JP6015827B2 - Shape measuring device, control method for shape measuring device, and program - Google Patents

Description

  The present invention relates to a shape measuring device by optical scanning, a control method for the shape measuring device, and a program.

  As a typical measurement method of a non-contact type shape measuring apparatus that obtains shape information by optically scanning an object, there is a light cutting method (slit light projection method). The measurement principle of the light section method is based on the principle of triangulation. When slit light is projected onto an object and the object is imaged from another angle, a slit image deformed according to the shape of the object is obtained. The coordinates of the object can be calculated geometrically based on the projection optical system of the slit light, the imaging optical system of the slit image, the projection angle of the slit light, and the observation angle of each point of the deformed slit image. Furthermore, three-dimensional information of the object can be obtained by scanning the slit light on the object.

  In such a measurement, the information obtained from the captured slit image is discrete data for each pixel (pixel) constituting the image, so that the coordinate position of the object can be calculated with higher precision sub-pixel accuracy. Various techniques have been proposed. For example, there is a technique in which the center of gravity is calculated from the spatial distribution (luminance distribution) of the received light intensity, and the resolution is higher than the value determined by the pixel pitch. Patent Document 1 proposes a method of improving measurement accuracy by excluding erroneous center-of-gravity information due to multiple reflection or the like in this center-of-gravity calculation. Specifically, when the brightness threshold value for extracting the brightness distribution for performing the center of gravity calculation is changed, if the calculated center of gravity information has a difference of a certain degree or more, the information is excluded as invalid. Further, Patent Document 2 proposes a method of obtaining a luminance peak point by Gaussian fitting of measured discrete data.

JP 2002-22423 A JP 2009-74814 A

  However, these methods have a problem that a measurement target is limited when measurement with higher accuracy is required, and a problem that measurement takes time. Specifically, the method according to Patent Document 1 has a problem that it is difficult to measure a measurement target whose center of gravity position changes when the brightness threshold value of the brightness distribution for performing the center of gravity calculation is changed. For example, in the case of an object where reflection of light is observed not only from the surface of the object but also from a slight depth, such as subsurface scattering, the brightness distribution does not become symmetric and is processed as effective information. It was difficult to measure. Further, in the method according to Patent Document 2, there is a problem that it takes time for Gaussian fitting, and it is difficult to measure with high accuracy in the case of an object whose reflected light luminance distribution does not match Gaussian fitting. was there.

  SUMMARY An advantage of some aspects of the invention is to solve at least a part of the problems described above, and the invention can be implemented as the following forms or application examples.

  Application Example 1 The shape measurement apparatus according to this application example includes an optical scanning unit that scans an object with projection light, an imaging unit that acquires a measurement image by reflected light of the projection light from the object, and the acquired measurement image. Reflection position detection means for calculating the reflection position of the object based on the reflection position detection means, the reflection position detection means from the measurement luminance pattern distributed to the plurality of pixels constituting the acquired measurement image, the object reflection position candidate coordinates The measured luminance pattern is compared with a plurality of reference luminance patterns prepared at intervals of a predetermined resolution within the comparison coordinate range including the reflection position candidate coordinates, and the measured luminance pattern and each reference luminance pattern are compared. The pattern similarity is calculated, and the reflection position coordinates of the object are obtained from the coordinate information specified by the reference luminance pattern having the largest pattern similarity.

According to this application example, instead of simply recognizing the pixel position indicating the maximum luminance of the acquired measurement image as the reflection position of the object, the result of comparing the measurement luminance pattern distributed over a plurality of pixels with the reference luminance pattern Thus, the reflection position coordinates of the object are obtained. By applying this method, a reference luminance pattern for specifying the reflection position coordinates of the object can be individually prepared in advance according to the reflection characteristics of the object. As a result, more accurate measurement can be performed in accordance with the reflection characteristics of the object.
Also, since the degree of similarity between the measured luminance pattern distributed in multiple pixels and the reference luminance pattern is calculated and the calculated similarity is compared, the reflection position coordinates of the object can be obtained. Compared to the process of performing the above, the time required is short and high-speed measurement is possible.

  Application Example 2 In the shape measuring apparatus according to the application example described above, a plurality of reference luminance patterns are provided in advance as a plurality of reference pattern sets in accordance with the distribution characteristics of reflected light of a plurality of types of objects. It is preferable that a corresponding reference pattern set can be selected depending on the type.

According to this application example, according to reflection characteristics that vary depending on the type of object, a reference pattern set having a corresponding distribution characteristic is prepared in advance, and by selecting a reference pattern set for comparison that is optimal for measurement, More accurate measurement is possible. Specifically, for example, when the reflection from the object includes not only the reflection from the object surface but also the reflection from the lower layer, the distribution characteristic is not a distribution approximated by a single Gaussian distribution, and is inherent to the object. Distribution. By preparing a reference pattern set according to this object-specific distribution in advance, it is possible to perform comparison with which pattern matching with higher accuracy can be obtained, so that measurement with higher accuracy can be performed.
Also, by preparing a reference pattern set of reflection characteristics corresponding to various objects assumed as measurement targets in advance, it is possible to measure various objects with high accuracy simply by selecting the optimal reference pattern set for measurement. Is possible.
Therefore, it is possible to provide a shape measuring apparatus capable of measuring with higher accuracy without limiting the measurement target due to the reflection characteristics.

  Application Example 3 In the shape measuring apparatus according to the application example described above, it is preferable that the resolution and the comparison coordinate range can be set in advance during measurement.

According to this application example, since the resolution can be set at the time of measurement, the time required for comparison can be adjusted within a necessary and sufficient range. Specifically, for example, in the case of measurement that does not require so much measurement accuracy, the time required for comparison can be shortened by setting the resolution to the pixel interval or about half of the pixel interval. it can.
In addition, since a coordinate range for pattern comparison can be set during measurement, the time required for comparison can be adjusted within a necessary and sufficient range. Specifically, in setting the comparison coordinate range, when the range to be compared before and after the calculated reflection position candidate coordinates is narrowed, the number of pattern comparisons is reduced, so the measurement time is shortened.

  Application Example 4 In the shape measuring apparatus according to the application example described above, the resolution is preferably equal to or less than the pixel interval.

  According to this application example, the reflection position of the object is specified by comparing the measurement luminance pattern distributed in the plurality of pixels constituting the measurement image with the plurality of reference luminance patterns prepared at a pitch equal to or less than the pixel interval. Therefore, the calculated reflection position coordinates of the object are detected with a finer precision than the pixel interval having luminance information. That is, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 5 In the shape measuring apparatus according to the application example described above, the coordinates indicating the peak point of the reference luminance pattern having the highest pattern similarity are used as the reflection position coordinates of the object.

