JP5252820B2 - 3D measurement method and 3D shape measurement apparatus using the same - Google Patents

Description

  The present invention relates to a three-dimensional measurement method and apparatus for measuring a three-dimensional shape in a measurement region of a measurement object by image processing using a phase shift method.

  In general, as a method for measuring the three-dimensional shape of an object in a non-contact manner, for example, there is a phase shift method capable of high-speed and high-precision measurement. In this phase shift method, first, a process of projecting a sinusoidal fringe pattern onto a measurement object and imaging with a camera is executed a plurality of times while shifting the phase of the fringe pattern at a constant interval (for example, π / 2). . Next, the phase information of the fringe pattern is calculated from the change of the luminance value at each pixel of the plurality of captured images obtained by imaging, and the calculation of the three-dimensional shape in the measurement region of the measurement object based on this phase information I do.

  Therefore, in the phase shift method, if the reliability of the above-described luminance value decreases due to the influence of the reflectivity of the measurement object, noise superposition during imaging, distortion of the sine wave shape, or the like, accurate phase information of the fringe pattern Cannot be obtained, and the measurement accuracy of the three-dimensional shape may be reduced. Accordingly, various techniques have been developed to improve the measurement by the phase shift method, such as evaluating the reliability of the luminance value in the measurement process of the three-dimensional shape.

  For example, in Patent Document 1, a plurality of projection optical units that project a fringe pattern are arranged, the measurement object is projected and imaged from different directions, and the luminance value of each pixel of a plurality of captured images obtained thereby is captured. A technique is disclosed in which the magnitude of change (the strength of contrast) is compared, and the three-dimensional information of the measurement object is calculated based on data from the projection direction with the highest contrast.

Further, in Patent Document 2, a phase value reliability determination unit that extracts a luminance saturation region, a contrast reduction region, a boundary region of the contrast reduction region, and a phase error region is provided, and the above-described region where the phase information of the calculated fringe pattern is not reliable Discloses a technique that enables highly accurate three-dimensional measurement as a whole by acquiring phase information with high reliability by separately performing re-measurement or the like.
JP 2004-309240 A JP 2006-23178 A

  However, in the conventional technology as described above, it is possible to cope with a decrease in measurement accuracy due to a decrease in contrast or an erroneous measurement due to saturation of a luminance value. In the vicinity of this boundary, imaging blurring occurred and erroneous measurement occurred. Therefore, it has been necessary to make the reflectance of the measurement object constant by, for example, painting the measurement object white during measurement. Therefore, the phase shift method is used in fields where painting on the measurement object is permitted, for example, reverse engineering and modeling, but is not permitted in painting, for example, visual inspection of workpieces on the production line. Was not used.

  Incidentally, the appearance inspection of the workpiece is currently performed by image processing in which two-dimensional image features such as edges are extracted from the luminance information of the captured image and various relationships in the measurement region are obtained by feature measurement. On the other hand, in recent years, as the cost of products that make up the measurement object has been reduced and the complexity of the measurement object has progressed, further reduction in the loss rate of the measurement object is required, and more accurate inspection is required. Has been. For example, in the appearance inspection of a molded product, it is necessary to identify dust placed on the molded product, foreign matter embedded in the surface of the molded product, discoloration of the surface, and the like. Such discriminability is required because the dust placed on the molded product can be wiped off in a later process, so it is not judged as defective, but the foreign material or resin embedded in the resin of the molded product This is because the discoloration cannot be corrected and is judged to be defective.

  However, in the above-described image processing, the dust on the molded product, the foreign matter embedded in the surface of the molded product, the discoloration of the surface, and the like are different from the normal part, so they are detected as uniform defects. However, it is difficult to identify and detect each of them, and those judged as defective for these reasons have to be checked again by the operator or regarded as defective and discarded. Here, if the operator checks again, it takes extra work man-hours, and if it is discarded, it will throw away non-defective items if wiped off. It was generating useless costs leading to up.

  The present invention has been made to solve the above-described problem. Even when a measurement object has a plurality of reflectances, the influence of blurring of imaging is eliminated, and a three-dimensional shape in a measurement region is accelerated. It is another object of the present invention to provide a three-dimensional measurement method and an apparatus using the same, in which discrimination performance in a measurement region is improved by enabling measurement with high accuracy.

