JP6247191B2 - Perspective distortion measuring apparatus and perspective distortion measuring method - Google Patents

Description

  The present invention relates to a perspective distortion measuring apparatus and a perspective distortion measuring method for measuring the degree of distortion of a measurement object such as glass or resin having light transmittance.

  In general, glass having optical transparency or the like is used for automobile window glass or architectural window glass, and resin or the like having optical transparency is used for a motorcycle hood or a monitor screen. When the degree of distortion is large in such light-transmitting glass or resin, the external scenery or displayed image that is visible through the glass or resin is distorted. Therefore, it is necessary to evaluate whether or not the strain caused by these glass or resin is within an allowable range.

For example, in Patent Literature 1, a plurality of display points arranged at predetermined intervals displayed on a screen are photographed through a transparent object, and inspection position coordinates of the photographed display points and a display stored in advance are stored. There is disclosed a distortion inspection apparatus that inspects perspective distortion by calculating a displacement amount from a reference position of a point.
In addition, for example, in Patent Document 2, the length of each line segment is measured by imaging a target composed of a plurality of line segments forming a predetermined angle through the plate-like body, and the measurement result and the plate-like body are removed. A perspective distortion detection method for detecting a perspective distortion by comparing the reference length of each line segment in FIG.

International Publication No. 2008/149712 Japanese Patent Laid-Open No. 06-144006

However, the strain inspection apparatus described in Patent Document 1 and the perspective distortion detection method described in Patent Document 2 measure strain based on a plurality of display points (light emission points). For this reason, in Patent Document 1 and Patent Document 2, it is difficult to measure a fine strain that enters a gap between display points, such as a fine ridge generated when a resin-made transparent plate material is press-molded. In such a case, it is conceivable to reduce the display point and shorten the interval at which the display point is arranged. However, because a plurality of display points overlap or cannot be captured due to distortion, it is impossible to capture one by one. It becomes difficult to identify one display point.
Accordingly, the present invention has been made paying attention to the above-described problem, and an object thereof is to provide a perspective distortion measuring apparatus and a perspective distortion measuring method capable of measuring perspective distortion with high spatial resolution. .

  To achieve the above object, a perspective distortion measuring apparatus according to an aspect of the present invention is a pattern that sequentially displays a plurality of stripe pattern images composed of a plurality of alternating light and dark stripes extending in one direction as a light and dark pattern. A display unit, an imaging unit that images a plurality of light and dark patterns displayed by the pattern display unit via the measurement target, and a perspective distortion that calculates a perspective distortion of the measurement target surface based on a plurality of imaging results by the imaging unit A perspective distortion calculation unit that extracts a first light-dark appearance order, which is a light-dark appearance order, for each pixel of the imaging unit from the imaging results of a plurality of stripe pattern images among the plurality of imaging results. The first code value indicating the coordinate value on the pattern display unit is calculated for each pixel from the first light and dark appearance order.

Further, the perspective distortion measuring method according to one aspect of the present invention includes a pattern display step of sequentially displaying a plurality of stripe pattern images composed of a plurality of alternating light and dark stripes extending in one direction as a light and dark pattern, and a pattern display. An imaging step of imaging a plurality of light and dark patterns displayed in the process via the measurement target, and a perspective distortion calculation step of calculating a perspective distortion of the measurement target surface based on a plurality of imaging results of the imaging step In the perspective distortion calculation step, the first light / dark appearance order, which is the light / dark appearance order, is extracted for each pixel from the image pickup results of the plurality of stripe pattern images among the plurality of image pickup results. The first code value indicating the coordinate value on the pattern display unit is calculated for each pixel.
Furthermore, the surface quality inspection method according to one aspect of the present invention is a method for inspecting a resin or glass molded product derived from perspective distortion caused by at least one of injection molding, press molding, component mounting, assembly, heat treatment, and finished product inspection. A quality defect is inspected by using the above-described method for measuring perspective distortion.

  According to the perspective distortion measuring device, the perspective distortion measuring method, and the surface quality inspection method according to the present invention, the perspective distortion can be measured with high spatial resolution.

