interposed between the image capturing unit 72 and the centroid calculating unit 73. The light intensity averager 75 averages the light intensity of the image cells without variation, and takes into account the light intensity of the surrounding image cells for the central image cell.

  The three-dimensional shape recognition device disclosed in Patent Document 1 having the above configuration performs an averaging process as shown in FIG. Unevenness is reduced, and the calculation accuracy of the center of gravity is improved. That is, for the pixel A in FIG. 8A, the average value of the luminance of the pixels A, B, C, D, and E shown in FIG. 8B, or the pixels A and B shown in FIG. , C, D, E, F, G, H, and I are used, the luminance unevenness of the pixels is reduced, and the calculation accuracy of the center of gravity is improved. Then, after performing the above averaging process on the luminance values of the optical section line image indicated by the X coordinate and Z coordinate of FIG. 8D, which is an enlargement of FIG. 8A, the center of gravity position is calculated. doing.

  A method for calculating the position of the center of gravity of the light section line is disclosed in Japanese Patent Laid-Open No. 6-74724 (Patent Document 2). The method for calculating the barycentric position of the optical cutting line disclosed in Patent Document 2 is to take a picture by moving the camera a plurality of times smaller than the inter-pixel pitch in the barycentric calculation direction of the optical cutting line, thereby increasing the resolution. Thus, the center of gravity calculation data is increased to improve the accuracy of calculating the center of gravity position of the optical section line image.

JP-A-3-249509 JP-A-6-74724

In the light section line image, pixels having specific luminance may appear at random due to interference between light diffusely reflected on the measurement target surface. In the method of calculating the barycentric position of the optical section line image disclosed in Patent Document 1 described above, since the averaging process is performed using the surrounding pixels for every pixel, the barycentric position calculation result is influenced by a specific luminance. Will receive. For example, if a pixel having an abnormally high luminance is included, the luminance of the surrounding pixels is increased by performing the averaging process, and the center of gravity position calculation result is shifted to a pixel position having a high luminance. For this reason, when measuring the surface shape of the measurement target, if there is a pixel having a specific luminance, an error occurs in the gravity center position calculation, and there is a problem that the measured surface shape includes noise.

  In addition, if the pixel position having a specific luminance varies with the passage of time, the position of the center of gravity also varies, resulting in a problem that the repeatability of displacement measurement is reduced.

  Also, in the method of calculating the center of gravity position of the optical cutting line disclosed in Patent Document 2, the center of gravity position is calculated while including pixels having specific brightness. If there is a pixel having luminance, there is a problem that an error occurs in the gravity center position calculation, and noise is included in the measured surface shape.

  The present invention has been made in view of the above problems, and an object of the present invention is to obtain a displacement measuring method and a displacement measuring apparatus that perform center-of-gravity position calculation that is not affected by pixels having a specific luminance and improve the accuracy of displacement measurement. Is.

  The displacement measuring method according to the present invention is a displacement measuring method for calculating a center of gravity position of a light cutting line image obtained by photographing a light cutting line generated by irradiating a measuring object with slit light and obtaining a displacement of the measuring object. In the above, the luminance of each pixel in the centroid position calculation direction of the light section line image is approximated to an ideal luminance distribution by the least square method, and further using the weighted least square method using the difference between the luminance value and the approximate value. Re-approximation to an ideal luminance distribution is performed, and the position of the center of gravity of the light section line image is calculated from the expected value of the distribution to determine the displacement of the measurement object.

  Further, the displacement measuring device according to the present invention includes a slit light irradiation device that irradiates a measurement target with slit light, a photographing device that captures a light cutting line generated by irradiating the measurement target with the slit light, and An image processing device that obtains the displacement of the measurement target from the optical cutting line image obtained by the photographing device, and the image processing device is configured to calculate the center of gravity position of the optical cutting line image obtained by the photographing device. After approximating the luminance of each pixel to an ideal luminance distribution using the least square method, the difference between the luminance value and the approximate value is used to approximate the luminance distribution to an ideal luminance distribution using the weighted least square method. The center-of-gravity position of the optical section line image is calculated from the expected value to determine the displacement of the measurement target.

  According to the present invention, with the above configuration, it is possible to perform displacement measurement and surface shape measurement with less noise, and displacement measurement and surface shape measurement with suppressed measurement variation over time.

It is a schematic diagram which shows the structure of the displacement measuring apparatus which concerns on Embodiment 1 of this invention. It is a schematic diagram which shows the optical cutting line image of the displacement measuring apparatus which concerns on Embodiment 1 of this invention. It is a schematic diagram explaining the displacement measuring method which concerns on Embodiment 1 of this invention. It is a schematic diagram explaining the displacement measuring method which concerns on Embodiment 1 of this invention. It is a schematic diagram explaining the displacement measuring method which concerns on Embodiment 2 of this invention. It is a schematic diagram explaining the displacement measuring method which concerns on Embodiment 2 of this invention. It is a schematic diagram which shows the structure of the conventional three-dimensional shape recognition apparatus. It is a schematic diagram explaining the gravity center position calculation method of the conventional three-dimensional shape recognition apparatus.

DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of a displacement measuring method and a displacement measuring apparatus according to the invention will be described with reference to the drawings. The invention is not limited by these embodiments, and the embodiments can be combined, or the embodiments can be appropriately changed or omitted within the scope of the invention.

Embodiment 1 FIG.
FIG. 1 is a schematic diagram showing a configuration of a displacement measuring apparatus according to Embodiment 1 of the present invention. In FIG. 1, arrows indicated by x, y, and z indicate directions to be described next. That is, x indicates a direction parallel to the major axis of the slit light, y indicates a direction perpendicular to the major axis of the slit light, Z-axis is a direction perpendicular to a plane composed of the x-axis and the y-axis, That is, the height direction is shown.

  The measurement object 1 is an object whose height or surface shape is measured in this embodiment. The slit light irradiation device 2 irradiates the measurement object 1 with slit light 3 having a Gaussian light intensity distribution, and a portion where the slit light 3 is irradiated on the surface of the measurement object 1 becomes an optical cutting line 4. The imaging device, for example, the camera 5 is installed in parallel to the x-axis direction and is installed so that the optical axis is parallel to the yz plane, and images the light section line 4. An image captured by the camera 5 is a light section line image 6 as shown in FIG. The light section line image 6 is a digital image.

  The arrows indicated by x ′ and y ′ in FIG. 3 correspond to the arrows x and y in FIG. 1, respectively, and the arrow x ′ indicates the length direction of the line of the optical section line image 6. An arrow y ′ indicates the line width direction of the light section line image 6. The image processing device 7 captures the optical section line image 6 obtained by photographing with the camera 5, calculates the center position of the luminance in the direction of the arrow y 'in FIG. 3, and displays the center position using the principle of triangulation. 1 is converted to a position in the arrow z direction (height direction). The focus adjustment device 8 adjusts the focus of the slit light 3 irradiated from the slit light irradiation device 2 to the measurement object 1.

  The light source of the slit light irradiation device 2 is a point light source using a single transverse mode laser, and its intensity distribution is a Gaussian distribution. The point light source is uniformly spread by a lens in one direction to be converted into slit light 3, and the measurement target 1 is irradiated with slit light 3 having a Gaussian intensity distribution in the y-axis direction.

  As a point light source in which the intensity distribution is a Gaussian distribution, LED illumination or halogen illumination may be used, and it may be transmitted through a lens, a filter, an optical waveguide, or the like that converts the light intensity distribution into a Gaussian distribution. The point light source may be converted into slit light 3 by scanning using a galvanometer mirror or a polygon mirror.

  Furthermore, after the LED illumination light or halogen illumination light is slit with a lens, it may be transmitted through a lens, a filter, an optical waveguide, or the like that converts the light intensity distribution into a Gaussian distribution. May be converted into a point light source with a lens and slitted with a lens, galvanometer mirror, or polygon mirror, and then transmitted through a lens, filter, optical waveguide, or the like that converts the intensity of light into a Gaussian distribution.

  The focus adjustment device 8 adjusts the focus of the slit light 3 output from the slit light irradiation device 2 using a zoom lens or a varifocal lens (focus variable lens). In addition, instead of the focus adjustment device 8, the point light source before being converted into the slit light 3 may be focus-adjusted using a zoom lens or a varifocal lens (variable focus lens) inside the slit light irradiation device 2. The focus of the slit light 3 may be adjusted by changing the distance between the slit light irradiation device 2 and the measurement target 1.

The displacement measuring apparatus according to the first embodiment is configured as described above, and a method for measuring the height of the measuring object 1 using the apparatus will be described.
The slit light irradiation device 2 irradiates the measurement target 1 with slit light 3 having a Gaussian light intensity distribution. The focus adjustment device 8 is adjusted so that the focus of the slit light 3 is defocused with respect to the measurement target 1, and the light cutting line 4 is set to be blurred. The light cutting line 4 is photographed by the camera 5 and taken into the image processing device 7 as a light cutting line image 6. Since the light cutting line 4 is blurred and the intensity distribution is widened, the luminance distribution of the portion corresponding to the light cutting line 4 in the light cutting line image 6 is also widened, and the number of pixels corresponding to the light cutting line 4 is increased.

  Such a light section line image 6 is taken into the image processing device 7, and the luminance centroid is calculated for the pixel array 9 in the y'-axis direction as shown in FIG. First, a Gaussian distribution Gi such as Equation (1) approximated to the luminance distribution of the pixel array 9 is obtained using the least square method.