  In order to achieve the above object, the invention of claim 1 uses a phase shift method in which a fringe pattern is projected onto a measurement object by shifting the phase of the fringe pattern and projected a plurality of times to obtain a plurality of measurement objects. A three-dimensional measurement method for measuring a three-dimensional shape in a measurement region of a measurement object having a reflectance of a luminance obtained by differentiating a luminance value distribution obtained from an image in the measurement region in a direction orthogonal to a fringe pattern When obtaining a correction relationship between the differential value and the phase correction value for correcting the influence of the difference in reflectance of the measurement object on the phase value obtained from the fringe pattern in advance, when measuring a three-dimensional shape, A luminance differential value is calculated from the measurement region, the phase value affected by the difference in reflectance is corrected based on the phase correction value obtained from the correction relationship, and the corrected phase value is converted into the phase By performing calculation using the shift method, and measures the three-dimensional shape of the measurement region.

According to a second aspect of the present invention, in the three-dimensional measurement method according to the first aspect, the measurement object is composed of two types having different surface reflectances, and the respective luminance values (L1, L2) at the time of measurement. ) in the case is known, said correction relation, and (a) the respective luminance values (L1, L2), wherein the luminance value in a direction perpendicular to (b) stripe pattern of the measuring region and the luminance value distribution for The standard deviation σ estimated from the shape of the MTF obtained based on the luminance value at the ideal edge obtained by binarizing the average value ((L1 + L2) / 2) of the respective luminance values as a threshold value; As a correction table of phase correction values corresponding to the above values, when measuring a three-dimensional shape, the phase value affected by the difference in reflectivity is corrected based on the correction table, and the In the measurement area It is intended to measure the three-dimensional shape that.

According to a third aspect of the invention, in the first or second aspect of the invention, after obtaining the luminance value distribution of the measurement region, the image is smoothed to calculate a luminance differential value. .

According to a fourth aspect of the invention, in the invention according to any one of the first to third aspects, the calculated luminance differential value is further increased in the direction from a plurality of correction relationships obtained in advance. A specific correction relationship is selected using a luminance second-order differential value obtained by differentiation.

A fifth aspect of the present invention is the invention according to any one of the first to fourth aspects, wherein the difference in reflectance is limited only when the calculated luminance differential value is equal to or greater than a preset specified value. The phase value affected by is corrected.

According to a sixth aspect of the present invention, a fringe pattern projecting unit that projects a fringe pattern whose light luminance changes on a measurement object, an imaging unit that images the measurement object on which the fringe pattern is projected, and the imaging unit are output. Obtained from an image in the measurement region, in a three-dimensional shape measurement apparatus comprising: a three-dimensional measurement processing unit that performs a calculation for measuring a three-dimensional shape in a measurement region of a measurement target based on image data to be measured The correction relationship between the luminance differential value obtained by differentiating the luminance value distribution in the direction perpendicular to the fringe pattern and the phase correction value for correcting the influence of the difference in reflectance of the measurement object on the phase value obtained from the fringe pattern. A storage unit that is obtained and stored in advance, and the three-dimensional measurement processing unit calculates a differential luminance value from the measurement region when measuring a three-dimensional shape, and based on the phase correction value obtained from the correction relationship. And correcting the phase value affected by the difference in reflectance, and calculating the corrected phase value using a phase shift method, whereby the measurement object has a plurality of reflectances. The three-dimensional shape in the region is measured.

  According to the first aspect of the present invention, it is possible to calculate the phase value excluding the influence due to the difference in the reflectance of the measurement object, so that the three-dimensional shape in the measurement region can be measured with high accuracy.

  According to the second aspect of the present invention, even when the measurement object has a portion that differs greatly in the imaging direction, it is possible to measure with high accuracy by eliminating the influence of the difference in reflectance.

According to the invention of claim 3 , it is possible to reduce the influence of the noise on the luminance differential value, and to measure the three-dimensional shape with higher accuracy.

According to the invention of claim 4, since the influence due to the difference in reflectance of the measurement object can be surely eliminated, the three-dimensional shape can be measured with higher accuracy.

According to the fifth aspect of the invention, it is possible to eliminate the influence of fine (instantaneous) fluctuations in the luminance value due to noise, so that the three-dimensional shape can be measured with higher accuracy and at higher speed.

According to the invention of claim 6 , even if the measurement object has a plurality of reflectances, a three-dimensional shape measuring apparatus capable of measuring the three-dimensional shape in the measurement region with high accuracy can be obtained.

