Measurement device, computer program, system, and article manufacturing method

Abstract

To provide a measuring device that prevents an error caused by an image used for measurement to allow accurate measurement.SOLUTION: A measuring device receives light from a measuring object by an image pickup device, and calculates a measured value related to the measuring object by using the cross-correlation function of two images acquired by the image pickup device. The measuring device has correction means that corrects the measured value according to the ratio of, for example, peak spreading and a low frequency component corresponding to the configuration of spatial frequency components of the two images.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a measuring device, a computer program, a system, and a method for manufacturing an article. [Background technology]

[0002] A conventional non-contact measurement device is disclosed in Patent Document 1. This measurement device irradiates a measurement object with a laser to generate speckles, photoelectrically converts the speckle distributions before and after movement to obtain two image signals, and calculates the amount of deformation of the measurement object based on the positions of the extreme values ​​of the correlation function between the two. Although Patent Document 1 performs measurement using speckles, similar measurements are also possible using ordinary image information using incoherent illumination.

[0003] In Patent Document 2, in the displacement measuring device, the optical magnification is determined according to the amount of displacement measured in order to improve accuracy. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Special Publication No. 59-52963 [Patent Document 2] Japanese Patent Application Laid-Open No. 2003-222504 Summary of the Invention [Problem to be solved by the invention]

[0005] Patent Document 2 discloses a method for correcting distortion caused by the light-receiving optical system of a displacement measurement device. In a light-receiving optical system with large distortion, the effect of distortion varies depending on the amount of displacement. When the amount of displacement is small, the optical magnification is equivalent to the design value, but as the amount of displacement increases, the deviation between the optical magnification and the design value increases. Therefore, by measuring the amount of distortion in the optical system in advance, the optical magnification is corrected according to the amount of displacement.

[0006] However, errors that occur in displacement measuring devices are not limited to those caused by optical magnification such as distortion, and the method of correcting optical magnification as in Patent Document 2 has the problem that it is not possible to sufficiently correct other errors.

[0007] Therefore, an object of the present invention is to provide a measuring device or the like that can suppress errors caused by images used in measurement and perform highly accurate measurements. [Means for solving the problem]

[0008] In order to achieve this object, a measurement device according to one aspect of the present invention comprises: A measuring device for measuring a measurement object, The aforementioned Light from the object to be measured is received by the image sensor To do Using the cross-correlation function of two images acquired by 、 The measurement value for the object to be measured is a processing unit for obtaining The processing unit calculates the spread of the peak shape of the cross-correlation function. The measured value is corrected according to the Ruko It is characterized by the following. [Effects of the Invention]

[0009] According to the present invention, it is possible to provide a measuring device or the like that is capable of suppressing errors caused by images used in measurement and performing highly accurate measurements. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 2 is a configuration diagram showing an example of a length measuring meter as a measuring device according to the first embodiment. [Figure 2] 4 is a flowchart showing an example of a measurement flow of the length measuring meter of the first embodiment. [Figure 3] 3 is a flowchart showing a specific example of the displacement amount calculation in step S14 of FIG. 2. [Figure 4]10A is a diagram showing the surface dependency of measurement error data when measuring a metal sample and a paper sample as the measurement object 2, and FIG. 10B is a diagram showing the line sensor output data of the metal sample and the paper sample. [Figure 5] 10A is a diagram showing the WD dependency of measurement error data when measuring a metal sample as the measurement object 2, and FIG. 10B is a diagram showing line sensor output data when the WD is at the center and at the edge. [Figure 6] 10A is a diagram showing the speed dependency of the measurement error data on the measurement object 2, and FIG. 10B is a diagram showing the line sensor output data when the speed of the measurement object 2 is different. [Figure 7] FIG. 10 is a schematic diagram for explaining the occurrence of sub-pixel estimation errors. [Figure 8] FIG. 8(A) is a diagram showing the cross-correlation function when high frequency components are dominant as the composition of spatial frequency components contained in the line sensor output, and FIG. 8(B) is a diagram showing the cross-correlation function when low frequency components are dominant as the composition of spatial frequency components contained in the line sensor output. [Figure 9] 10 is a graph plotting the relationship between the spread of the peak shape of the cross-correlation function and the length measurement error, based on the results of measurements performed under various measurement conditions. [Figure 10] FIG. 10 is a schematic diagram for explaining quadratic function fitting used in sub-pixel estimation calculation. [Figure 11] (A) is a diagram showing the measurement error when no correction is performed under various conditions, (B) is a diagram showing the correction effect of embodiment 1, and (C) is a diagram showing the measurement error when correction is performed under the same conditions. [Figure 12] FIG. 10 is a diagram illustrating an example in which linear function fitting is used as a method for calculating sub-pixel estimation of a cross-correlation function and the spread of a peak shape. [Figure 13] FIG. 10 is a diagram showing a control system including a measurement device and a robot arm according to a third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the present invention is not limited to the following embodiments. In each drawing, the same members or elements are designated by the same reference numerals, and duplicate descriptions will be omitted or simplified.