  According to this application example, the luminance peak point of the reference luminance pattern that is most similar to the measurement luminance pattern distributed in the plurality of pixels constituting the measurement image is specified as the reflection position coordinates of the object. Since the detection is based on the luminance distribution, the reflection position coordinates of the object are detected with a finer accuracy than the pixel interval having the luminance information. Therefore, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 6 In the shape measuring apparatus according to the application example, the pattern similarity is a normalized cross-correlation value between the measurement luminance pattern and the reference luminance pattern, and is obtained by comparing the measurement luminance pattern with a plurality of reference luminance patterns. When the obtained distribution of normalized cross-correlation values is approximated by a quadratic function, the reflection position coordinates of the object are obtained from position information where the value of the quadratic function shows the maximum value.

  According to this application example, the pattern similarity is evaluated based on the normalized cross-correlation value between the measured luminance pattern and the reference luminance pattern. Therefore, the pattern shape similarity is evaluated without being affected by the difference in reflected luminance. . Further, by obtaining the reflection position coordinates based on the maximum value obtained by approximating the distribution of normalized cross-correlation values with a quadratic function, it is possible to detect the position with higher accuracy. As a result, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 7 In the shape measuring apparatus according to the application example described above, the reference luminance pattern is composed of a pixel luminance pattern and a subpixel luminance pattern, and the pixel luminance pattern is derived from a function in which one or a plurality of Gauss functions are synthesized. The sub-pixel luminance pattern is derived from the pixel luminance pattern.

  According to this application example, it is possible to perform comparison for position detection with a pattern more similar to the measured luminance pattern based on the actual reflection characteristics. As a result, it is possible to detect the position with high accuracy with little error due to the object. As a result, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 8 In the shape measurement apparatus according to the application example described above, the reference luminance pattern includes a pixel luminance pattern and a subpixel luminance pattern. The pixel luminance pattern is derived from the measured luminance pattern, and the subpixel luminance pattern is a pixel. It is derived from the luminance pattern.

  According to this application example, the reference luminance pattern is derived based on the distribution information of the measured luminance pattern obtained from the reflected light of the actual object. Specifically, a pixel luminance pattern is generated using distribution information of a measured luminance pattern, instead of generating a similar pattern by combining a plurality of functions. In this way, even an object whose reflected light luminance distribution does not match Gaussian fitting can be detected with high accuracy with little error. Further, when it is difficult to synthesize a similar reference luminance pattern from a plurality of Gaussian functions, or when it takes time to synthesize, a reference luminance pattern can be created more easily.

  Application Example 9 The control method of the shape measuring apparatus according to this application example includes a step of scanning an object with projection light, a step of acquiring a measurement image by reflected light of the projection light from the object, and the acquired measurement image Are prepared at predetermined resolution intervals in a comparison coordinate range including a measurement luminance pattern and a reflection position candidate coordinate from a measurement luminance pattern distributed in a plurality of pixels constituting The step of calculating the pattern similarity between the measured luminance pattern and each reference luminance pattern by comparing a plurality of reference luminance patterns, and reflecting the object from the coordinate information specified by the reference luminance pattern having the largest pattern similarity Obtaining position coordinates.

According to this application example, instead of simply recognizing the pixel position indicating the maximum luminance of the acquired measurement image as the reflection position of the object, the result of comparing the measurement luminance pattern distributed over a plurality of pixels with the reference luminance pattern Thus, the reflection position coordinates of the object are obtained. By applying this method, a reference luminance pattern for specifying the reflection position coordinates of the object can be individually prepared in advance according to the reflection characteristics of the object. As a result, measurement with higher accuracy is possible in accordance with the reflection characteristics of the object without causing a difference in measurement results due to the reflection characteristics of the object.
Also, since the degree of similarity between the measured luminance pattern distributed in multiple pixels and the reference luminance pattern is calculated and the calculated similarity is compared, the reflection position coordinates of the object can be obtained. Compared to the process of performing the above, the time required is short and high-speed measurement is possible.

  Application Example 10 In the control method of the shape measuring apparatus according to the application example described above, a step in which a plurality of reference luminance patterns are provided in advance as a plurality of reference pattern sets in accordance with distribution characteristics of reflected light of a plurality of types of objects, and measurement In this case, it is preferable to include a step of selecting a corresponding reference pattern set according to the type of the object.

According to this application example, according to reflection characteristics that vary depending on the type of object, a reference pattern set having a corresponding distribution characteristic is prepared in advance, and by selecting a reference pattern set for comparison that is optimal for measurement, More accurate measurement is possible. Specifically, for example, when the reflection from the object includes not only the reflection from the object surface but also the reflection from the lower layer, the distribution characteristic is not a distribution approximated by a single Gaussian distribution, and is inherent to the object. Distribution. By preparing a reference pattern set according to this object-specific distribution in advance, it is possible to perform comparison with which pattern matching with higher accuracy can be obtained, so that measurement with higher accuracy can be performed.
Also, by preparing a reference pattern set of reflection characteristics corresponding to various objects assumed as measurement targets in advance, it is possible to measure various objects with high accuracy simply by selecting the optimal reference pattern set for measurement. Is possible.
Therefore, it is possible to provide a shape measuring apparatus capable of measuring with higher accuracy without limiting the measurement target due to the reflection characteristics.

  Application Example 11 In the method for controlling the shape measuring apparatus according to the application example described above, it is preferable that the resolution and the comparison coordinate range are set in advance for measurement.

According to this application example, since the resolution can be set at the time of measurement, the time required for comparison can be adjusted within a necessary and sufficient range. Specifically, for example, in the case of measurement that does not require so much measurement accuracy, the time required for comparison can be shortened by setting the resolution to the pixel interval or about half of the pixel interval. it can.
In addition, since a coordinate range for pattern comparison can be set during measurement, the time required for comparison can be adjusted within a necessary and sufficient range. Specifically, in setting the comparison coordinate range, when the range to be compared before and after the calculated reflection position candidate coordinates is narrowed, the number of pattern comparisons is reduced, so the measurement time is shortened.

  Application Example 12 In the control method of the shape measuring apparatus according to the application example, the resolution is preferably equal to or less than the pixel interval.