(First embodiment)
A three-dimensional measurement method according to the first embodiment of the present invention will be described with reference to FIGS. This embodiment is applied because the surface of the measurement object is composed of two types having different reflectivities, the respective luminance values (L1, L2) at the time of measurement are known, and further, optical It is conceivable that the value of σ representing the degree of blurring of imaging by the system is also known (constant in the measurement region of the measurement object). For example, this corresponds to a case where three-dimensional measurement is performed on a white paper that is a flat surface and characters are drawn with a single color ink. FIG. 1 shows a schematic configuration of a three-dimensional shape measuring apparatus that can execute the three-dimensional measuring method of the present embodiment. The three-dimensional shape measurement apparatus 1 is for measuring a three-dimensional shape in a measurement region of a measurement object in a non-contact manner using a phase shift method, and the luminance of light on the measurement object 6 is sinusoidal. A fringe pattern projection unit 2 (stripe pattern projection unit) that projects a changing fringe pattern, an imaging unit 3 (imaging unit) that images a measurement object 6 on which the stripe pattern is projected, a fringe pattern projection unit 2 and an imaging unit 3, a 3D measurement processing unit 4 (3D measurement processing unit) that performs various processes for measuring a 3D shape in the measurement region based on various data output from these units, and a calculation result And a display unit 5 for displaying captured images and the like.

  The fringe pattern projection unit 2 includes an illumination unit 21, a fringe pattern generation unit 22, and a projection optical unit 23. The illumination unit 21 includes a light source and a condenser lens that converts light from the light source into parallel light, and the stripe pattern generation unit 22 includes a transmissive liquid crystal and a polarizing plate. The projection optical unit 23 is a lens for projecting the fringe pattern created by the fringe pattern generation unit 22 onto the measurement object 6. Thereby, the fringe pattern projection unit 2 can project the fringe pattern, and further, by removing the fringe pattern generation unit 22, it can project the uniform light. In the present embodiment, the transmissive liquid crystal and the polarizing plate are used for the fringe pattern generation unit 22. However, in order to shift the phase at a predetermined interval from the combination of the reflective liquid crystal and the polarizing plate, the fringe pattern baked on the film. Equivalent functions can be realized with a combination of the actuator and digital mirror device (DMD).

  The imaging unit 3 includes an imaging element 31 for photoelectrically converting a subject image and an imaging optical unit 32 including an optical lens for creating an image of the measurement object 6 on the imaging element 31. Thereby, an image such as a fringe pattern projected onto the measurement object 6 and deformed on the surface thereof is formed by the imaging optical unit 32. The image sensor 31 converts an image such as a stripe pattern into an image signal and outputs the image signal.

  The three-dimensional measurement processing unit 4 includes a fringe pattern data creation unit 41, a projection / imaging control unit 42, a memory 43, and a calculation unit 44. The fringe pattern data creation unit 41 creates fringe pattern data for generating a fringe pattern in the fringe pattern generation unit 22. The projection / imaging control unit 42 transfers the fringe pattern data to the fringe pattern projection unit 2 and performs control for projecting and imaging based on the fringe pattern data. The memory 43 stores the camera resolution, the incident angle, the fringe pattern pitch, and the number of phase shift steps set in the fringe pattern projection unit 2, the captured image captured by the imaging unit 3, and image processing described later on the captured image. The data output from each unit, such as an image subjected to, is recorded. The calculation unit 44 reads out the data in the memory 43 and performs various calculations such as calculation of the phase value of the fringe pattern and image processing of the captured image in order to measure the three-dimensional shape of the measurement target 6 in the measurement region. . The display unit 5 is a display or the like that displays the processing results in the above-described units.

  By the way, when the measurement object has a plurality of reflectances, the phase value of the calculated fringe pattern near the boundary of the light and darkness of the captured image is shifted from the correct value due to the blur of the imaging of the imaging unit 3. Incorrect measurement occurs at the boundary between the light and dark of the measurement object. Therefore, in the above-described memory 43, the difference between the luminance differential value obtained by differentiating the luminance value distribution obtained from the image in the measurement region in the direction orthogonal to the stripe pattern and the reflectance of the measurement object is the phase value of the stripe pattern. The correction relationship with the phase correction value for correcting the influence on the image is stored before measurement.

  This correction relationship is to execute a simulation using a computer or the like to project a sine wave stripe pattern on a plane with contrast and then restore the phase value of the stripe pattern when blurring occurs due to the optical system. Ask for. Here, the procedure of this simulation will be described. 2 to 10 show the following states in this simulation.

  First, a state where there is a contrast between a bright area and a dark area on the surface of an object to be measured is a state in which only two brightness values (L1, L2) of brighter and darker exist as shown in FIG. , y coordinate is set and expressed by x-direction luminance value distribution.