It is an apparatus block diagram which shows the perspective distortion measuring apparatus which concerns on the 1st Embodiment of this invention. It is explanatory drawing which shows a stripe pattern image. It is a flowchart which shows the measuring method of perspective distortion of this embodiment. It is a figure for demonstrating the calculation method of a 1st code value. It is explanatory drawing which shows area division by the 1st reference code value. It is explanatory drawing which shows area division by the 1st code value. It is explanatory drawing which shows the shift | offset | difference of the boundary part by the 1st code value and the 1st standard code value. It is a graph which shows the 1st standard code value with respect to the pixel located in a line with the y-axis direction. It is a graph which shows the 1st code value with respect to the pixel located in a line with the y-axis direction. It is a graph which shows the difference of the 1st code value and the 1st standard code with respect to the pixel located in a line with the y-axis direction. It is explanatory drawing which shows a stripe pattern image and a multi slit pattern. It is a flowchart which shows the measuring method of perspective distortion which concerns on the 2nd Embodiment of this invention. It is explanatory drawing which shows the calculation method of a 2nd code value. It is explanatory drawing which shows the calculation method of a 2nd reference | standard code value. It is explanatory drawing which shows the calculation method of the difference of a code value. It is a flowchart which shows the measuring method of perspective distortion which concerns on the 3rd Embodiment of this invention. It is a block diagram which shows the modification of an apparatus structure. It is a block diagram which shows the modification of an apparatus structure. It is explanatory drawing which shows the modification of a stripe pattern image and a multi slit pattern. In an Example, it is a figure which shows the photograph when a measuring object is imaged on the background of a screen. It is an image which shows the photograph when the slit pattern image and multi slit pattern in Example 1 are projected. It is explanatory drawing which shows progress which calculates the difference of the 2nd code value and 2nd reference code value in Example 1. FIG. It is a figure which shows the result of mapping the difference of the 2nd code value in Example 1, and a 2nd reference | standard code value. It is a graph which shows the 2nd code value and the 2nd standard code value in the II position of Drawing 23. It is a graph which shows the difference value of the 2nd code value of Drawing 24, and the 2nd standard code value. In Example 1, it is a graph which shows the 2nd code value and the 2nd standard code value at the time of applying the 3rd embodiment. It is a graph which shows the difference value of the 2nd code value of Drawing 26, and the 2nd standard code value. In Example 1, it is a figure which shows the result of having mapped the difference of a 2nd code value and a 2nd reference | standard code value at the time of applying 3rd Embodiment. It is an image which shows the photograph when the slit pattern image in Example 2 and a multi slit pattern are projected. It is explanatory drawing which shows progress which calculates the difference of the 2nd code value in Example 2, and a 2nd reference | standard code value. It is a figure which shows the result of mapping the difference of the 2nd code value in Example 2, and a 2nd reference | standard code value. FIG. 32 is a graph showing a second code value and a second reference code value at the position II-II in FIG. 31. It is a graph which shows the difference value of the 2nd code value of Drawing 32, and the 2nd standard code value. In Example 2, it is a graph which shows the 2nd code value and the 2nd standard code value at the time of applying the 3rd embodiment. It is a graph which shows the difference value of the 2nd code value of Drawing 34, and the 2nd standard code value. In Example 2, it is a figure which shows the result of having mapped the difference of the 2nd code value and the 2nd reference code value at the time of applying the 3rd embodiment.

DESCRIPTION OF EMBODIMENTS Hereinafter, modes for carrying out the present invention (hereinafter referred to as embodiments) will be described in detail with reference to the drawings.
<1. First Embodiment>
[1-1. Device configuration]
First, a perspective distortion measuring apparatus according to a first embodiment of the present invention will be described with reference to FIGS. 1 and 2. The perspective distortion measuring apparatus according to the present embodiment is an apparatus that measures the perspective distortion of the measurement object 4 and includes an imaging unit 1, a pattern display unit 2, and a control unit 3, as shown in FIG.

The imaging unit 1 is an imaging device such as a video camera using a CCD element or the like, and images a plurality of light and dark patterns displayed on the pattern display unit 2 via the measurement target 4. In addition, the imaging unit 1 captures an imaging target region including at least all locations where the perspective distortion of the measuring target 4 is measured. The imaging unit 1 sends a plurality of captured images respectively corresponding to the plurality of captured light and dark patterns to the control unit 3.
The pattern display unit 2 includes a projector 2a and a screen 2b. The projector 2a is a display device that displays an image on a screen 2b such as a projector. The projector 2a acquires a plurality of light and dark patterns from the control unit 3 to be described later, and displays the acquired light and dark patterns in order on the flat screen 2b. Note that the number of imaging pixels of the imaging unit 1 and the number of projection pixels of the projector 2a can be appropriately selected according to the target measurement accuracy. For example, the number of imaging pixels of the imaging unit 1 may be 640 pixels × 480 pixels, and the number of projection pixels of the projector 2a may be 1024 pixels × 768 pixels. Further, from the viewpoint of a balance between the two precisions, the number of projection pixels may be about 1 to 2 times each of the number of image pickup pixels.

  In the present embodiment, as illustrated in FIG. 1, the imaging unit 1, the measurement object 4, and the screen 2 b are provided in a straight line in the z-axis direction. The screen 2b is provided with a flat display surface parallel to the xy plane. In addition, the projector 2a displays, on the screen 2b, a stripe pattern image composed of a plurality of stripes extending in the x-axis direction and alternately arranged in the y-axis direction at equal intervals as the light / dark pattern.