  According to this application example, the reflection position of the object is specified by comparing the measurement luminance pattern distributed in the plurality of pixels constituting the measurement image with the plurality of reference luminance patterns prepared at a pitch equal to or less than the pixel interval. Therefore, the calculated reflection position coordinates of the object are detected with a finer precision than the pixel interval having luminance information. That is, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 13 In the control method of the shape measuring apparatus according to the application example described above, the coordinates indicating the peak point of the reference luminance pattern having the largest pattern similarity are used as the reflection position coordinates of the object.

  According to this application example, the luminance peak point of the reference luminance pattern that is most similar to the measurement luminance pattern distributed in the plurality of pixels constituting the measurement image is specified as the reflection position coordinates of the object. Since the detection is based on the luminance distribution, the reflection position coordinates of the object are detected with a finer accuracy than the pixel interval having the luminance information. Therefore, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 14 In the control method of the shape measurement apparatus according to the application example, the pattern similarity is a normalized cross-correlation value between the measurement luminance pattern and the reference luminance pattern, and the measurement luminance pattern and the plurality of reference luminance patterns When the distribution of normalized cross-correlation values obtained by the comparison is approximated by a quadratic function, the reflection position coordinates of the object are obtained from position information where the value of the quadratic function shows the maximum value.

  According to this application example, the pattern similarity is evaluated based on the normalized cross-correlation value between the measured luminance pattern and the reference luminance pattern. Therefore, the pattern shape similarity is evaluated without being affected by the difference in reflected luminance. . Further, by obtaining the reflection position coordinates based on the maximum value obtained by approximating the distribution of normalized cross-correlation values with a quadratic function, it is possible to detect the position with higher accuracy statistically. As a result, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 15 In the control method of the shape measuring apparatus according to the application example described above, the reference luminance pattern includes a pixel luminance pattern and a subpixel luminance pattern, and the pixel luminance pattern is a function in which one or a plurality of Gaussian functions are synthesized. The sub-pixel luminance pattern is derived from the pixel luminance pattern.

  According to this application example, it is possible to perform comparison for position detection with a pattern more similar to the measured luminance pattern based on the actual reflection characteristics. As a result, it is possible to detect the position with high accuracy with little error due to the object. As a result, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Application Example 16 In the control method of the shape measuring apparatus according to the application example described above, the reference luminance pattern includes a pixel luminance pattern and a sub-pixel luminance pattern, and the pixel luminance pattern is derived from the measured luminance pattern, and the sub-pixel luminance pattern Is derived from the pixel luminance pattern.

  According to this application example, the reference luminance pattern is derived based on the distribution information of the measured luminance pattern obtained from the reflected light of the actual object. Specifically, a pixel luminance pattern is generated using distribution information of a measured luminance pattern, instead of generating a similar pattern by combining a plurality of functions. In this way, even an object whose reflected light luminance distribution does not match Gaussian fitting can be detected with high accuracy with little error. Further, when it is difficult to synthesize a similar reference luminance pattern from a plurality of Gaussian functions, or when it takes time to synthesize, a reference luminance pattern can be created more easily.

  Application Example 17 A program according to this application example causes the shape measuring apparatus to function including the control method described above.

  According to this application example, by using the program including the control method described above to function, a shape measuring device capable of measuring with higher accuracy without limiting the measurement target due to the reflection characteristics is provided. can do. In addition, it is possible to provide a shape measuring apparatus that can adjust the time required for position detection within a necessary and sufficient range and can efficiently measure the position.

(A); Block diagram which shows the structure of the shape measuring apparatus concerning Embodiment 1, (b): The perspective view explaining the shape measuring method. (A): Explanatory drawing which shows the example of a measurement image, (b): Explanatory drawing which shows the example of the luminance distribution of the reflected light which the pixel has caught. Explanatory drawing which shows the example of the reflective luminance distribution which does not follow a single Gaussian distribution. (A): a graph showing an example of the measured luminance pattern Dm, (b): a graph showing an example of creating the distribution function Pm (y). (A): a graph showing a pixel luminance pattern, (b): a graph for explaining a sub-pixel luminance pattern. The graph which shows reference luminance pattern Ds-4-Ds4. Flow chart of shape measurement. 6 is a flowchart of a reflection position detection subroutine. (A), (b); The graph which shows the relationship between the measurement brightness | luminance pattern Dm and the reference brightness | luminance pattern Dsi. The graph which shows typically the example in which noise is contained in reflection luminance distribution. 10 is a flowchart of reflection position detection according to the second embodiment. The graph which shows distribution of a normalized cross correlation value.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Embodiments of the invention will be described below with reference to the drawings. In the following drawings, the scale may be different from the actual scale for easy understanding.

(Embodiment 1)
FIG. 1A is a block diagram illustrating a configuration of the shape measuring apparatus 100 according to the first embodiment, and FIG. 1B is a perspective view illustrating a shape measuring method.
The shape measuring device 100 is a non-contact type shape measuring device, and includes an optical scanning unit 10, an imaging unit 20, a reflection position detecting unit 30, and the like. When the slit light 1 emitted from the optical scanning unit 10 is projected onto the object 2 placed on the stage 14 and the object 2 is imaged from another angle by the imaging unit 20, a slit image deformed in accordance with the shape of the object 2 is obtained. It is done. The coordinates of the object 2 can be calculated geometrically based on the projection angle of the slit light 1 and the observation angle of each point of the deformed slit image. Furthermore, the three-dimensional information of the object 2 can be obtained by scanning the slit light 1 on the object 2.

  In FIG. 1B, the slit length direction of the slit light 1 is defined as the Xp direction, and the direction in which the slit light 1 is scanned on the object 2 while intersecting the Xp direction is defined as the Yp direction. The stage 14 is configured with an XpYp plane.

  The optical scanning unit 10 includes a light source 11, an optical system 12, a scan drive unit 13, and the like. The light source 11 irradiates the slit light 1 through the optical system 12. The scan drive unit 13 drives the optical system 12 to change the projection angle of the slit light 1. By changing the projection angle of the slit light 1, the entire object 2 is scanned with the slit light 1. The light source 11 is a laser light source, but may be an LED (Light Emitting Diode) light source. The scanning method is not driven by the optical system 12, the projection angle of the slit light 1 is fixed, the stage 14 is moved in the Y direction, and the object 2 placed on the stage 14 moves under the slit light 1. It may be a method of moving.