  Here, i represents a pixel position in the y′-axis direction, and Gi represents a luminance obtained by approximating the pixel row 9 with a Gaussian distribution at the pixel position i. N, m, and σ are parameters indicating the overall luminance distribution, the expected value of the luminance distribution (center of gravity position), and the spread of the luminance distribution, respectively. These three parameters are obtained by the least square method. is there.

  Next, the difference between the luminance distribution of the pixel array 9 and the Gaussian distribution approximated is obtained, and the weight of each pixel is determined as in the following equation.

  Here, fi is the luminance at the pixel position i in the pixel row 9, di is the difference between fi and Gi, wi is the weight of the pixel at the pixel position i, and W is a threshold for determining the weight. In this equation, the larger the absolute value of the difference di is, the smaller the weight is. Further, if the absolute value of the difference di is larger than a preset threshold value W, the weight is set to 0.

The weight wi only needs to be smaller as the absolute value or the squared value of the difference di increases, and the weight may be set to 0 if the absolute value or the squared value of the difference di is larger than the threshold value W. Alternatively, the above two methods may be mixed. Therefore, instead of the expression (3), the absolute value of the difference di or the inverse of the squared value may be used as the weight wi. Further, the weight may be set to 0 if the absolute value or the squared value of the difference di is larger than the threshold value W, and the weight may be set to 1 if it is smaller than the threshold value W.

  When the weight wi of each pixel is determined, a Gaussian distribution approximated to the luminance distribution of each pixel column is obtained again using a weighted least square method. The parameter m obtained here is the barycentric position of the pixel row 9. For example, assume that the luminance distribution of the pixel column 9 is as shown in FIG. Here, in order to facilitate the calculation, the luminance range was normalized with 0 to 1 without changing the gradation number, and the logarithm of the luminance value was taken. The Gaussian distribution approximated to the logarithmic distribution of the luminance values by the least square method is as shown by a dotted line in FIG. Further, the Gaussian distribution approximated by the weighted least square method is as shown by a solid line in FIG. However, the threshold value W = 1.0. The parameters thus obtained are N≈0.0658, m≈−0.0510, and σ≈1.1180, respectively. In the case of FIG. 4, the barycentric position of the luminance distribution is obtained by converting the parameter m into the luminance position. 7010 pixels.

  Finally, the height or surface shape of the measurement object 1 can be measured using triangulation from the barycentric position obtained by the above method.

  Therefore, when a pixel 10 (see FIG. 2) having a specific luminance is included in the light section line image 6 due to interference between light diffusely reflected on the surface of the measurement object 1, the pixel 10 having a specific luminance is excluded. Alternatively, since the gravity center position is calculated by reducing the degree of influence using weighting, the gravity center position does not shift to a pixel having a specific luminance. Therefore, when the pixels 10 having specific luminance are randomly generated, it is possible to reduce fine noise appearing in the height measurement result or the surface shape measurement result.

  Further, when the position of the pixel 10 having a specific luminance varies with the passage of time, it is possible to reduce the variation of the height measurement result or the surface shape measurement result with the passage of time.

  If the averaging process is performed on the optical section line image 6 as disclosed in Patent Document 1, minute luminance unevenness and noise included in the optical section line image 6 can be reduced. If the pixel having the pixel value is included, the luminance value of the specific pixel decreases, but the luminance value of the surrounding pixel increases, and the position of the center of gravity shifts to the pixel having the specific luminance. Therefore, in order to suppress noise and temporal variations in the height measurement result and the surface shape measurement result, it is effective to calculate the barycentric position by removing pixels having a specific luminance or reducing the influence level. .

Embodiment 2. FIG.
Next, a displacement measuring method and a displacement measuring device according to Embodiment 2 will be described. The second embodiment is different from the first embodiment in the method of calculating the barycentric position of the light section line image 6.
First, the luminance distribution of the pixel column 9 in the y′-axis direction and the luminance distributions of the adjacent pixel columns 11 and 12 shown in FIG. It is the pixel column 9 that obtains the position of the center of gravity. The average of the luminance distribution of the pixel column 11 and the luminance distribution of the pixel column 12 is obtained, and the weight of each pixel is determined for the pixel column 9 as follows.

However, ai is the average value of the luminance of the pixel column 11 and the pixel column 12 at the pixel position i in the y′-axis direction, fi is the luminance of the pixel position i in the pixel column 9, di is the difference between the luminance fi and ai of each pixel, wi is a weight of the pixel position i, and W is a threshold for determining the weight. In this equation, the weight is set to 1 if the absolute value of the difference di is within the preset threshold value W, and the weight is set to 0 if the absolute value is greater than the threshold value W.