  The control unit 3 includes a sequence controller 3a, a perspective distortion calculation unit 3b, a pattern projection unit 3c, and an image buffer 3d. The image buffer 3d acquires and stores a plurality of captured images captured by the imaging unit 1. The pattern projection unit 3c stores a plurality of stripe pattern images as a plurality of light and dark patterns, selects a stripe pattern image to be displayed from a plurality of stripe pattern images based on an instruction from the sequence controller 3a, and transmits it to the projector 2a. . The perspective distortion calculation unit 3b calculates perspective distortion from a plurality of captured images captured by the imaging unit 1 and stored in the image buffer 3d. The sequence controller 3a transmits the bright and dark patterns to the projector 2a of the pattern projection unit 3c, captures the captured image into the image buffer 3d, transfers image data of the captured image to the perspective distortion calculation unit 3b, and fluoroscopy The distortion calculation operation of the distortion calculation unit 3b is controlled.

Here, in the present embodiment, the pattern projection unit 3c stores five types of stripe pattern images 6a to 6e shown in FIG. 2 as light and dark patterns, and displays the stripe pattern images 6a to 6e on the pattern display unit 2 in order. Let The stripe pattern images 6a to 6e are stripe-like images that extend in the x-axis direction and are divided into 2 ⁿ equal parts alternately in the y-axis direction. The stripe pattern images 6a to 6e are equally divided into 2, 4, 8, 16, and 32 regions in the y-axis direction.
The measuring object 4 is made of light-transmitting glass or resin. Further, the measurement object 4 may have been subjected to at least one of injection molding, press molding, component mounting, assembly, heat treatment, and finished product inspection.

[1-2. Method for measuring perspective distortion]
Next, with reference to FIGS. 1-7, the measuring method of the perspective distortion which concerns on this embodiment is demonstrated in detail.
First, as shown in FIG. 2, the sequence controller 3a causes the pattern projection unit 3c to select a stripe pattern image 6a to be initially displayed as a light / dark pattern (S100).

  Next, the projector 2a displays the stripe pattern images 6a to 6e on the screen 2b by projecting the stripe pattern images 6a to 6e selected by the pattern projection unit 3c on the screen 2b (S102). At this time, if the process performed immediately before is step S100, the selected stripe pattern image 6a is displayed. If the process performed immediately before is step S108 described later, the selected stripe pattern images 6b to 6e are displayed.

  Furthermore, the imaging unit 1 captures the stripe pattern images 6a to 6e displayed in step S102 (S104). At this time, the measuring object 4 is provided between the imaging unit 1 and the screen 2b. Therefore, the imaging unit 1 captures the stripe pattern images 6a to 6e that are observed through the measurement target. The captured image captured in step S104 is stored in the image buffer 3d.

Thereafter, the sequence controller 3a controls the stripe pattern images to be projected and imaged, and determines whether or not all the stripe pattern images 6a to 6e have been imaged (S106).
When all the stripe pattern images 6a to 6e are not captured in step S106, the sequence controller 3a displays the stripe pattern images 6b to 6e in the display order next to the stripe pattern images 6a to 6d displayed in the immediately preceding step S102. Is selected (S108). In the present embodiment, five types of stripe pattern images 6a to 6e shown in FIG. 2 are sequentially displayed and imaged.

  On the other hand, when all the stripe pattern images 6a to 6e have been imaged in step S106, the perspective distortion calculation unit 3b performs first from a plurality of imaging results stored in the image buffer 3d based on an instruction from the sequence controller 3a. Is calculated (S110). At this time, the perspective distortion calculation unit 3b performs encoding by binarizing and synthesizing each pixel of the image based on the five types of stripe pattern images 6a to 6e stored in the image buffer 3d, and 32 pixels are obtained for each pixel. A first code value indicating the area is calculated. The first code value is a value indicating coordinates on the screen 2b, and is a value indicating a region obtained by dividing the screen 2b into 32 parts at equal intervals in the y direction.

  The encoding of each pixel will be described with reference to FIG. FIG. 4 shows a case where each pixel is binarized using the stripe pattern images 6a to 6c and the first code value is calculated for each pixel for the sake of simplicity. In the example shown in FIG. 4, the imaging result of each pixel is bright for the captured image of the stripe pattern image 6a which is the primary pattern, the stripe pattern image 6b which is the secondary pattern, and the stripe pattern image 6c which is the tertiary pattern. The case was set to 1, and the case where the imaging result was dark was set to 0 and binarized. Further, by synthesizing the values in the primary pattern to the tertiary pattern in the order of appearance, the code value of each pixel is obtained and divided into 8 areas in the y direction (first code values indicated by 0 to 7 in FIG. 4). can do. In the present embodiment, the perspective distortion calculation unit 3b divides each pixel into 32 regions in order to divide the region from the imaging results of the stripe pattern images 6a to 6e.