  The imaging unit 20 includes an imaging device 21, an optical system 22, an imaging drive unit 23, an ADC (Analog to Digital Converter) circuit 24, and the like. The imaging element 21 is an area sensor such as a CCD type or a MOS type, and captures light received through the optical system 22 as a measurement image under the control of the imaging drive unit 23. The measurement image is output from the ADC circuit 24 as information of the measurement luminance pattern Dm. The measured luminance pattern Dm is composed of luminance information received by each of a plurality of pixels constituting the imaging element 21. Each pixel is arranged in a matrix at a pixel pitch f.

  The reflection position detection unit 30 includes a luminance pattern analysis unit 31, a control unit 32, and the like. The luminance pattern analysis unit 31 analyzes the measurement luminance pattern Dm under the control of the control unit 32 and calculates the reflection position of each point on the object 2 of the slit light 1. The control unit 32 also controls the optical scanning unit 10 and the imaging unit 20, generates a reference pattern set PS described later, and outputs the shape measurement result of the object 2 based on the reflection position calculation result.

Next, a method for shape measurement by the shape measurement apparatus 100 will be described.
FIG. 2A is an explanatory diagram illustrating an example of a measurement image. The X axis and Y axis in FIG. 2A are the coordinates of the pixels constituting the measurement image, and correspond to the directions of Xp and Yp in FIG.

  The polygonal line drawn in FIG. 2A is an example of a slit image (bright line 3) in which the linear slit light 1 is projected onto the object 2 and transformed into a trapezoidal shape in accordance with the shape of the object 2. From the position on the Y axis of the bright line 3, the incident angle (observation angle) of the slit light 1 that is reflected by the object 2 and incident on the imaging means 20 can be obtained. From this observation angle, the projection angle of the slit light 1 that is known information, and the distance between the projection base point and the reference point on the stage 14, the height of the reflection point of the object 2 can be obtained by the triangulation method. .

In order to accurately measure the height of the object 2, the value of the position (y = y ₀ ) of the bright line 3 in FIG. 2A is accurately calculated, and the position corresponding to y ₀ is specified with high accuracy. There is a need. However, the bright line 3 is generally obtained as image information having a certain width. When attention is paid to the approximate center (x, y) = (x ₀ , y ₀ ) of the bright line 3, the bright line 3 at x ₀ is discrete for each pixel having a width in the Y direction, with y ₀ being substantially the center. Therefore, it is necessary to specify the position of y ₀ corresponding to the actual position within this width range. This will be specifically described below.

Figure 2 (b) is a diagram showing an example of a luminance distribution of reflected light pixels in the vicinity of y ₀ at x ₀ line is caught, show how the bright line 3 has a distribution of brightness in the width direction Yes. 2B expands the Y-axis scale to the width of the bright line 3 with respect to the Y-axis scale of FIG.
In FIG. 2B, the bright lines 3 are distributed as a measured luminance pattern Dm having a width of the pixels n-4 to n + 5. The measurement luminance pattern Dm is composed of discrete data of luminance information distributed to each pixel. Therefore, it is desired that the actual reflection position of the object 2 is accurately calculated and specified from the discrete data of luminance.

When the slit light 1 is projected, the brightness of the light reflected by the object 2 generally follows a Gaussian distribution in many cases, and the peak point of the Gaussian distribution often corresponds to the reflection position of the object 2. A curve indicated by an alternate long and short dash line in FIG. 2B shows a distribution when the discrete data of each pixel is Gaussian fitted. As can be seen from this figure, the position y _p of the pixel n are obtained the maximum luminance, there is deviation of e _s between the peak point y ₀ of the Gaussian distribution corresponding to the actual reflection position, measuring brightness This shows that the position can be specified with high accuracy by Gaussian fitting the pattern Dm. In other words, when the actual luminance distribution follows a Gaussian distribution, it is possible to detect a position with high accuracy from discrete data by calculating the peak point of the fitted Gaussian distribution.

  However, in the case of an object whose luminance distribution of reflected light is not suitable for Gaussian fitting, the peak point of the Gaussian distribution that is footed may not indicate the actual reflection position. That is, depending on the reflection characteristics of the object, the measurement accuracy may deteriorate. In addition, every time measurement is performed, Gaussian fitting is performed on pixel information (discrete data), which requires time. Therefore, according to the present embodiment, the measurement target is not limited by the reflection characteristics, can be measured with higher accuracy, and does not require a Gaussian fitting process each time, and can be measured in a short time. Propose. The method will be described below.

FIG. 3 is an explanatory diagram illustrating an example of a reflected luminance distribution that does not follow a single Gaussian distribution.
When the surface of the object 2 is translucent or when the surface is coated, the reflection with respect to the projection of the slit light 1 is not only the reflection from the surface of the object 2 but also the deep layer (under the surface of the object 2). ) May be combined. Further, in a substance such as marble, in addition to surface reflection, subsurface scattering (subsurface scattering) is observed.
In FIG. 3, a distribution P <b> 1 indicates a luminance distribution due to reflection from the surface, and a distribution P <b> 2 indicates a luminance distribution due to reflection from below the surface of the object 2. In actual measurement, a distribution Pm in which these reflections are combined is observed. Peak point of the distribution Pm (y = y _0), can cause measurement errors may shift amount e _r between substantially equal peak point of the distribution P1 to the actual reflection position (y = y _r). A method for measuring an object exhibiting a distribution that does not follow such a single Gaussian distribution will be described below as an example.

Example 1
1. Outline of Measurement An outline of the measurement method according to this embodiment will be described.
FIG. 4A is a graph showing an example of the measured luminance pattern Dm, and FIG. 4B is a graph showing an example of creating a distribution function Pm (y) for fitting to the measured luminance pattern Dm.
First, a luminance distribution due to reflection from the object 2 is measured in advance, and a measured luminance pattern Dm is acquired. Next, a distribution function Pm (y) for fitting to the measured luminance pattern Dm is obtained by synthesizing one or a plurality of Gauss functions. At this time, information on the peak point y _{0 of the} distribution function Pm (y) and the reflection position y _r of the object 2 is specified together with the distribution function Pm (y). The reflection position y _r of the object 2 is obtained by extracting a Gauss function related to the surface reflection portion from the Gauss function used for the synthesis and setting the peak value to y _r or actually performing other physical measurements. There is a method to ask for.