[0012] <Embodiment 1> The inventors have found that measurement errors tend to increase when low-frequency components are dominant in the spatial frequency components contained in the image used. For example, when measuring objects with significantly different surface roughness, measurement errors increase when low-frequency components are dominant in the spatial frequency components of the surface pattern of the objects.

[0013] Furthermore, measurement errors also increase when, for example, the distance to the object being measured changes, causing the image to become blurred, or when the exposure time is constant and the object's speed increases, causing the image to shake. In other words, measurement errors increase when the proportion of frequency components lower than a predetermined frequency in the distribution of spatial frequency components in the image of the object being measured is equal to or greater than a predetermined proportion. Therefore, it has been discovered that it is desirable to correct the measurement values ​​in such cases. Therefore, the measurement device 1 of the first embodiment is configured to correct the measurement error in accordance with the configuration of the spatial frequency components.

[0014] 1 is a configuration diagram showing an example of a length measuring meter (displacement measuring device) as a measuring device 1 of embodiment 1. The measuring device 1 of this embodiment measures the displacement of a measuring object 2 arranged opposite the measuring device 1 in a non-contact manner. The measuring object 2 moves in the direction of the arrow in the figure. The light beam emitted from the light source 3 is focused onto the measurement object 2 by the light focusing member 4, and illuminates the measurement object 2.

[0015] The light source 3 can be appropriately selected from a laser diode, an LED, a halogen lamp, etc. When a laser diode is selected, the image obtained will be an image composed of speckles, and when an incoherent light source such as an LED or a halogen lamp is selected, an image that reflects the pattern on the surface of the measurement object 2 will be obtained.

[0016] The focusing element 4 is composed of a single lens or a group of lenses. When using a laser diode, it is desirable to perform aberration correction so that the light can be focused as a plane wave. Furthermore, if the distance between the measurement device 1 and the object 2 to be measured changes, oblique incidence illumination will cause speckle displacement, so it is desirable to configure coaxial epi-illumination. On the other hand, when an incoherent light source such as an LED or halogen lamp is selected, it is sufficient to illuminate the light-receiving area, and aberrations and the like are not particularly important, so the light source can be selected appropriately depending on the size of the area to be illuminated.

[0017] A portion of the light beam diffusely reflected from the illuminated measurement object 2 is collected onto an image sensor, such as a sensor 6, via a light receiving optical system composed of a light collecting member 5, an aperture stop 7, and a light collecting member 8. In this embodiment, a double-telecentric optical system is used as the light receiving optical system. The light-collecting members 5 and 8 are positioned so that their focal points coincide with each other, and an aperture stop 7 is installed at that position. By adopting a double-telecentric optical system, the magnification of the optical system is less likely to change even when the distance between the measurement device 1 and the measurement object 2 changes, and a configuration can be realized that is less susceptible to effects such as positional deviation due to changes in the temperature of the installation environment.

[0018] Hereinafter, the distance between the mounting reference plane of the measuring device 1 and the object to be measured 2 will be referred to as WD (Working Distance). The focusing elements 5 and 8 are composed of a single lens or a group of lenses. The magnification of the optical system is determined by the ratio of the focal lengths of the focusing elements 5 and 8. It can be selected appropriately depending on the desired resolution. If changes in WD and the installation position of the sensor 6 can be ignored, a normal imaging optical system can also be selected. Furthermore, if changes in WD cannot be ignored but changes in the installation position of the sensor 6 can be ignored, it is also possible to select an object-side telecentric optical system.