  Next, as shown in FIG. 3, a luminance value distribution is generated by superimposing four types of sine waves of initial phases 0, π / 2, π, and 3π / 2 on this luminance value distribution (FIG. 2). At this time, superimposition is performed so that the phase direction of the sine wave is in the x direction. Furthermore, as shown in FIG. 4, a blurring state is generated by performing a moving average using a normal distribution with a standard deviation σ on each of these luminance value distributions (FIG. 3). The brightness value distributions before and after the blur are the brightness value distributions (FIGS. 3A to 3D) of the four captured images when there is no blur on the imaging unit 3 side. This corresponds to the luminance value distribution ((FIGS. 4A to 4D) of four captured images in a certain case.

  Next, the phase value of the fringe pattern is calculated from the luminance value distribution (FIGS. 3A to 3D) in the state before blurring (no blur), and the luminance value distribution in the state after blurring (with blur) Similarly, the phase value of the fringe pattern is calculated from (FIGS. 4A to 4D). FIG. 5 is a plot of the phase values of these fringe patterns. Here, in the vicinity of the boundary between light and dark (bright area and dark area), there is a portion where an error occurs in the two phase values. Since the state before blurring is a correct phase value, the difference between these two phase values (phase difference P), that is, the phase value after blurring, which is obtained by subtracting the phase value before blurring is the value of imaging. This is a phase error due to blur. This phase difference is used as a phase correction value when correcting the phase value described later.

  FIG. 6 plots a luminance differential value D and a phase difference P obtained by differentiating the luminance value distribution shown in FIG. 2 in the x direction with the horizontal axis as the pixel position. Here, the change in the phase difference P substantially coincides with the change in the luminance differential value D. However, in the range where the luminance secondary differential value D2 obtained by further differentiating the luminance differential value D in the x direction is not 0, the phase difference P It can be seen that the image is slightly darker.

  FIG. 7 is a DP graph showing the relationship between the luminance differential value D and the phase difference P. Overall, the graph looks like a figure of 8. Here, the D-P graph is divided into two by the sign of the above-described luminance second-order differential value D2, and each is approximated by a cubic function passing through the origin. The larger the slope near the origin, the graph is when the luminance second derivative D2> 0. At this time, the DP graph is expressed as (Equation 1).

Here, since the coefficients a ₁ , b ₁ , c ₁ , a ₂ , b ₂ , and c ₂ in (Equation 1) and (Equation 2) differ depending on the values of L1, L2, and σ, respectively, L1, L2, σ The coefficients a ₁ , b ₁ , c ₁ , a ₂ , b ₂ , c ₂ of (Equation 1) are obtained for each combination. FIG. 8 shows a state in which only one parameter of L1, L2, and σ is changed. When only L1 in FIG. 8A is changed, the inclination near the origin when D2> 0 hardly changes, and when only L2 in FIG. 8B is changed, D2 is changed. It can be seen that the inclination near the origin when <0 hardly changes. It can also be seen that when σ in FIG. 8C is changed, the overall inclination changes regardless of the sign of D2.

The coefficients a ₁ , b ₁ , c ₁ , a ₂ , b ₂ , and c ₂ obtained by such a simulation are in the above correction relationship, and correspond to the values of L1, L2, and σ as shown in FIG. A table of the above correction relation is created as a table and used for correcting phase errors during actual measurement. Here, the DP graph is approximated by using a cubic function, but may be approximated by using an odd-order function that exhibits characteristics such as the above-described figure of 8. Further, as shown in FIG. 10, the phase difference P itself with respect to the luminance differential value D is used as a table by interpolating between plot points in the DP graph obtained by the simulation without performing such approximation by the odd-order function. You may keep it. Hereinafter, these tables are referred to as DP correction tables.

  Next, a measurement procedure of the three-dimensional shape measurement method according to this embodiment will be described. The following measurement procedure is mainly executed by the calculation unit 44. FIG. 11 shows a flowchart of the present embodiment. FIG. 12 shows a schematic outline of this measurement procedure. Here, the following is an example of a four-phase phase shift method in which the phase of the sine wave of the fringe pattern to be projected is changed by ¼ period, but in this embodiment, three or more phases are used. The phase shift method may be used. Note that each point in the captured image corresponds to each pixel.

First, as shown in step S <b> 1, a process is performed in which predetermined light is projected and imaged on the measurement object 6 to acquire a predetermined captured image. Here, the phase change / projection processing according to step S1 will be specifically described. FIG. 13 shows the flowchart. The processing of the next step LS11 is combined with the processing of step LE11, and during this time (in the present embodiment, steps S11 to S13) is a processing that is repeated by sequentially increasing the value of i from 0 to 3. That is, the following steps S11, S12, and S13 are executed by shifting the phase of the fringe pattern described later by π / 2. First, processing for generating a fringe pattern is executed based on the sinusoidal fringe pattern data L _i (u, v) created by the fringe pattern data creation unit 41 in step S11. The fringe pattern data L _i (u, v) is expressed as (Equation 3).