Next, from this distribution function Pm (y), a plurality of position detection reference luminance patterns Dsi (i =...) Having position information for each sub-pixel resolution δy exceeding the pixel resolution (pixel pitch f). -3, -2, -1, 0, 1, 2, 3, ...). The reference luminance pattern Dsi is a collection of discrete data that the pixels arranged with the pixel pitch f have on the distribution function Pm (y).
Specifically, first, obtains the reference brightness pattern Ds0 when one pixel is overlapped with the reflection position y _0, and shifted by resolution δy then the reflection position y ₀ the extent necessary to plus or minus on the Y axis A plurality of reference luminance patterns Dsi are generated. Each of the reference luminance patterns Dsi has the positional information of the displacement amount from the reflection position y _0.
The plurality of reference luminance patterns Dsi are generated in advance for each type of object having different reflection characteristics, stored as a reference pattern set PS for each type of object, and correspond to that object in measuring the shape of the actual object. It is preferable to select and use the reference pattern set PS to be used.

Next, a correlation value between the measured luminance pattern Dm obtained by measurement and each reference luminance pattern Dsi is calculated. Here, the reflection position of the object 2 can be specified from the position information (deviation amount from the reflection position y ₀ ) of the reference luminance pattern Dsi with the highest correlation value.
In this embodiment, the normalized cross-correlation value R is used as the correlation value. Normalized cross correlation (NCC) is an absolute measure normalized by the mean and standard deviation, and the closer the value is to 1, the higher the similarity.

Details of the present embodiment will be described below.
2. Preparation In the preparation stage, a reference luminance pattern Dsi is created to prepare a reference pattern set PS.
First, prior to measuring the shape, the reflection luminance distribution of the object 2 is measured to check the reflection characteristics.
FIG. 4A shows the measured luminance pattern Dm obtained by actual measurement.
Next, a distribution function Pm (y) that substantially fits the distribution of the measured luminance pattern Dm is created. The distribution function Pm (y) is created by synthesizing a plurality of Gaussian functions as shown in Equation 1 below. The constants C ₁ , C ₂ ,..., Σ ₁ , σ ₂ ,..., Y ₁ , y ₂ , etc. are appropriately set so as to substantially fit the distribution of the measured luminance pattern Dm.

FIG. 4B shows a state in which the distribution function Pm (y) is created by combining the Gaussian functions P1 and P2 and the fitting with the measured luminance pattern Dm is substantially confirmed. Further, from the peak point of the Gaussian function P1, seeking reflection position y _r of the object 2 in the distribution function Pm (y).
The distribution function Pm (y), the peak point y _{0 of the} distribution function Pm (y), the reflection position y _r , and the deviation amount e _r = y ₀ −y _r obtained here are used only for the object 2 or A reference luminance distribution function dedicated to an object group exhibiting equivalent reflection characteristics and reflection position information in this distribution are stored in an external storage device or the like.

  The creation of the distribution function Pm (y) is not limited to the method of combining a plurality of Gaussian functions as described above. For example, the discrete data of the measured measurement luminance pattern Dm is used as it is, and the discrete data A method of interpolating each point with a spline function or the like may be used.

Next, a reference luminance pattern Dsi is created from the distribution function Pm (y).
The reference luminance pattern Dsi is a set of luminance distribution data having position information for evaluating the correlation by comparing with the measured luminance pattern Dm obtained by imaging the object 2 and specifying the reflection position. The reference luminance pattern Dsi is composed of a pixel luminance pattern (reference luminance pattern Ds0) and other sub-pixel luminance patterns, and the generation methods are different.

  FIG. 5A is a graph showing a pixel luminance pattern, and shows the distribution function Pm (y) and the value of the distribution function Pm (y) at each pixel position. In this graph, the luminance distribution data of each point (d1 to d13) indicated by a circle is a pixel luminance pattern.

First, the peak point y ₀ in the distribution function Pm (y) is shifted so as to overlap the position of the pixel n, and a basic reference luminance pattern Ds 0 is created as a pixel luminance pattern. This corresponds to d1 to d13 in FIG. 5A, and is obtained from the intersection point on the graph as the value of the distribution function Pm (y) at each position from the pixel n-5 to the pixel n + 7.

Next, a reference luminance pattern Dsi is created as a subpixel luminance pattern.
The sub-pixel luminance pattern is a reference luminance pattern Dsi for detecting the position with the sub-pixel resolution δy, and is created by shifting the reference luminance pattern Ds0 by the sub-pixel resolution δy in the pixel direction around the position of the pixel n. . That is, the peak point y ₀ of the distribution function Pm (y) is shifted from the position of the pixel n in the Y direction by a pitch corresponding to the desired resolution δy, and the distribution function Pm (y) corresponding to each pixel position is changed. A reference luminance pattern Dsi is created using values as a data set.

  FIG. 5B shows a specific example of the subpixel luminance pattern. Here, one reference luminance pattern Dsi is created to set the resolution δy to a quarter of the pixel pitch. The peak point of the distribution function Pm (y) is shifted by -1/4 pixel in the Y direction, and the value of the distribution function Pm (y) at each position from the pixel n-5 to the pixel n + 7 is circled at the intersection on the graph. Is shown. This reference mark pattern Ds-1 is obtained when the data set (d1 to d13) at the circle positions overlaps the -1/4 pixel position.

Similarly, the reference luminance pattern Dsi (= Ds−4 to Ds4) is generated from the −4/4 pixel position to the +4/4 pixel position at a 1/4 pixel pitch (= resolution δy).
FIG. 6 is a graph showing the reference luminance patterns Ds-4 to Ds4 (except for Ds0) obtained in this way. This set of reference luminance patterns Ds-4~Ds4 is a reference pattern set PS resolution δy quarter-pixel (sub-pixel), the reference pattern set and y _0, y _r, stores the value of e _r To complete the preparation.

3. Measurement FIG. 7 shows a flowchart of shape measurement according to this embodiment. The shape measurement method will be specifically described with reference to this flowchart.
The shape measuring apparatus 100 includes a program for causing a function including a control method for shape measurement according to the flowchart.

First, the object 2 is set on the stage 14 (FIG. 1B) (step SA1).
Next, a set of reference luminance patterns Ds-4 to Ds4 prepared for the object 2 is selected as a reference pattern set PS for measurement (step SA2). Also, it should read the value of y _0, y _r, e _r corresponding simultaneously.
Next, the positions of the optical scanning means 10 and the stage 14 are adjusted, and the distance between the projection base point and the reference point on the stage 14 is obtained. Alternatively, the position is adjusted to a preset distance (step SA3).
Next, the slit light 1 is projected from the optical scanning unit 10 (step SA4), and a slit image is captured by the imaging unit 20 (step SA5). As a result, for example, the image of FIG. 2A is obtained.