[0019] The sensor 6 is composed of a photoelectric conversion element array such as a CCD element or a CMOS element. The sensor 6 is a line sensor or an area sensor. In the case of an area sensor, it is possible to detect two-dimensional displacement, and in the case of a line sensor, it is possible to detect one-dimensional displacement. Here, one-dimensional length measurement (amount of displacement) using a line sensor will be described. However, the measurement in embodiment 1 is not limited to length measurement (amount of displacement).

[0020] The light beam focused on the sensor 6 is photoelectrically converted and then output to a signal processing unit 9, where it undergoes processes such as dynamic range correction and gamma correction to generate image data (data for each pixel). The image data generated by the signal processing unit 9 is supplied to a control unit 10. The control unit 10 includes a CPU as a computer and a memory as a storage medium that stores a computer program. The control unit 10 calculates the amount of displacement of the measurement object 2 based on the image data in accordance with the computer program, outputs the measurement value (amount of displacement) to an external device, and functions as control means that controls the operation of each unit of the entire length measuring device as the measuring device 1.

[0021] FIG. 2 is a flowchart showing an example of the measurement flow of the measurement device of embodiment 1, and the operation of each step in the flowchart of FIG. 2 is performed by the computer in the control unit 10 executing a computer program stored in memory. Upon starting measurement, in step S10, the measuring device 1 sequentially acquires images at the set sampling rate using the sensor 6. In step S11, the first acquired image is set as a reference image, in step S12, images are acquired sequentially, and in step S13, the image acquired in step S12 is set as a measurement image.

[0022] Then, in step S14, the amount of displacement is calculated by calculating the correlation between the reference image and the measurement image, and in step S15, the amount of displacement is output as a measurement value (amount of displacement). In step S16, it is determined whether or not to end the measurement operation based on the operation output of an operation unit (not shown), and if not, the process returns to step S12 and repeats the processes of steps S12 to S16. If it is determined in step S16 that the measurement operation is to be ended, the measurement flow of FIG. 2 ends.

[0023] Note that if sampling progresses and the reference image deviates from the measurement area, processing such as updating the reference image may be performed. In this way, in the measuring device 1 of the first embodiment, light from the measurement object is received by the image sensor, and a measurement value for the measurement object is calculated as, for example, a length measurement value (amount of displacement) using the cross-correlation function of two images acquired by the image sensor.

[0024] FIG. 3 is a flowchart showing a specific example of the displacement amount calculation in step S14 of FIG. 2, and the operation of each step of the flowchart of FIG. 3 is performed by the computer in the control unit 10 executing a computer program stored in memory. The displacement amount is calculated by calculating the cross-correlation function between the reference image and the measurement image, and determining the displacement from the position of the extremum. The calculation of the cross-correlation function is performed, for example, in frequency space. That is, in step S101, the reference image is Fourier transformed, and in step S102, the image is limited to a predetermined frequency band by a band-pass filter.

[0025] Next, in step S103, the measurement image is Fourier transformed, and in step S102, the image is limited to the same frequency band as in step S102 using a band-pass filter. Note that a window function may be applied when performing the Fourier transform. The band-pass filters in steps S102 and S103 are configured so that each frequency component of the Fourier transformed data can be set to transmit or not transmit.

[0026] Next, in step S105, one of the conjugate complex numbers of the Fourier transformed image is taken and multiplied, and in step S106, an inverse Fourier transform is performed to obtain a correlation function. In step S107, the maximum value (extremum) of the cross-correlation function is detected, and in step S108, the correlation position at the maximum value (extremum) is detected. Note that the maximum value (extremum) is determined in pixel units.

[0027] Furthermore, in S109, a sub-pixel estimation calculation is performed to achieve high accuracy by calculating with a resolution equal to or smaller than the size of one pixel. In the first embodiment, during the sub-pixel estimation calculation, an extreme value of the cross-correlation function and values ​​before and after it are used to approximate using, for example, a quadratic function, and the extreme value of the approximate function is calculated as the sub-pixel estimation value. In addition to a quadratic function, other approximation methods may also be used, such as approximation using the intersection of straight lines or approximation using a Gaussian distribution. In this way, the measurement value (displacement amount) is calculated based on the sub-pixel estimation value calculated based on the cross-correlation function.