  Note that the weight may be set to 0 if the difference di is large compared to the pixel row 9 in the pixel row 9 and the neighboring pixel row. Therefore, the pixel column to be compared with the pixel column 9 may be only the pixel column 11 or only the pixel column 12. A plurality of pixel columns may be used. When there are a plurality of pixel columns to be compared, the difference di may be a difference from the average value of the luminance distributions of the plurality of pixel columns or a difference from the maximum value of the plurality of pixel columns. Moreover, the difference with the minimum value of a some pixel row | line | column may be sufficient. Further, the comparison with the threshold value W may use a value obtained by squaring the difference di instead of the absolute value of the difference di.

  If the weight of each pixel is determined, a Gaussian distribution of equation (1) approximated to the luminance distribution of each pixel column is obtained using a weighted least square method. The parameter m obtained here is the barycentric position of the pixel row 9. For example, assume that the luminance distribution of the pixel column 9 is as shown in FIG. When the average value of the pixel column 11 and the pixel column 12 which are adjacent to the pixel column 9 is obtained and the difference from the luminance distribution of the pixel column 9 is taken, the pixel indicated by the dotted line in FIG. However, the threshold value W = 20. Here, in order to facilitate the calculation, the luminance range was normalized with 0 to 1 without changing the gradation number, and the logarithm of the luminance value was taken. The Gaussian distribution approximated to the logarithmic distribution of the luminance values by the weighted least square method is as shown by a solid line in FIG. The parameters thus obtained are N≈0.0701, m≈−0.0441, and σ≈1.1054. In the case of FIG. 6, the center of gravity position of the luminance distribution is 72.7294 pixels obtained by converting m into the luminance position. It becomes. Therefore, the center-of-gravity position calculation method according to the second embodiment also obtains the same effect as that of the first embodiment.

DESCRIPTION OF SYMBOLS 1 Measurement object 2 Slit light irradiation apparatus 3 Slit light 4 Optical cutting line 5, 71 Camera 6 Optical cutting line image 7 Image processing apparatus 8 Focus adjustment apparatus 9, 11, 12 Pixel row 10 Pixel 70 with peculiar luminance 70 Slit light generation Device 72 Image capture unit 73 Center of gravity calculator 74 Coordinate calculator 75 Light intensity averager

Claims (6)

In the displacement measurement method for calculating the center of gravity position of the light cutting line image obtained by photographing the light cutting line generated by irradiating the measurement target with slit light, and obtaining the displacement of the measurement target,
After approximating the luminance of each pixel in the calculation direction of the center of gravity of the light section line image to the ideal luminance distribution by the least square method, it is ideal by the weighted least square method using the difference between the luminance value and the approximate value. A displacement measuring method characterized by re-approximate to a proper luminance distribution and calculating the center of gravity position of the light section line image from the expected value of the distribution to obtain the displacement of the measurement object. In the displacement measurement method for calculating the center of gravity position of the light cutting line image obtained by photographing the light cutting line generated by irradiating the measurement target with slit light, and obtaining the displacement of the measurement target,
Comparing the arbitrary luminance distribution in the centroid position calculation direction of the light section line image and the luminance distribution in the vicinity thereof, if there is a pixel with extremely different luminance values, the pixel is excluded and ideal by the least square method A displacement measuring method characterized by approximating a luminance distribution and calculating a center of gravity position of the optical section line image from an expected value of the arbitrary luminance distribution to obtain a displacement of the measurement object.   The displacement measuring method according to claim 1, wherein a focus of the slit light is adjusted. A slit light irradiation device for irradiating a measurement object with slit light;
An imaging device that captures a light cutting line generated by irradiating the measurement object with the slit light;
An image processing device for obtaining a displacement of the measurement object from a light section line image obtained by the imaging device,
The image processing apparatus includes:
After approximating the luminance of each pixel in the centroid position calculation direction of the light section line image obtained by the above imaging device to the ideal luminance distribution by the least square method, weighting is performed using the difference between the luminance value and the approximate value. A displacement measuring apparatus which re-approximates an ideal luminance distribution by a least square method and calculates a center of gravity position of the optical section line image from an expected value of the distribution to obtain a displacement of the measurement object. A slit light irradiation device for irradiating a measurement object with slit light;
An imaging device that captures a light cutting line generated by irradiating the measurement object with the slit light;
An image processing device for obtaining a displacement of the measurement object from a light section line image obtained by the imaging device,
The image processing apparatus includes:
Compare any luminance distribution in the centroid position calculation direction of the light section line image obtained by the above imaging device with the luminance distribution in the vicinity, and if there is a pixel with extremely different luminance values, exclude that pixel and minimize it A displacement measuring apparatus that approximates an ideal luminance distribution by a square method and calculates a center of gravity position of the optical section line image from an expected value of the arbitrary luminance distribution to obtain a displacement of the measurement object.   The displacement measuring device according to claim 4, further comprising a focus adjusting device that adjusts a focus of the slit light.

Images