Next, processing for detecting a reflection position from the captured slit image is performed.
First, X = 0 is set (step SA6), and the reflection position at X = 0 is detected (step SA7). The reflection position detection step SA7 will be described later as a subroutine.

Next, it is confirmed whether or not the value of X has reached a predetermined value corresponding to the length of one necessary line (slit light 1), and it is determined whether or not the necessary one line has been processed (step SA8). If the processing has not been completed (No), X is incremented and the pixel coordinates X are indexed in the X direction (step SA9), and the reflection position at the next X position is detected (step SA7). Step SA7 to step SA9 are repeated until the necessary one line processing is completed.
When the process is finished (Yes in step SA8), it is confirmed whether the optical scanning range is finished and the measurement can be finished (step SA10). When the process is not finished (No), the optical scanning unit 10 performs optical measurement. The scanning is indexed (step SA11), and the next slit light 1 is projected (step SA4).

  Step SA4 to step SA11 are repeated until the necessary optical scanning is completed. When the optical scanning is completed (Yes in Step SA10), the shape measurement result of the object 2 is output (Step SA12), and the measurement is finished.

FIG. 8 is a flowchart of the reflection position detection subroutine (step SA7).
FIGS. 9A and 9B are graphs showing the relationship between the measured luminance pattern Dm and the reference luminance pattern Dsi. The method for detecting the reflection position will be specifically described with reference to these drawings.

  First, a pixel is scanned in the Y direction at the pixel coordinate X selected in step SA6 or step SA9, and a pixel having the maximum luminance is detected as a reflection position candidate coordinate (step SB1). For example, in FIG. 9A, the pixel n is a pixel having the maximum luminance.

  Next, the entire reference pattern set PS is adjusted to the position of the pixel n (step SB2). Specifically, as shown in FIG. 9A, the reference luminance pattern Ds0 is shifted in the Y direction so that the position of the maximum luminance matches the position of the pixel n. Similarly, the Y coordinate value of the entire reference pattern set PS (reference luminance patterns Ds-4 to Ds4) is shifted in the Y direction by the same amount.

  Next, the measured luminance pattern Dm and the reference luminance pattern Ds-4 are compared (step SB3). Specifically, a normalized cross-correlation value R is calculated, and the degree of correlation between the measured luminance pattern Dm and the reference luminance pattern Ds-4 is quantified. Next, the measured luminance pattern Dm and the reference luminance pattern Ds-3 are compared, and the correlation value is similarly calculated (steps SB5 and SB3). The processing up to the reference luminance pattern Ds4 is repeated, and it is confirmed whether the processing up to the reference luminance pattern Ds4 is completed (step SB4). If completed, the reference luminance pattern Dsi having the largest correlation value is specified (step SB6).

FIG. 9B shows the relationship between the measured luminance pattern Dm having the largest correlation value and the reference luminance pattern Ds2. It can be seen that the respective data are substantially the same, and the normalized cross-correlation value R shows a maximum value of about 1. As a result, when the resolution δy is a quarter of the pixel pitch, it is detected that the peak position y ₀ of the reflection distribution is at an intermediate position (+2/4 pixel position) between the pixel n and the pixel n + 1. by the reflection position y _r from the deviation amount e _r is identified, the processing of the reflection position detection is completed.

  In the present embodiment, the reference pattern set PS is prepared so that the resolution δy is a quarter of the pixel pitch, and the comparison range is set to the range of two pixels before and after. However, the present invention is not limited to this. In order to obtain higher measurement accuracy, the reference pattern set PS may be prepared with a finer pitch, for example, a sufficient pitch, and each correlation evaluation may be performed within a necessary range.

  As described above, according to the shape measuring apparatus 100 according to the present embodiment, the method for controlling the shape measuring apparatus, and the program that includes and functions the control method, the following effects can be obtained.

  Rather than simply recognizing the pixel position indicating the maximum luminance of the acquired measurement image as the reflection position of the object 2, the object 2 is obtained by comparing the measurement luminance pattern Dm distributed over a plurality of pixels with the reference luminance pattern Dsi. The reflection position coordinates of are obtained. By applying this method, it is possible to prepare a reference luminance pattern Dsi for comparison having a luminance distribution according to the reflection characteristic of the object 2 in advance, and by comparing the luminance distribution patterns, more accurate High measurement is possible.

  In addition, the normalized cross-correlation value R between the measured luminance pattern Dm and the reference luminance pattern Dsi distributed over a plurality of pixels is calculated, and the magnitude of the calculated normalized cross-correlation value R is compared, whereby the reflection position of the object 2 is calculated. Coordinates can be obtained. The number of data used for the calculation is discrete data in sub-pixel units, and the number of data is small compared to the process of performing Gaussian fitting each time, so the time required for the calculation is short and high-speed measurement is possible. .

  In addition, a reference pattern set PS having a corresponding distribution characteristic is prepared in advance according to reflection characteristics that vary depending on the type of the object 2, and an optimum comparison reference pattern set PS is selected for measurement. High measurement is possible. Specifically, for example, when the reflection from the object 2 includes not only the surface of the object 2 but also the reflection from the lower layer, the distribution characteristic is not a distribution approximated by a single Gaussian distribution. The distribution is specific to the object 2. By preparing a reference pattern set PS in accordance with the distribution unique to the object 2 in advance, it is possible to perform comparison with which more accurate pattern matching is obtained, and thus it is possible to perform measurement with higher accuracy.

In addition, by preparing a reference pattern set PS having reflection characteristics corresponding to various objects assumed as measurement targets in advance, the accuracy of various objects can be improved simply by selecting an optimal reference pattern set PS for measurement. Good measurement is possible.
Therefore, it is possible to provide a shape measuring apparatus capable of measuring with higher accuracy without limiting the measurement target due to the reflection characteristics.

  In addition, since the resolution δy can be set prior to measurement, the time required for comparison can be adjusted within a necessary and sufficient range. Specifically, for example, in the case of measurement that does not require so much measurement accuracy, measurement is performed with a finer resolution δy by setting the resolution δy to the pixel pitch f or about half of the pixel pitch f. Compared to the case, the time required for measurement can be shortened.

  In addition, since a coordinate range for pattern comparison prior to measurement can be set, the time required for comparison can be adjusted within a necessary and sufficient range. Specifically, in setting the comparison coordinate range, if the range to be compared before and after the calculated reflection position candidate coordinates (pixels with the maximum brightness) is narrowed, the number of pattern comparisons decreases, so measurement Time is shortened.