[0028] In addition, in the first embodiment, in step S110, a correction coefficient is calculated based on the spread of the peaks of the cross-correlation function. Note that the spread of the peak shape of the cross-correlation function changes depending on the configuration of the spatial frequency components, so in the first embodiment, the correction coefficient is acquired based on the spread of the peaks of the cross-correlation function. That is, in step S110, the correction coefficient is calculated depending on the configuration of the spatial frequency components.

[0029] Then, in step S111, the measurement value (length measurement value), which is the result of the sub-pixel estimation calculation, is corrected using the correction coefficient to calculate the displacement of the measurement object 2. Here, step S111 functions as a correction step (correction means) that corrects the measurement value. Note that if the spread of the peak shape is equal to or greater than a predetermined value, the measurement value is corrected, and if the spread of the peak shape is smaller than the predetermined value, no correction is performed since the error can be ignored.

[0030] The above correction coefficients will be explained using Figures 4 to 6. Figures 4 to 6 show examples of measurement errors under various conditions and line sensor output data obtained during measurement. Figure 4(A) shows the surface dependency of measurement error data when measuring a metal sample and a paper sample as the measurement object 2, and (B) shows the line sensor output data for the metal sample and the paper sample.

[0031] 4(A) and (B), the measurement conditions, including the optical system and signal processing, were the same. Note that Fig. 4(A) shows the length measurement error data based on the sub-pixel estimation calculation when the correction in step S111 is not performed. The measurement error for the metal sample was approximately -0.02%, while the measurement error for the paper sample was a large -0.12%.

[0032] Furthermore, as shown in FIG. 4B, the line sensor output data for the metal sample contains more high frequency components than the line sensor output data for the paper sample. As shown in FIGS. 4A and 4B, it can be seen that when the surface of the measurement object 2 is rough and the low frequency components increase, the length measurement (displacement amount) error data increases.

[0033] Next, Figure 5(A) shows the WD dependency of measurement error data when measuring a metal sample as the measurement object 2, and (B) shows the line sensor output data when the WD is at the center and at the end. Figure 5(A) shows the measurement error (displacement amount) of a metal sample when the WD is different, with WD=0 [mm] on the horizontal axis representing the design value, and the measurement error when the WD is changed is plotted. In Figures 5(A) and (B), the same metal sample is used as the measurement object 2, and the measurement conditions are the same except for the WD.

[0034] Note that Fig. 5A shows length measurement error data based on the sub-pixel estimation calculation when no correction is performed in step S111. Also, as shown in Fig. 5B, the line sensor output data when the WD is at the center has more high frequency components than the line sensor output data when the WD is at the end.

[0035] When a double-telecentric optical system is used, the change in optical magnification is small even when the working distance is changed. However, as shown in Figure 5(A), the measurement error data deteriorates with changes in working distance, resulting in an error of about -0.1%. This error is larger than the error dependent on changes in optical magnification, indicating that it is due to an error factor other than changes in optical magnification. In other words, as shown in Figures 5(A) and (B), as the working distance of the measurement target 2 increases and the low-frequency component increases, the measurement (displacement) error data increases.

[0036] Next, Fig. 6(A) is a diagram showing the speed dependency of the measurement error data of the measurement object 2, and (B) is a diagram showing the line sensor output data when the speed of the measurement object 2 is different. In Fig. 6(A) and (B), the measurement conditions are the same except for the speed. Note that Fig. 6(A) shows the measurement (displacement amount) error data based on the sub-pixel estimation calculation when the correction in step S111 is not performed.

[0037] As shown in Figure 6(A), as the speed increases, the error increases, resulting in an error of about -0.1%. This tendency can occur due to distortion in the optical system, but it is generally possible to keep distortion small by using a telecentric lens. The error that occurred in this case was larger than the error dependent on distortion, indicating that the error was caused by speed, not by the optical system. In other words, as shown in Figures 6(A) and (B), as the speed of the measurement target 2 increases and the low-frequency components increase, the length measurement (displacement amount) error data increases.

[0038] 4 to 6, when the correction in step S111 is not performed, there are error factors that are not dependent on the characteristics of the optical system. Looking at the line sensor output data in Figures 4 to 6, there is a tendency for errors to increase when low-frequency components become dominant in the spatial frequency components contained in the line sensor output.