  Here, (u, v) is a coordinate on the liquid crystal of the fringe pattern projection unit 2, b (u, v) is a bias value, a (u, v) is an amplitude value, φ (u) is a phase value, iπ / 2 is a phase shift amount (i = 0, 1, 2, 3). The luminance of the fringe pattern changes with respect to the u direction, and u is a function of the phase φ and can also be expressed as u (φ).

Next, a process of projecting the stripe pattern of the stripe pattern data L _i (u, v) onto the measurement object 6 is executed by the stripe pattern projection unit 2 (S12), and the measurement object 6 onto which the stripe pattern is projected is displayed. Processing for imaging with the imaging unit 3 is executed (S13). At this time, the coordinates on the captured image are set to (x, y), and the x direction of the captured image is set to be a direction orthogonal to the sinusoidal stripe pattern. Next, a process of recording the luminance value I _i (x, y) at each point in the captured image in the memory 43 is executed (S14). At this time, the position of each pixel of the captured image is represented by coordinates (x, y) on the captured image. The processing from step LS11 to LE11 corresponds to the contents shown in FIG.

  Next, a process of projecting uniform light onto the measurement object 6 is executed by the fringe pattern projection unit 2 (S15), and a process of imaging the measurement object 6 onto which the uniform light is projected by the imaging unit 3 Is executed (S16). Further, a process of recording this one captured image in the memory 43 is executed (S17). This captured image is set as an image It. The image It is obtained by averaging the four captured images captured by shifting the phase of the fringe pattern and projecting the image, in addition to obtaining the image It by projecting and capturing uniform light. It may be acquired. The above is the description of the phase change / projection processing according to step S1.

  When the process of step S1 described above is executed, the process of generating a differential image Dt in the x direction for the luminance of the image It in step S2 is executed. This is for differentiating the luminance value distribution obtained from the image It in the x direction to calculate the luminance differential value at each point on the captured image. Here, the same differentiation operator as that used when the DP correction table is obtained in the above-described simulation is used.

  The process of the next step LS1 is combined with the process of step LE1, and the process during this period (in this embodiment, steps LS2 to LE2) is repeated until it is executed for all y coordinate values. The process in step LS2 is combined with the process in step LE2, and the process in this period (in the present embodiment, steps S3 to S5) is repeated until all x coordinate values are executed. Accordingly, the following steps S3, S4, and S5 are executed for all pixels of the captured image.

Next, as shown in step S3, processing for calculating the phase value of the fringe pattern is executed. Specifically, a sine wave is applied to each pixel from the luminance value I _i (x, y) (i = 0,1,2,3) of each pixel recorded in the memory 43, and the fringe pattern in each pixel The phase value φ (x, y) of is calculated. The phase value φ (x, y) is expressed as (Equation 4). The processing in step S3 corresponds to the contents shown in FIG.

  Next, as shown in step S4, processing for correcting an error in the phase value of the fringe pattern due to imaging blur is executed. Specifically, the phase correction value P is acquired from the DP correction table corresponding to the values of L1, L2, and σ that are known in advance, and the phase value φ ′ (x, y) obtained by correcting the phase value φ (x, y) ). The phase value φ ′ (x, y) is expressed as (Equation 5). The processing in step S4 corresponds to the contents shown in FIG.

  Next, as shown in step S5, three-dimensional coordinates X (x, y), Y (x, y), and Z (x, y) are calculated for the pixel position (x, y) of the captured image. Processing is executed. Specifically, the fringe pattern projected on the measurement object 6 is modulated by the three-dimensional shape of the measurement object 6, but the phase value φ (u) at the time of projection and the phase value φ obtained from the captured image. Since (x, y) is the same, the correspondence between the coordinates on the image sensor 31 and the coordinates on the liquid crystal is obtained, and the three-dimensional coordinates of the surface of the measurement object 6 are obtained by the principle of triangulation. The three-dimensional coordinates X (x, y), Y (x, y), and Z (x, y) are expressed as (Equation 6).

Here, u (f) is a function for associating the phase with the pixel on the liquid crystal element, and p ^c and p ^p are parameters obtained by calibration of the imaging unit 3 and the fringe pattern projection unit 2, respectively. 7) and (Formula 8) are satisfied. The processing in step S5 corresponds to the content shown in FIG.

  By the above procedure, the three-dimensional shape in the measurement region of the measurement object is measured. Therefore, since the phase value excluding the influence due to the difference in reflectance of the measurement object can be calculated, the three-dimensional shape in the measurement region can be measured with high accuracy.