  The first code value is a value indicating each divided area when the screen 2b is divided into eight areas equally spaced in the y direction according to the projected image as in the example illustrated in FIG. is there. Therefore, when the imaging field of view of the imaging unit 1 and the area on which the image on the screen 2b is projected match and there is no object that distorts light between the imaging unit 1 and the screen 2b, the first of each pixel By calculating a code value of 1, each pixel is divided into regions at equal intervals in the y direction. On the other hand, when there is an object having perspective distortion between the imaging unit 1 and the screen 2b as shown in FIG. 1, the area of each pixel divided by the calculated first code value depends on the amount of distortion. Move or transform. Therefore, it is possible to determine the presence or absence of perspective distortion by determining whether or not the area divided by the calculated first code value is an area divided at equal intervals in the y direction.

  FIG. 5 and FIG. 6 show an example of area division by the first code value. 5 and 6, for the sake of simplicity, each pixel is assumed to be the same as in FIG. 4 on the premise that the imaging field of the imaging unit 1 and the area on which the image on the screen 2 b is projected. The state divided into 8 regions by the first code value calculated based on the stripe pattern images 6a to 6c is shown. FIG. 5 shows a case where the measurement object 4 has no perspective distortion, and FIG. 6 shows a case where the measurement object 4 has a perspective distortion. 5 and 6, reference numeral 10 denotes the entire image, reference numeral 40 denotes an area on the image showing the measurement object 4, reference numerals 11A to 11H and 12A to 12H denote areas corresponding to the first code value, and reference numerals 11a to 11a. 11g and 12a-12g show the boundary part of each adjacent area | region 11A-11H and 12A-12H, respectively.

  As in the example illustrated in FIG. 5, when the visual field coincides with the pattern projection unit on the screen and there is no perspective distortion in the measurement target 4, the regions 12 </ b> A to 12 </ b> H divided by the first code value are images. The whole 10 is divided into eight equal parts in the y-axis direction, and becomes a rectangle having the same length and width. For this reason, the boundary portions 12a to 12h are straight lines that are horizontal in the x-axis direction. On the other hand, as in the example illustrated in FIG. 6, when the measurement target 4 has perspective distortion, the regions 11 </ b> A to 11 </ b> H divided by the first code value are rectangles obtained by dividing the entire image 10 into eight equal parts in the y-axis direction. However, the area and shape are changed from the state of FIG. 5 according to the position and amount of perspective distortion. For this reason, in the example shown in FIG. 6, for example, the boundary portions 11a to 11h are not straight lines parallel to the x-axis direction, but are curves having distortion in the y-axis direction. In this embodiment, although not shown, each pixel is divided into 32 sections. As described above, the first code value calculated in step S110 can determine the presence or the occurrence position of the perspective distortion by evaluating the distribution of the first code value for each pixel. For example, the presence or absence of perspective distortion can be determined by evaluating whether the shape of each region divided by the distribution of the first code value for each pixel is a rectangular shape equally divided in the y-axis direction. . In addition, since the determination of the perspective distortion is performed in units of pixels continuously in the x-axis direction, the perspective distortion can be measured with high spatial resolution.

  After step S110, the perspective distortion calculation unit 3b calculates the generation position and generation amount of the perspective distortion from the first code value and the first reference code value calculated in advance (S112). Similar to the first code value, the first reference code value is a value indicating coordinates on the screen 2b, and indicates a reference region in a state where there is no perspective distortion. In the present embodiment, the first reference code value is calculated by performing the processing from step S100 to step S110 in a state where there is no measurement object 4. Similarly to the first code value, each pixel is divided into a plurality of regions by the first reference code value. At this time, the state in which there is no measurement object 4 is the same as the state in which there is a measurement object 4 that is transparent and has no perspective distortion. Therefore, by performing the processing from step S100 to step S110, the region divided by the first reference code value is a rectangular shape divided at equal intervals in the y direction, for example, as in the example shown in FIG. It becomes the area of.

  Here, in step S112, the region where the first code value and the first reference code value are the same is compared for each pixel to calculate the generation position and amount of perspective distortion from the difference in the shape of the region. . For example, FIG. 7 compares the area divided by the solid line obtained from the first code value shown in FIG. 6 with the area divided by the broken line obtained from the first reference code value shown in FIG. FIG. In the example illustrated in FIG. 7, since there is a perspective distortion in the central portion of the measurement target 4 in the x-axis direction, the first code value is obtained in the central portion of the region 40 indicating the measurement target 4 in the x-axis direction. Deviation occurs between the boundary portion and the boundary portion obtained from the first reference code value. Since the amount of deviation corresponds to the amount of perspective distortion, the amount of distortion can be measured by detecting the amount of deviation.

  Step S112 will be described more specifically with reference to FIGS. 8 to 10 show the first code value, the first reference code value, and the first code value for each pixel arranged in the y-axis direction in the region of one pixel in the x-axis direction indicated by reference numeral 14 in FIG. And a first reference code value. 8 and 9, eight regions in which pixels in the y-axis direction are divided by boundary portions 12a to 12g and 11a to 11g based on the first code value and the first reference code value indicated by numerical values of 0 to 7. 12A to 12H and 11A to 11H.