[0039] In the case shown in Figure 4, when comparing the line sensor outputs of a metal sample and a paper sample, there is a large difference in the composition of the spatial frequency components contained in the line sensor output, reflecting the difference in surface characteristics, typified by the surface roughness of each sample. In other words, the line sensor output of the paper sample, which has a large measurement error, is dominated by low-frequency components compared to the line sensor output of the metal sample.

[0040] Similarly, in the case shown in Fig. 5, blurring occurs due to defocusing caused by changes in WD, and in the case shown in Fig. 6, shaking occurs due to an increase in speed. It can be seen that when the effects of these blurring and shaking cause low-frequency components to become dominant as spatial frequency components contained in the line sensor output, the measurement error increases.

[0041] This tendency occurs because when low-frequency components are dominant as part of the spatial frequency components contained in the output of the line sensor used for measurement, the peak broadening of the cross-correlation function increases, thereby increasing the impact of sub-pixel estimation errors. The following describes the relationship between the occurrence of sub-pixel estimation errors and the amount of sub-pixel estimation errors caused by the broadening of the peak shape of the cross-correlation function.

[0042] FIG. 7 is a schematic diagram for explaining the occurrence of sub-pixel estimation errors. When calculating the cross-correlation function of two line sensor outputs acquired at different times in real space, one output is shifted by one pixel, and the resulting multiplication and sum is then calculated.

[0043] Figure 7 shows a graph of the overlap state of the two line sensor outputs for a typical shift amount and the corresponding cross-correlation function value. The number of pixels in the line sensor is N, and the shift amount is from -N to +N pixels. The overlap area of ​​the output data of the two line sensors is indicated by diagonal lines. This example shows the case where a displacement of m pixels (m is an integer other than zero) is measured.

[0044] When the displacement amount is equal to an integer pixel, ideally the peak of the cross-correlation function should have a symmetrical shape. However, as shown in Figure 7, asymmetry occurs in the overlapping area of ​​the two line sensor outputs, and therefore the peak shape of the cross-correlation function also becomes asymmetric. The occurrence of asymmetry can be explained as follows.

[0045] That is, in the calculation of the cross-correlation function in real space, the values ​​of the two line sensor outputs exist in the overlapping portion, so multiplication is performed as is, but there is nothing to multiply outside the overlapping portion, so it does not contribute to the calculation of the cross-correlation function. At the 0th pixel of the cross-correlation function, all N pixels for the two line sensor outputs become the overlapping portion.

[0046] At the mth pixel, the overlapping area is (Nm) pixels, and at the 2mth pixel, the overlapping area is (N-2m) pixels. At the mth pixel, the values ​​of the two line sensor outputs in the overlapping area match, so the value of the cross-correlation function is very large. At other times, the values ​​of the two line sensor outputs do not match, even in the overlapping area, so the values ​​of the cross-correlation function are randomly small compared to the value of the cross-correlation function at the mth pixel.

[0047] Here, when comparing the 0th pixel and the 2mth pixel of the cross-correlation function, a difference of 2m pixels occurs in the overlapping portion as described above. Because a difference occurs in the range contributing to the calculation of the cross-correlation function, the value of the cross-correlation function for the 0th pixel is generally larger than the value of the cross-correlation function for the 2mth pixel. In this way, asymmetry occurs in the peak shape when the cross-correlation function is calculated. This asymmetry in the peak shape of the cross-correlation function then affects sub-pixel estimation.

[0048] The fitting function used for subpixel estimation is generally a linear function or a symmetric function such as a quadratic function. That is, the subpixel estimate is calculated by fitting a linear function or a quadratic function based on the cross-correlation function. However, when fitting with asymmetry as shown in Figure 7, the estimated position of the peak is (m-Δ) pixels. (Δ>0) This means that the measured displacement has an error in the short direction. If m is a negative integer, it means movement in the opposite direction, but the result is similar.