(Second Embodiment)
A three-dimensional measurement method according to a second embodiment of the present invention will be described with reference to the drawings. This embodiment is applied to two types of objects having different surface reflectances of the measurement object, the respective luminance values (L1, L2) at the time of measurement are known, and the optical system It is conceivable that the value of σ representing the degree of blurring of imaging due to is different in the captured image (not constant in the measurement region of the measurement object). For example, this corresponds to a case where three-dimensional measurement is performed on a single-color shape portion having a portion whose height direction is greatly different, in which characters are drawn with different single-color inks. The three-dimensional shape measurement apparatus that can execute the three-dimensional measurement method of the present embodiment is the same as that of the first embodiment (hereinafter the same).

  FIG. 14 shows a flowchart of the present embodiment. Here, steps equivalent to the steps shown in FIG. 11 described above are given the same reference numerals, and the same reference numerals are also given in the following flowcharts. First, steps S1 and S2 are executed.

  Next, as shown in step S6, a process for calculating the value of σ is executed. Here, the calculation process of σ according to step S6 will be specifically described. FIG. 15 shows the flowchart. First, as shown in step S61, a process of extracting a collection of pixels at a position where the luminance value changes as a correction target range in which the value of σ is to be calculated is executed. Specifically, the entire differential image Dt is located while the luminance value in the x direction changes from L1 to L2 or from L2 to L1 (at this time, the luminance differential value changes from 0 and returns to 0). A collection of pixels is extracted by being divided as one correction target range, and the number p of the correction target ranges is extracted.

  The process of the next step LS61 is combined with the process of step LE61, and the process during this period (in this embodiment, steps S62 and S63) is repeated until it is executed for all the p correction target ranges described above. is there.

  Next, as shown in step S62, processing for calculating the MTF within the correction target range is executed. Specifically, for the correction target range, based on the luminance value in the x direction of the image It and the luminance value at the ideal edge estimated from the values of L1 and L2, MTF estimation by the edge spread function is performed. Do. Here, the ideal edge is obtained by binarizing the luminance value in the x direction of the image It with a threshold value of (L1 + L2) / 2. As is known, this MTF is an index representing the transmission performance when the optical system is regarded as one transmission system, and is defined for each spatial frequency, and is expressed as (Equation 9). .

  Where ESF (x) is the edge spread function (imaged edge profile), E (x) is the ideal edge profile, h (x) is the Hanning window with [0, W] as the window region, ξ is a spatial frequency, and f [] is a Fourier transform.

  Next, as shown in step S63, a process of calculating the value of σ within the correction target range from the MTF is executed. Specifically, the MTF for the standard deviation σ value when the blur of imaging is assumed to be a normal distribution is obtained in advance, and the MTF obtained by (Equation 9) is closest to the shape of the MTF for which σ value. Is determined by the least square method, and an optimal value of σ is selected. The above is the description of the σ calculation process according to step S6.

  The processing after the next step LS1 is the same as in the first embodiment. In this way, by correcting the error of the phase value of the fringe pattern using the calculated σ value and L1 and L2 that are known in advance, the measurement object has a part that differs greatly in the imaging direction. Thus, even when the degree of blur of imaging is not constant in the measurement region of the measurement object, it is possible to measure with high accuracy by eliminating the influence of the difference in reflectance.

(Third embodiment)
A three-dimensional measurement method according to a third embodiment of the present invention will be described with reference to the drawings. As a case where the present embodiment is applied, the reflectance of the surface of the measurement object is diverse, so the values of L1 and L2 vary depending on the position in the image It, and the image It can be represented by only two types of luminance values. It may be inappropriate.

  FIG. 16 shows a flowchart of the present embodiment. First, similarly to the second embodiment, the processing from steps S1 to S6 is executed. Next, as shown in S <b> 7, the image It is divided into regions based on the luminance values, and for each image in the divided region, processing for obtaining a representative luminance value that represents the luminance value of the image is executed. At this time, the region division is performed by integrating adjacent pixels, extracting a contour line, or the like, and the average luminance value of the image is used as the representative luminance value for each image in the region.

  The process of the next step LS3 and the process of the next step LS4 are processes corresponding to the process of step LS1 and the process of step LS2. Accordingly, the following steps S3, S8, S4, and S5 are executed for all pixels of the captured image.

Next, the process of step S3 is executed, and as shown in step S8, the process of obtaining L1 and L2 in the pixel of interest is executed. Specifically, the representative luminance value (L _a ) of the region to which the pixel of interest belongs, and the representative luminance value (L _b ) of the region closer to its own pixel among the regions adjacent to this region in the x direction The brighter representative luminance value is L1, and the darker representative luminance value is L2. Therefore, L1 and L2 are expressed as (Equation 10) and (Equation 11).