In FIG. 10, the difference between the first code value and the first reference code value is determined between the boundary portion 11a and the boundary portion 12a, between the boundary portion 11b and the boundary portion 12b, and between the boundary portion 11c and the boundary portion 12c. Between the boundary portion 11d and the boundary portion 12d, between the boundary portion 11e and the boundary portion 12e, between the boundary portion 11f and the boundary portion 12f, and between the boundary portion 11f and the boundary portion 12f, respectively. . Therefore, it is possible to confirm that perspective distortion has occurred in the region delimited by the boundary where the difference has occurred, and further, the gap between the boundary between the first code value and the first reference code value. Is calculated as the amount of distortion.
In the present embodiment, the perspective distortion generation position and distortion amount shown in FIG. 9 can be continuously measured in the x-axis direction in units of pixels. Since such a measuring apparatus and measuring method can measure perspective distortion with high spatial resolution in the x-axis direction, it is difficult to measure with a method of measuring distortion based on a plurality of display points. Perspective distortion due to wrinkles can be measured.

<2. Second Embodiment>
[2-1. Device configuration]
Next, with reference to FIG. 11, the structure of the perspective distortion measuring apparatus according to the second embodiment of the present invention will be described. The perspective distortion measuring apparatus according to the present embodiment measures the perspective distortion of the measurement object 4 using the stripe pattern images 6a to 6e and the multi-slit pattern 6f as the light and dark patterns. The perspective distortion measuring apparatus according to the present embodiment includes an imaging unit 1, a pattern display unit 2, and a control unit 3, as in the first embodiment described with reference to FIG.

The configurations of the imaging unit 1 and the pattern display unit 2 are the same as those in the first embodiment.
The control unit 3 includes a sequence controller 3a, a perspective distortion calculation unit 3b, a pattern projection unit 3c, and an image buffer 3d.
The pattern projection unit 3c stores a plurality of stripe pattern images and a multi-slit pattern as a plurality of light and dark patterns. The pattern projection unit 3c selects a plurality of stripe pattern images to be displayed from the plurality of stripe pattern images based on an instruction from the sequence controller 3a, and transmits the selected stripe pattern image to the projector 2a. Thereafter, the pattern projection unit 3c transmits a multi-slit pattern to the projector 2a based on an instruction from the sequence controller 3a. Furthermore, the pattern projection unit 3c transmits a multi-slit pattern scanned based on an instruction to the sequence controller 3a to the projector 2a. Details of scanning of the multi-slit pattern will be described later.

  The image buffer 3d acquires and stores a plurality of captured images captured by the imaging unit 1. The perspective distortion calculation unit 3b calculates perspective distortion from a plurality of captured images captured by the imaging unit 1 and stored in the image buffer 3d. The sequence controller 3a transmits the bright and dark patterns to the projector 2a of the pattern projection unit 3c, captures the captured image into the image buffer 3d, transfers image data of the captured image to the perspective distortion calculation unit 3b, and fluoroscopy The distortion calculation operation of the distortion calculation unit 3b is controlled.

Here, in the present embodiment, the pattern projection unit 3c stores five types of stripe pattern images 6a to 6e and multi-slit patterns 6f shown in FIG. 11 as light and dark patterns, and each of the stripe pattern images 6a to 6e and multi-slits. The pattern 6f is displayed on the pattern display unit 2 in order. The stripe pattern images 6a to 6e are the same as those in the first embodiment. The multi-slit pattern 6f is a plurality of stripes extending in the x-axis direction and arranged at equal intervals in the y-direction. Further, the stripe width of the multi-slit pattern 6f is smaller than the stripe pattern image 6 having the smallest stripe width among the stripe pattern images 6a to 6e. The interval between the plurality of stripes of the multi-slit pattern 6f is the same distance as the stripe width of the stripe pattern image 6e.
The perspective distortion measuring apparatus according to the present embodiment is the same as the first embodiment with respect to the configuration other than the above.

[2-1. Method for measuring perspective distortion]
Next, with reference to FIGS. 11 to 15, the method for measuring perspective distortion according to the present embodiment will be described in detail.
First, as shown in FIG. 12, the sequence controller 3a causes the pattern projection unit 3c to select a stripe pattern image 6a to be initially displayed as a light / dark pattern (S200). Step S200 is performed similarly to step S100 of the first embodiment.

Next, the projector 2a displays the stripe pattern images 6a to 6e on the screen 2b by projecting the stripe pattern images 6a to 6e selected by the pattern projection unit 3c on the screen 2b (S202). Step S202 is performed similarly to step S102 of the first embodiment.
Furthermore, the imaging unit 1 captures the stripe pattern images 6a to 6e displayed in step S202 (S204). Step S204 is performed similarly to step S104 of the first embodiment.