  The reflection position of the object 2 is specified by comparing the measurement luminance pattern Dm distributed in a plurality of pixels constituting the measurement image with a plurality of reference luminance patterns Dsi prepared at intervals equal to or less than the pixel pitch f. Therefore, the calculated reflection position coordinates of the object 2 are detected with a finer precision than the pixel pitch f having luminance information.

  Further, the reflection position coordinates of the object 2 are specified from the luminance peak point of the reference luminance pattern Dsi that is most similar to the measured luminance pattern Dm distributed in a plurality of pixels constituting the measurement image. Since the detection is based on the luminance distribution, the reflection position coordinates of the object 2 are detected with a finer accuracy than the pixel pitch f having luminance information. Therefore, it is possible to provide a shape measuring device capable of measuring with higher accuracy.

  Further, since the pattern similarity is evaluated based on the normalized cross-correlation value R between the measured luminance pattern Dm and the reference luminance pattern Dsi, the pattern shape similarity is evaluated without being affected by the difference in reflected luminance.

  Further, by combining the reference luminance pattern Dsi from one or a plurality of Gaussian functions, it is possible to make a comparison for position detection with a pattern more similar to the measured luminance pattern Dm based on actual reflection characteristics. As a result, it is possible to detect the position with high accuracy with little error due to the object.

  Further, since the reference luminance pattern Dsi can be created using the distribution information of the measured luminance pattern Dm, even if the luminance distribution of reflected light is not suitable for Gaussian fitting, there is little error. Accurate position detection is possible. Further, when it is difficult to synthesize a similar reference luminance pattern Dsi from a plurality of Gaussian functions, or when it takes time to synthesize, the reference luminance pattern Dsi can be created more easily.

(Embodiment 2)
Next, the shape measuring apparatus 100 according to the second embodiment and a method for controlling the shape measuring apparatus will be described. In the description, the same components as those in the first embodiment are denoted by the same reference numerals, and redundant description is omitted.

  In actual measurement, the luminance distribution of reflected light may include noise due to disturbance light, the roughness of the reflecting surface, and the presence of foreign matter. Here, an example of measurement when noise is included will be described as a second embodiment.

FIG. 10 is a graph schematically illustrating an example in which noise is included in the reflected luminance distribution.
In the reflected light distribution Pm ′, there is noise in a portion corresponding to the pixel n and the pixel n + 1, and a measured luminance pattern Dm ′ in which the luminance of each pixel is reduced is obtained. In the first embodiment (Example 1), as shown in FIG. 9A, the pixel n indicating the maximum luminance is selected as the reflection position candidate coordinates, but in this example, the pixel n + 2 is used as the reflection position candidate coordinates. The position of the maximum luminance of the selected reference luminance pattern Ds0 is shifted in the Y direction so as to coincide with the position of the pixel n + 2 as shown in FIG. As described above, when the reflected light includes noise, the reflection position candidate coordinates may be detected with a deviation from the original position. As is clear from FIG. 10, the reflection position is detected at a position away from the actual reflection position.

  In such a case, that is, when it is assumed that the reflection position candidate coordinates are deviated pixel by pixel due to the influence of noise, the selection of the reflection position candidate coordinates is not performed by the pixel showing the maximum luminance, but a plurality of This is performed by evaluating the correlation between the pixel luminance pattern and the measured luminance pattern Dm ′. Specifically, a plurality of reference luminance patterns Ds0 obtained by shifting the position of the maximum luminance of the reference luminance pattern Ds0 to a plurality of pixel positions before and after the pixel n + 2 indicating the maximum luminance, and the measured luminance pattern Dm ′. Correlation evaluation is performed, and the pixel having the highest correlation value is set as a reflection position candidate coordinate.

FIG. 11 shows a flowchart of a reflection position detection subroutine according to the present embodiment. This embodiment is different only in the subroutine part (step SA7) of reflection position detection in the shape measurement flowchart (FIG. 7) of the first embodiment. A method for detecting the reflection position will be specifically described with reference to the flowchart of FIG.
The shape measuring apparatus 100 includes a program for causing a function including a control method for shape measurement according to the flowchart.

First, a plurality of reference luminance patterns Ds0 are created (step SC1). Specifically, when the comparison range is two pixels before and after the pixel n, five pixels obtained by shifting the position of the maximum luminance of the reference luminance pattern Ds0 to five pixel positions from the pixel n to the pixel n + 4. Reference luminance patterns Ds01 to Ds05 are generated.
Next, a correlation value between each reference luminance pattern Ds0i and the measured luminance pattern Dm ′ is calculated (steps SC2 to SC4).
Next, the pixel having the largest correlation value as a result of the calculation is set as a reflection position candidate coordinate (step SC5). In this way, a pixel closer to the actual reflection position can be selected as the reflection position candidate coordinates.

  In the subsequent flow, similar to Steps SB2 to SB6 of the first embodiment, the correlation value is compared and evaluated by applying the reference pattern set PS prepared by sub-pixels with the pixel selected as the reflection position candidate coordinates as the center. The reflection position is specified from the reference luminance pattern Dsi from which the maximum correlation value is obtained.

  In the above example, the pixels as the reflection position candidate coordinates are obtained using the reference luminance patterns Ds01 to Ds05 obtained from the five pixel positions from the pixel n-1 to the pixel n + 3. May be determined according to the magnitude of noise.

  Further, when it is assumed that the reflection position candidate coordinates are shifted in units of pixels due to the influence of noise as described above, the range of the reference pattern set PS used in the method of the first embodiment is not limited to the above method. Alternatively, a method of preparing in a range over a plurality of pixels and evaluating a correlation value with all reference luminance patterns Dsi may be used. However, in this case, it is necessary to consider that the measurement time increases because the number of comparison processes in the subpixel increases.

  As described above, according to the shape measuring apparatus 100 according to the present embodiment, the method for controlling the shape measuring apparatus, and the program that includes and operates the control method, the presence of disturbance light, the roughness of the reflecting surface, foreign matter, and the like. Even when noise is included in the luminance distribution of the reflected light, the reflection position can be specified with high accuracy by comparing and evaluating the correlation value between the measurement luminance distribution and the reference pattern set PS.

  Note that the present invention is not limited to the above-described embodiment, and various modifications and improvements can be added to the above-described embodiment. A modification will be described below.