  Next, steps S4 and S5 are executed. In this way, the values of L1 and L2 are obtained for each pixel of interest, and the error in the phase value of the fringe pattern is corrected, so that a plurality of reflectivities greatly differ in each part on the measurement object. However, it is possible to measure with high accuracy by eliminating the influence of the difference in reflectance.

(Fourth embodiment)
A three-dimensional measurement method according to the fourth embodiment of the present invention will be described with reference to the drawings.
This embodiment can be applied to the case where the brightness value varies due to noise in the projection system or the imaging system.

  FIG. 17 shows a flowchart of the present embodiment. First, the process of step S1 is executed. Next, a process of applying a median filter or a moving average filter to the image It in step S9 to smooth the image It is executed. The processing after the next step S2 is the same as that of the third embodiment.

  By smoothing the image It in this way and generating the differential image Dt, it is possible to reduce the influence of the noise on the luminance differential value and to measure the three-dimensional shape with higher accuracy. Can do.

(Fifth embodiment)
A three-dimensional measurement method according to a fifth embodiment of the present invention will be described with reference to the drawings. As in the fourth embodiment, the present embodiment may be applied when the luminance value varies due to noise in the projection system or the imaging system.

  FIG. 18 shows a flowchart of the present embodiment. Also in the present embodiment, the processes from Steps S1 to S7 are executed as in the fourth embodiment. The process of the next step LS5 and the process of the next step LS6 are processes corresponding to the process of step LS1 and the process of step LS2. Accordingly, the following steps S3, S8, S10, S4 and S5 are executed for all pixels of the captured image.

  Next, the processes of steps S3 and S8 are executed, and the process of determining whether or not the absolute value of the luminance differential value is greater than or equal to a preset specified value in step S10 is executed. If the absolute value of the luminance differential value is greater than or equal to the specified value, the process of step S4 is executed. FIG. 19 shows a pixel range (correction execution range) in which the absolute value of the luminance differential value is equal to or greater than a specified value.

  Next, the process of step S5 is performed. As described above, as long as the luminance differential value is equal to or larger than a predetermined value set in advance, the influence of fine (instantaneous) fluctuation of the luminance value due to noise can be eliminated by correcting the phase value. The three-dimensional shape can be measured with higher accuracy and higher speed.

  In addition, this invention is not restricted to the structure of the said various embodiment, A various deformation | transformation is possible in the range which does not change the meaning of invention. For example, the correction relationship may be stored in a memory provided separately from the memory 43 that records data output from the fringe pattern projection unit 2 and the imaging unit 3.

The block block diagram which shows the three-dimensional shape measuring apparatus which can perform the three-dimensional measuring method which concerns on the 1st Embodiment of this invention. (A) is the schematic which shows on a plane with a contrast in the simulation for calculating | requiring correction | amendment relationship, (b) is a figure which shows the luminance value distribution of the x direction of Fig.2 (a). 2A is a diagram in which a sine wave with an initial phase of 0 is superimposed on the diagram of FIG. 2B, FIG. 2B is a diagram in which a sine wave with an initial phase of π / 2 is superimposed, and FIG. The figure which piled up the sine wave of (pi), (d) is the figure which piled up the sine wave of the initial phase 3π / 2 similarly. (A) is the figure which performed the moving average to the figure of Fig.3 (a), (b) is the figure which performed the moving average to the figure of FIG.3 (b), (c) is the figure of FIG.3 (c). The figure which performed the moving average, (d) is the figure which performed the moving average to the figure of FIG.3 (d). (A) is a figure which shows the phase value of the x direction before and after blurring computed from Drawing 3 (a)-(d) and Drawing 4 (a)-(d), and (b) is Drawing 5 (a) ) Is an enlarged view of a portion A, and FIG. 5C is an enlarged view of a portion B of FIG. The figure which shows the luminance differential value D which differentiated the luminance value distribution of FIG. 2, and the phase difference P shown in FIG. 5 for every position of a pixel. A graph (D-P graph) showing the relationship between the luminance differential value D and the phase difference P. (A) shows changes in DP graph when L1 is a variable and L2 and σ are constants. (B) shows changes in DP graph when L2 is a variable and L1 and σ are constants. FIG. 4C is a diagram showing a change in the DP graph when σ is a variable and L1 and L2 are constants. The figure which shows the table at the time of approximating the relationship between the luminance differential value D and the phase difference P with an odd-order function. The figure which shows the table when the said relationship is not approximated by an odd-order function. The flowchart which shows the measurement procedure of the said measuring method. (A) thru | or (d) is a schematic diagram which shows the main processes in the said measurement procedure. The flowchart which shows the procedure of the phase change and projection process by the said measuring method. The flowchart which shows the measurement procedure of the three-dimensional measuring method which concerns on the 2nd Embodiment of this invention. The flowchart which shows the procedure of (sigma) calculation by the said measuring method. The flowchart which shows the measurement procedure of the three-dimensional measuring method which concerns on the 3rd Embodiment of this invention. The flowchart which shows the measurement procedure of the three-dimensional measuring method which concerns on the 4th Embodiment of this invention. The flowchart which shows the measurement procedure of the three-dimensional measuring method which concerns on the 5th Embodiment of this invention. The figure which shows the correction | amendment execution range in the said measuring method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 3D shape measuring apparatus 2 Stripe pattern projection unit (stripe pattern projection unit)
3 Imaging unit (imaging unit)
4 3D measurement processing unit (3D measurement processing unit)
DESCRIPTION OF SYMBOLS 5 Display unit 6 Measurement object 21 Illumination part 22 Stripe pattern production | generation part 23 Projection optical part 31 Imaging element 32 Imaging optical part 41 Stripe pattern data creation part 42 Projection / imaging control part 43 Memory (storage part)
44 Calculation unit