Thereafter, the sequence controller 3a controls the stripe pattern images to be projected and imaged, and determines whether or not all the stripe pattern images 6a to 6e have been imaged (S206). Step S206 is performed similarly to step S106 of the first embodiment.
If all the stripe pattern images 6a to 6e are not captured in step S206, the sequence controller 3a displays the stripe pattern images 6b to 6e in the display order next to the stripe pattern images 6a to 6d displayed in the immediately preceding step S202. Is selected (S208). Step S208 is performed similarly to step S108 of the first embodiment.

After step S208, the process of step S202 is performed, and the stripe pattern images 6b to 6e selected in step S208 are projected on the pattern display unit 2. Note that the processes in steps S202 to S208 are repeated until the stripe pattern image 6e is captured.
On the other hand, when all the stripe pattern images 6a to 6e are captured in step S206, the sequence controller 3a causes the pattern projection unit 3c to select the multi-slit pattern 6f (S210).

Next, the projector 2a projects the multi-slit pattern 6f selected by the pattern projection unit 3c onto the screen 2b, thereby displaying the multi-slit pattern 6f on the screen 2b (S212).
Further, the pattern projection unit 3c scans the multi-slit pattern 6f in the y direction (S214). At this time, the imaging unit 1 continuously captures the multi-slit pattern 6f displayed on the screen 2b and scanned in the y direction at the same time as the multi-slit pattern 6f is scanned. At this time, in the imaging result, the captured image of the multi-slit pattern 6f and the imaging timing are stored simultaneously. The scanning of the multi-slit pattern is completed when the scanning is performed for an interval at which stripes of the multi-slit pattern 6f are arranged. In addition, the imaging by the imaging unit 1 is ended with the end of scanning of the multi-slit pattern 6f. The captured image captured in step S214 is stored in the image buffer 3d.

  After step S214, the perspective distortion calculation unit 3b calculates a first code value for each pixel from the imaging results of the stripe pattern images 6a to 6e among the plurality of imaging results stored in the image buffer 3d (S216). The calculation method of the first code value is the same as that in step S110 of the first embodiment. In step S216, the perspective distortion calculation unit 3b divides each pixel into regions for each calculated first code value.

  Thereafter, the perspective distortion calculation unit 3b calculates a second code value from the first code value and the imaging result of the multi-slit pattern 6f among the plurality of imaging results stored in the image buffer 3d (S218). Similar to the first code value, the second code value is a value indicating coordinates on the screen 2b, but is a finer coordinate than the first code value. In the present embodiment, the second code value indicates the coordinates in pixel units on the screen in the y-axis direction.

The second code value is calculated based on the first code value, a plurality of captured images of the multi-slit pattern 6f that are continuously captured, and the capturing timing of each captured image. Specifically, the perspective distortion calculation unit 3b evaluates the luminance for each pixel of a plurality of captured images in the order of imaging, and the imaging timing at which the luminance becomes the maximum luminance, that is, on the screen imaged by the video camera The timing at which the slit light is projected onto the corresponding point is extracted for each pixel, and the extracted imaging timing is converted into a value corresponding to the number of moving pixels of the slit on the screen to calculate the intermediate code value.
Next, the calculated intermediate code value is synthesized in units of areas divided by the first code value, whereby a second code value that changes monotonously for each pixel in the y-axis direction on the screen is calculated. (S218).

  The method for calculating the second code value will be specifically described with reference to FIG. In the example illustrated in FIG. 13, a case where a multi-slit pattern 6 f including eight stripes that are bright portions is used will be described for the sake of simplicity. FIG. 13 is an explanatory diagram showing a method for calculating the second code value in the area 14 in the example shown in FIG.

  First, the imaging timing with the maximum luminance is extracted for each pixel, and the extracted imaging timing is converted into a value corresponding to the number of moving pixels of the slit on the screen, so that it is shown in the upper graph of FIG. An intermediate code value is calculated. Since the multi-slit pattern 6f scans the screen at a constant speed and stops when the next slit pattern exists in the initial state, the conversion of the imaging timing is performed at the time that is the imaging timing. This is performed by multiplying a coefficient such that the entire scanning time is 1/8 of the total number of pixels in the y-axis direction of the image. Here, each stripe of the multi-slit pattern 6f moves within the regions 11A to 11H divided by the first code value by being scanned. Therefore, as shown in the upper graph of FIG. 13, the code value (intermediate code value) becomes 0 at the end of the y-axis negative direction side in each region 11A to 11H, and from there, the positive direction of the y-axis As the value moves to the side, the value gradually increases and becomes a code value corresponding to the time at which scanning is completed at the pixel at the end on the positive side in the y-axis. Next, in the region 11A to 11H divided by the first code value, the calculated intermediate code value and the first code value are combined, that is, the first code value is added to the total number of pixels in the y-axis direction of the image. By multiplying the coefficients so as to be 1/8 of and adding them, the second continuous code values shown in the lower graph of FIG. 13 are calculated.