[0049] 8A and 8B are schematic diagrams for explaining the relationship between the structure of spatial frequency components included in the line sensor output and the sub-pixel estimation error. Fig. 8(A) shows a cross-correlation function when the proportion of high frequency components in the overall spatial frequency components contained in the line sensor output is high, and Fig. 8(B) shows a cross-correlation function when the proportion of low frequency components in the overall spatial frequency components contained in the line sensor output is high, and Fig. 8(B) shows a cross-correlation function when the low frequency components are dominant as the composition of the spatial frequency components contained in the line sensor output.

[0050] In FIG. 8(A), the proportion of high-frequency components is relatively large compared to FIG. 8(B), so the spread of the peak shape of the cross-correlation function is narrower, and the magnitude of the sub-pixel estimation error is also small when the ideal cross-correlation function is asymmetric. On the other hand, as shown in Figure 8(B), when the proportion of low-frequency components is larger than that in Figure 8(A), the peak shape of the cross-correlation function becomes broader, and when the ideal cross-correlation function is asymmetric, the magnitude of the sub-pixel estimation error becomes larger.

[0051] As described above, in measurement methods such as length measurement using a cross-correlation function, asymmetry occurs in the peak shape of the cross-correlation function. When the low-frequency components in the spatial frequency components included in the sensor output increase, the spread of the peak shape of the cross-correlation function increases. When the spread of the peak shape of the cross-correlation function increases, the influence of the asymmetry of the peak shape increases, and the sub-pixel estimation error, i.e., the measurement error of length measurement, etc., increases. Furthermore, this error occurs in the direction of shortening the measurement quantity of length measurement, etc.

[0052] Figure 9 is a graph plotting the relationship between the spread of the peak shape of the cross-correlation function and the measurement error based on the results of measurements performed under various measurement conditions. As shown in Figure 9, a correlation is obtained between the spread of the peak shape of the cross-correlation function and the measurement error. As the proportion of low-frequency components in the spatial frequency components of the two images, which are the sensor outputs, increases, the peak shape of the cross-correlation function also broadens. Therefore, the peak spread on the horizontal axis of Figure 9 can also be replaced with an increase in the proportion of low-frequency components, which are part of the spatial frequency components.

[0053] In this embodiment, the relationship shown in the graph of FIG. 9 is used to correct the length measurement error. That is, in order to calculate the correction coefficient in step S110 in Fig. 3, the spread of the peak shape of the cross-correlation function acquired during the length measurement calculation is calculated. Then, based on the relationship between the spread of the peak shape and the length measurement error in Fig. 9, the length measurement error is estimated and the correction coefficient is determined.

[0054] At this time, the correction coefficient can be obtained by referring to an approximation formula such as a polynomial that uses the spread of the peak shape of the cross-correlation function as a variable for the length measurement (displacement amount) error, or a table stored in advance in memory. When calculating the spread of the peak shape of the cross-correlation function using an approximation formula, the calculation can be performed by utilizing the quadratic function fitting used in the sub-pixel estimation calculation in step S109.

[0055] 9 shows the relationship between peak broadening and measurement (displacement) error, and correction can be performed by calculating a correction coefficient using this relationship as an approximation or table, but for example, the relationship between peak broadening and the measurement error correction coefficient can also be used as an approximation or table. In this case, the measurement error correction coefficient corresponds to the reciprocal of the measurement error with its polarity reversed.

[0056] Furthermore, in this case, the proportion of low frequency components may be used instead of the peak broadening, and the relationship between the proportion of low frequency components and the correction coefficient may be expressed as an approximate formula or table, and correction may be performed using this. As can be seen from FIG. 9, when the peak spread is equal to or less than a predetermined peak spread value, i.e., when the proportion of low-frequency components is smaller than a predetermined proportion, the measurement error can be ignored. In such cases, the measurement error correction coefficient can be set to zero, and no correction is required.

[0057] FIG. 10 is a schematic diagram illustrating quadratic function fitting used in sub-pixel estimation calculation. The sub-pixel estimation calculation is performed using pixel values ​​before and after the maximum value of the cross-correlation function, which is set as the center of the peak. Here, an example is shown in which the maximum value is obtained at the m-th pixel. The quadratic function used for fitting is, for example, Equation 1.