(Modification 1)
In the first embodiment, the reference luminance pattern Dsi having the largest correlation value obtained by comparing the measured luminance pattern Dm and the plurality of reference luminance patterns Dsi is specified, and the position of the reference luminance pattern Dsi and the associated y _0, y _r, had sought reflection position from the information in e _r. In contrast, the present modified example, when approximating the distribution of the correlation value by a quadratic function, y ₀ the value of the quadratic function is location information and associated indicating the maximum value, y _r, from e _r, the object 2 It is characterized in that the coordinates of the reflection position are obtained. This will be specifically described below.

FIG. 12 is a graph showing the distribution of the calculated normalized cross-correlation value R.
In step SA7 (steps SB1 to SB6) of reflection position detection described above, the normalized cross-correlation value R is calculated by shifting the resolution by δy, the horizontal axis represents the shift amount of the resolution δy, and the vertical axis represents the normalized cross-correlation value R. Is plotted. By approximating and connecting the plotted points with a quadratic function, a point where the quadratic function has the maximum value is derived.

  In Embodiment 1 (Example 1), the reflection position coordinates are specified from the position of the reference luminance pattern Dsi that is the shift amount of the resolution δy indicating the maximum value of the calculated normalized cross-correlation value R. A more accurate result can be obtained by specifying the reflection position coordinates from the position information in which the value of is the maximum value. By obtaining the reflection position coordinates based on the maximum value obtained by approximating the distribution of the normalized cross-correlation value R with a quadratic function, it is possible to detect the position with higher accuracy statistically.

  DESCRIPTION OF SYMBOLS 1 ... Slit light, 2 ... Object, 3 ... Bright line, 10 ... Optical scanning means, 11 ... Light source, 12, 22 ... Optical system, 13 ... Scan drive part, 14 ... Stage, 20 ... Imaging means, 21 ... Imaging element, DESCRIPTION OF SYMBOLS 23 ... Imaging drive part, 24 ... ADC circuit, 30 ... Reflection position detection means, 31 ... Luminance pattern analysis part, 32 ... Control part, 100 ... Shape measuring device, Dm ... Measurement brightness pattern, Dsi ... Reference brightness pattern, PS ... Reference pattern set.

Claims (14)

Optical scanning means for scanning an object with projection light;
Imaging means for acquiring a measurement image from the reflected light of the projection light from the object;
A reflection position detecting means for calculating a reflection position of the object based on the acquired measurement image;
The reflection position detection means calculates a reflection position candidate coordinate of the object from a measurement luminance pattern distributed to a plurality of pixels constituting the acquired measurement image,
A plurality of prepared brightness intervals and a plurality of resolution intervals that are within a range of at least one-quarter and less than a quarter of the pitch of the plurality of pixels within a comparison coordinate range including the reflection position candidate coordinates. by comparing the reference brightness pattern, calculates a pattern similarity the measuring brightness pattern and Ri by normalized cross-correlation value between each of the reference luminance patterns,
When approximating the distribution before Symbol normalized cross-correlation value by a quadratic function, the position information value of the quadratic function is a maximum value, the shape measuring apparatus and obtaining the reflection position coordinates of the object . The plurality of reference luminance patterns are provided in advance as a plurality of reference pattern sets according to the distribution characteristics of the reflected light of the plurality of types of the objects,
The shape measuring apparatus according to claim 1, wherein, in the measurement, the corresponding reference pattern set can be selected according to the type of the object.   The shape measuring apparatus according to claim 1, wherein the resolution and the comparison coordinate range can be set in advance in measurement.   4. The shape measurement according to claim 1, wherein a coordinate indicating a peak point of the reference luminance pattern having the largest pattern similarity is a reflection position coordinate of the object. apparatus. The reference luminance pattern includes a pixel luminance pattern and a sub-pixel luminance pattern,
The pixel luminance pattern is derived from a function in which one or more Gaussian functions are combined,
The shape measurement apparatus according to claim 1, wherein the sub-pixel luminance pattern is derived from the pixel luminance pattern. The reference luminance pattern includes a pixel luminance pattern and a sub-pixel luminance pattern,
The pixel luminance pattern is derived from the measured luminance pattern;
The shape measurement apparatus according to claim 1, wherein the sub-pixel luminance pattern is derived from the pixel luminance pattern. Scanning an object with projected light;
Obtaining a measurement image by reflected light of the projection light from the object;
Calculating the reflection position candidate coordinates of the object from the measurement luminance pattern distributed to a plurality of pixels constituting the acquired measurement image;
A plurality of prepared brightness intervals and a plurality of resolution intervals that are within a range of at least one-quarter and less than a quarter of the pitch of the plurality of pixels within a comparison coordinate range including the reflection position candidate coordinates. Comparing a reference luminance pattern and calculating a pattern similarity between the measured luminance pattern and each of the reference luminance patterns;
Obtaining the reflection position coordinates of the object from the coordinate information specified by the reference luminance pattern having the largest pattern similarity, and
The pattern similarity is a normalized cross-correlation value between the measured luminance pattern and the reference luminance pattern,
Reflecting the position coordinates of the object, when approximating the distribution of the normalized cross-correlation value obtained by comparison with the previous SL measured brightness pattern and the plurality of reference brightness patterns in a quadratic function, the value of the quadratic function control method for shape measurement apparatus but, wherein the mel position information or RaMotomu indicating the maximum value. A step of preparing the plurality of reference luminance patterns as a plurality of reference pattern sets in accordance with distribution characteristics of the reflected light of the plurality of types of objects,
The method for controlling a shape measuring apparatus according to claim 7, further comprising a step of selecting the corresponding reference pattern set according to the type of the object in measurement.   The method for controlling a shape measuring apparatus according to claim 7 or 8, wherein the resolution and the comparison coordinate range are preset in measurement.   The method for controlling a shape measuring apparatus according to claim 7, wherein the resolution is equal to or less than the pixel interval.   11. The shape measurement according to claim 7, wherein a coordinate indicating a peak point of the reference luminance pattern having the largest pattern similarity is a reflection position coordinate of the object. Control method of the device. The reference luminance pattern includes a pixel luminance pattern and a sub-pixel luminance pattern,
The pixel luminance pattern is derived from a function in which one or more Gaussian functions are combined,
12. The shape measuring apparatus control method according to claim 7, wherein the sub-pixel luminance pattern is derived from the pixel luminance pattern. The reference luminance pattern includes a pixel luminance pattern and a sub-pixel luminance pattern,
The pixel luminance pattern is derived from the measured luminance pattern;
12. The shape measuring apparatus control method according to claim 7, wherein the sub-pixel luminance pattern is derived from the pixel luminance pattern.   A program for causing a shape measuring device to function including the control method according to any one of claims 7 to 13.

Images