  After step S218, the perspective distortion calculation unit 3b calculates the generation position and generation amount of the perspective distortion from the second code value and the preset second reference code value (S220). The second reference code value indicates a code value serving as a reference in a state where there is no perspective distortion. In the present embodiment, the second reference code value is a value calculated by performing the processing of Step S200 to Step S218 in a state where there is no measurement object 4, and is calculated as the second code value in Step S218. is there. FIG. 14 is an explanatory diagram showing an example of the calculation result of the second reference code value, which is obtained by converting the imaging timing extracted for pixels aligned in the y-axis direction corresponding to the region 14 of FIG. The intermediate code value (shown on the upper side of FIG. 14) and the second reference code value (shown on the lower side of FIG. 14) obtained by synthesizing the intermediate code values are shown. At this time, since the imaging unit 1 captures a state without perspective distortion, the number of pixels in the y-axis direction of each of the regions 12A to 12H is the same as shown in FIG. The second reference code value is a straight line.

In step S220, by subtracting the second reference code value from the second code value for each pixel, the amount of generation of perspective distortion in the area corresponding to each pixel can be calculated from the difference.
Step S220 will be specifically described with reference to FIG. The upper side of FIG. 15 is a graph showing the second code value and the second reference code value, and the lower side is a difference in code value between the second code value and the second reference code value. Note that the lower graph of FIG. 15 shows the scale of the vertical axis enlarged compared to the upper graph. As shown in FIG. 15, a difference in code value due to perspective distortion occurs between the second code value and the second reference code value. The position where the perspective distortion is generated can be calculated from the pixel in which such a code value difference occurs, and the amount of the perspective distortion generated can be calculated from the magnitude of the code value difference.

For example, a change amount per pixel of the second reference code value is obtained in advance, and a value obtained by subtracting the second reference code value from the second code value is a change amount of the code value per pixel on the screen. By dividing by 2, it is possible to detect how many pixels the viewpoint on the screen 2b has shifted in the region 40 of the measurement object 4. Furthermore, the perspective distortion at each point of the measurement target can be quantitatively measured by adding the geometric conditions of the imaging unit 1, the measurement target 4, and the pattern display unit 2. As described above, in this embodiment, the perspective distortion can be measured with a resolution corresponding to the number of pixels of the imaging unit 1 and the pattern display unit 2 in the x-axis direction and the y-axis direction. Perspective distortion can be measured. In the present embodiment, in addition to the fine local perspective distortion, the second reference code in a state where there is no measurement object is compared with the second code. For example, the measurement object 4 has a large curvature. In the case of having a portion, it is also possible to measure a perspective distortion that gradually changes due to the processed portion.
After step S220, the perspective distortion calculation unit 3b creates a map of perspective distortion by color or shade based on the perspective distortion calculated in step S220 (S222). In step S222, the perspective distortion is mapped in color or shading according to the amount of perspective distortion generated for the imaging region corresponding to each pixel.

<3. Third Embodiment>
Next, a perspective distortion imaging method according to the third embodiment of the present invention will be described with reference to FIG. In the perspective distortion imaging method according to the present embodiment, the perspective distortion is measured using the perspective distortion imaging apparatus having the same configuration as that of the second embodiment.
In the perspective distortion imaging method according to the present embodiment, first, the sequence controller 3a causes the pattern projection unit 3c to select a stripe pattern image 6a to be displayed first as a light / dark pattern (S300). Step S300 is performed similarly to step S200 of the second embodiment.

Next, the projector 2a displays the stripe pattern images 6a to 6e on the screen 2b by projecting the stripe pattern images 6a to 6e selected by the pattern projection unit 3c on the screen 2b (S302). Step S302 is performed similarly to step S202 of the second embodiment.
Further, the imaging unit 1 captures the stripe pattern images 6a to 6e displayed in step S302 (S304). Step S304 is performed similarly to step S204 of the second embodiment.

Thereafter, the sequence controller 3a controls the stripe pattern images to be projected and imaged, and determines whether or not all the stripe pattern images 6a to 6e have been imaged (S306). Step S306 is performed in the same manner as step S206 in the second embodiment.
If all the stripe pattern images 6a to 6e are not captured in step S306, the sequence controller 3a displays the stripe pattern images 6b to 6e in the display order next to the stripe pattern images 6a to 6d displayed in the immediately preceding step S302. Is selected (S308). Step S308 is performed similarly to step S208 of the second embodiment.

After step S308, the process of step S302 is performed, and the stripe pattern images 6b to 6e selected in step S308 are projected on the pattern display unit 2. Note that the processing in steps S302 to S308 is repeated until the stripe pattern image 6e is captured.
On the other hand, when all the stripe pattern images 6a to 6e are captured in step S306, the sequence controller 3a causes the pattern projection unit 3c to select the multi-slit pattern 6f (S310). Step S310 is performed similarly to step S210 of the second embodiment.