[0058]

number

[0059]

number

number

number

[0060]

number

number

[0061]

number

[0062] FIG. 11(A) is a diagram showing the measurement error when no correction is performed under various conditions. Here, the error is displayed as an absolute value. As shown in the figure, a large error of about 0.6% occurs under some conditions. FIG. 11(B) is a diagram showing the measurement error when correction is performed under the same conditions. Here, the error is displayed as an absolute value. In FIG. 11(B), the measurement error is 0.1% or less under all conditions. In this way, by performing the correction step of step S111 in embodiment 1, the measurement error is significantly improved.

[0063] In this way, by utilizing the information obtained from the sensor output, it is possible to correct the measurement value (amount of displacement). Note that, although the first embodiment shows a method of estimating the measurement error from the spread of the peak shape of the cross-correlation function, the configuration of the spatial frequency components contained in the line sensor output can also be calculated by calculating a power spectrum from Fourier transformed data.

[0064] Another method for calculating the composition of spatial frequency components contained in the line sensor output is to determine that the more crossing points with a threshold value, the more high-frequency components there are. Alternatively, the proportion of high-frequency components can be determined using differential data, for example. Conversely, the proportion of low-frequency components can be determined using a low-pass filter.

[0065] In this embodiment, the low frequency components may be frequency components below a predetermined frequency threshold, such as an average frequency, in the configuration of spatial frequency components (frequency spectrum distribution) included in the line sensor output. Alternatively, it is sufficient if the deviation value on the low frequency side in the configuration of spatial frequency components (frequency spectrum distribution) is equal to or less than a predetermined threshold. In addition, in this embodiment, the proportion of low frequency components refers to, for example, the proportion of frequency components lower than the average frequency in the configuration of spatial frequency components (frequency spectrum histogram) to the total frequency.

[0066] <Embodiment 2> Next, a measurement device according to a second embodiment of the present invention will be described. FIG. 12 is a diagram illustrating an example in which linear function fitting is used as a method for calculating sub-pixel estimation of a cross-correlation function and the spread of a peak shape.

[0067] A linear function fitting is performed using a combination of the maximum value of the cross-correlation function and its position, and the third magnitude value and its position, and a linear function fitting is performed using a combination of the second and fourth magnitude values ​​and their positions.Subpixel estimation is then made possible by calculating the intersection of the two linear function fitting results.

[0068] Furthermore, it is possible to calculate the spread of the peak shape of the cross-correlation function from the slope of the two linear function fitting results. Then, in step S110, the length measurement error is obtained based on the spread of the peak shape of the cross-correlation function calculated in this way and data in memory that stores, for example, the graph of Figure 9 as a function table. Then, in step S111, the length measurement error can be used to correct the sub-pixel estimation calculation result.

[0069] Although the first and second embodiments have been described using an example of a length measuring device as the measuring device 1, the measuring device 1 may be, for example, a distance measuring device that measures the distance to a measurement object or distance distribution based on the correlation function of two images, or may be, for example, a measuring device that measures the shape, position, or orientation of a measurement object.

[0070] In the above embodiment, correction coefficients and the like are obtained using approximate expressions or tables based on peak broadening and the proportion of low-frequency components, as explained in Fig. 9. However, control may also be performed such that the correction coefficient increases as the surface roughness of the object to be measured increases, for example, based on the correlation shown in Fig. 4. Furthermore, measurement values ​​may be calculated based on subpixel estimates calculated based on a cross-correlation function, and the measurement values ​​may be corrected when the surface roughness of the object to be measured is equal to or greater than a predetermined roughness.

[0071] Alternatively, control may be performed such that the larger the error in the distance WD between a predetermined reference plane of the measurement device and the measurement object, the larger the correction coefficient, based on the correlation shown in Fig. 5. Furthermore, the measurement value may be calculated based on a sub-pixel estimated value calculated based on the cross-correlation function, and the measurement value may be corrected if the error in the distance WD is equal to or greater than a predetermined error.

[0072] Alternatively, control may be performed such that the correction coefficient increases as the velocity of the object to be measured increases, based on the correlation shown in Fig. 6. Furthermore, the measurement value may be calculated based on a sub-pixel estimated value calculated based on a cross-correlation function, and the measurement value may be corrected when the velocity of the object to be measured is equal to or greater than a predetermined velocity.