Claims (6)

In the measurement area of the measurement object when the measurement object has a plurality of reflectivities by using the phase shift method of projecting and imaging the fringe pattern on the measurement object by shifting the phase of the fringe pattern multiple times A three-dimensional measurement method for measuring a three-dimensional shape,
A luminance differential value obtained by differentiating the luminance value distribution obtained from the image in the measurement region in a direction orthogonal to the fringe pattern, and the influence of the difference in reflectance of the measurement object on the phase value obtained from the fringe pattern Obtain a correction relationship with the phase correction value in advance,
When measuring a three-dimensional shape, a luminance differential value is calculated from the measurement region, a phase value affected by the difference in reflectance is corrected based on a phase correction value obtained from the correction relationship, and the correction A three-dimensional measurement method characterized in that a three-dimensional shape in the measurement region is measured by calculating a phase value obtained using a phase shift method. In the case where the measurement object is composed of two types having different surface reflectances, and the respective luminance values (L1, L2) at the time of measurement are known,
It said correction relationship, (a) the respective luminance values (L1, L2), (b ) the measurement region and the luminance value average of the respective brightness values the luminance value in a direction perpendicular to the stripe pattern of distribution for The phase corresponding to the value of the standard deviation σ estimated from the shape of the MTF obtained based on the luminance value at the ideal edge obtained by binarization using ((L1 + L2) / 2) as a threshold value Obtain it as a correction value correction table,
2. The three-dimensional shape in the measurement region is measured by correcting a phase value affected by the difference in reflectance based on the correction table when measuring a three-dimensional shape. The three-dimensional measurement method described in 1. Wherein after obtaining the luminance value distribution of the measurement area, the three-dimensional measurement method according to claim 1 or claim 2 and a smoothing processing of the image and calculates the differential luminance value. A specific correction relationship is selected using a luminance secondary differential value obtained by further differentiating the calculated luminance differential value in the direction from a plurality of correction relationships obtained in advance. The three-dimensional measurement method according to any one of claims 1 to 3 . Only if the calculated differential luminance value is greater than or equal to a prescribed value set in advance, any one of claims 1 to 4, characterized in that to correct the phase value affected by the difference in the reflectance The three-dimensional measurement method according to one item. A fringe pattern projection unit that projects a fringe pattern in which the brightness of light changes onto a measurement object;
An imaging unit that images a measurement object onto which the fringe pattern is projected;
A three-dimensional shape measurement apparatus comprising: a three-dimensional measurement processing unit that performs a calculation for measuring a three-dimensional shape in a measurement region of a measurement object based on image data output from the imaging unit;
A luminance differential value obtained by differentiating the luminance value distribution obtained from the image in the measurement region in a direction orthogonal to the fringe pattern, and the influence of the difference in reflectance of the measurement object on the phase value obtained from the fringe pattern A storage unit for preliminarily obtaining and storing a correction relationship with the phase correction value;
When measuring the three-dimensional shape, the three-dimensional measurement processing unit calculates a luminance differential value from the measurement region, and is influenced by the difference in reflectance based on the phase correction value obtained from the correction relationship. A three-dimensional shape in the measurement region when the measurement object has a plurality of reflectances is measured by correcting the phase value and calculating the corrected phase value using a phase shift method. A three-dimensional shape measuring apparatus.