[0073] <Embodiment 3> FIG. 13 is a diagram showing a control system including a measurement device and a robot arm in embodiment 3, and the control system 100 is composed of the measurement device 1 of embodiments 1 and 2, a signal processing unit 9, a control unit 10, a display unit 11, a transport unit 14, a robot arm 20, etc. Furthermore, the measuring device 1 is supported by a robot arm 20 as a support device when in use. A light beam emitted from a light source 3 housed in the measuring device 1 is focused on the measurement object 2 by a focusing member 4, and illuminates the measurement object 2 being transported by a transport unit 14 in the direction of the arrow.

[0074] The sensor of the measuring device 1 captures an image of the measurement object 2 illuminated by the light source 3 and transported by the transport unit 14, acquires image data, and inputs it to the control unit 10 via the signal processing unit 9. The control unit 10 then executes a measurement process to measure the length, shape, position, posture, distance, etc. of the measurement object 2 based on the image data, and calculates measurement values. Based on the information on the length, shape, position, posture, distance, etc., which are the measurement values ​​obtained in the measurement process, the control unit 10 sends a drive command to the robot arm 20 to control the operation of the robot arm 20. Furthermore, the measurement data measured by the measuring device 1 and the obtained images may be displayed on the display unit 11.

[0075] The robot arm 20 holds (grabs) the measurement object 2 with a robot hand (grasping portion) 21 at the tip based on the measurement value by the measuring device 1, and performs movement such as translation and rotation, and attitude control processing. Furthermore, by performing processes such as assembling the measurement object 2 to other parts using the robot arm 20, an article made up of multiple parts, such as an electronic circuit board or a machine, can be manufactured. Also, by performing further processing steps on the moved measurement object 2, a process can be carried out to manufacture the final article. A control unit for controlling the robot arm 20 may be provided separately from the control unit 10.

[0076] The present invention has been described in detail above based on its preferred embodiments, but the present invention is not limited to the above-mentioned embodiments 1 to 3, and various modifications are possible based on the gist of the present invention, and these modifications are not excluded from the scope of the present invention.

[0077] Note that a computer program that realizes all or part of the control functions of the above-described embodiments may be supplied to the measuring device 1, the control unit 10, etc. via a network or various storage media. The computer (or CPU, MPU, etc.) in the measuring device 1, the control unit 10, etc. may then read and execute the program. In this case, the program and the storage medium storing the program constitute the present invention. [Explanation of symbols]

[0078] 1: Measuring equipment 2: Measurement object 3:Light source 4: Light collecting element 5: Light collecting element 6: Sensor 7: Aperture diaphragm 8: Light collecting element

Claims

1. A measuring device for measuring a measurement object, a processing unit that obtains a measurement value related to the measurement object using a cross-correlation function of two images acquired by receiving light from the measurement object with an image sensor; The measuring device according to claim 1, wherein the processing unit corrects the measured value in accordance with the spread of the peak shape of the cross-correlation function.

2. 2. The measurement device according to claim 1, wherein the measurement value is obtained based on a sub-pixel estimate calculated based on the cross-correlation function.

3. 3. The measurement apparatus according to claim 2, wherein the subpixel estimated value is calculated by performing fitting using a linear function or a quadratic function based on the cross-correlation function.

4. A measuring device described in any one of claims 1 to 3, characterized in that the processing unit corrects the measurement value when the spread of the peak shape is greater than or equal to a predetermined value.

5. 2. The measuring apparatus according to claim 1, wherein the processing unit calculates the spread of the peak shape based on the maximum value of the cross-correlation function, the position at which the maximum value occurs, and values ​​at positions before and after the maximum value.

6. 2. The measuring apparatus according to claim 1, wherein the processing unit performs correction using an approximation formula or a table based on the relationship between the spread of the peak shape and the measurement error.

7. The measuring device according to any one of claims 1 to 3, characterized in that the processing unit corrects the measurement value when the proportion of frequency components lower than a predetermined frequency in the distribution of spatial frequency components of the two images is equal to or greater than a predetermined proportion.

8. A computer program for causing a computer to function as the processing unit of the measuring device described in any one of claims 1 to 7.

9. The measuring device according to any one of claims 1 to 7; a robot that holds and moves the object to be measured based on the measurement values ​​obtained by the measuring device.

10. measuring the object to be measured using the measuring device according to any one of claims 1 to 7; and manufacturing an article by processing the object to be measured based on the results of the measurement.

Images