Next, the projector 2a displays the multi-slit pattern 6f on the screen 2b by projecting the multi-slit pattern 6f selected by the pattern projection unit 3c onto the screen 2b (S312). Step S312 is performed in the same manner as step S212 in the second embodiment.
Further, the pattern projection unit 3c scans the multi-slit pattern in the y direction (S314). Step S314 is performed similarly to step S214 of the second embodiment.

Thereafter, the perspective distortion calculation unit 3b calculates a first code value for each pixel from the imaging results of the stripe pattern images 6a to 6e among the plurality of imaging results stored in the image buffer 3d (S316). Step S316 is performed similarly to step S216 of the second embodiment.
Next, the perspective distortion calculation unit 3b calculates a second code value from the first code value and the imaging result of the multi-slit pattern 6f among the plurality of imaging results stored in the image buffer 3d (S318). Step S318 is performed similarly to step S218 of the second embodiment.

  Further, the perspective distortion calculation unit 3b calculates a moving average value of the second code value in the scanning direction of the multi-slit pattern, and subtracts the calculated average value from the second code value, thereby measuring the measured perspective distortion. Local perspective distortion is extracted (S320). In step S320, the moving average value Ca (x, y) of the second code value at the position (x, y) of the captured image is m as the width of the pixel for which the moving average is taken, and the second average before averaging. When the code value is C (x, y), it is calculated by the following equation (1). In the present embodiment, by extracting the difference between the second code value and the moving average value, it is possible to exclude the influence of perspective distortion that gently changes in the direction in which the moving average value is calculated. For this reason, even when the measurement object 4 has a large curvature processed portion as detected in the second embodiment, the influence of the perspective distortion due to the processed portion can be excluded, and the local Only perspective distortion can be measured.

  After step S320, the perspective distortion calculation unit 3b creates a map of perspective distortion by color or shade based on the local perspective distortion calculated in step S320 (S322). Step S322 is performed similarly to step S222 of the second embodiment.

<4. Modification>
The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention pertains can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that these also belong to the technical scope of the present invention.

  For example, in the above embodiment, the stripe pattern images 6a to 6e are composed of a plurality of stripes extending in the y direction, but the present invention is not limited to such an example. For example, as illustrated in FIG. 13, the stripe pattern images 7a to 7e may be a plurality of stripes extending in the x direction. Thereby, a fine distortion changing in the x direction can be detected. In such a case, the multi-slit pattern 7f is a stripe extending in the y direction. Furthermore, in the perspective distortion measuring apparatus and measurement method according to the present invention, the perspective distortion in different directions may be measured using the stripe pattern images 6a to 6e and 7a to 7e. At this time, the first code in different directions may be calculated using the multi-slit patterns 6f and 7f, and the perspective distortion may be measured.

  In the first embodiment, each pixel is divided into 32 areas according to the first code value. However, the present invention is not limited to this example. For example, by further using a stripe pattern image in which the number of divisions is increased from 32, each pixel may be divided into a larger number of regions than 32 regions. Thereby, in the first embodiment, the spatial resolution can be improved, and in the second and third embodiments, the distance for scanning the multi-slit pattern can be shortened, thereby shortening the measurement time. be able to.

Furthermore, in the second and third embodiments, five types of stripe pattern images 6a to 6e are used. However, the present invention is not limited to this example as long as the format is ²ⁿ division. For example, the stripe pattern image may be four types of 6a to 6d or more divided patterns. At this time, the number of stripes of the multi-slit pattern is the same as the number of stripes of the light and dark pattern that is the minimum width in the stripe pattern image. The number of divisions of the stripe pattern image may be selected according to the magnitude of the distortion to be measured. When the distortion is small, the number of areas divided by the first code value is increased. Many types of stripe pattern images may be used. As a result, the scanning range of the multi-slit pattern can be narrowed, and measurement in a short time becomes possible. On the other hand, when the distortion is large, fewer types of stripe pattern images may be used so that the number of regions divided by the first code value is reduced. As a result, when a stripe pattern image is used, it is possible to accurately measure a large amount of perspective distortion that is difficult to measure.

  Furthermore, in the second and third embodiments, in step S214, the multi-slit pattern 6f is scanned by the interval at which the stripes of the multi-slit pattern 6f are arranged. However, the present invention is not limited to such an example. The multi-slit pattern 6f may be scanned wider than the interval between the stripes of the multi-slit pattern 6f. Thereby, the distortion in a boundary part can be detected reliably.

  Furthermore, in the third embodiment, the moving average of the second code values is used when extracting the local perspective distortion in step S320, but the present invention is not limited to such an example. For example, local perspective distortion may be extracted by performing second-order difference processing of the second code value. The second-order differential value Cd (x, y) of the second code value at the position (x, y) of the captured image is m for the width of the pixel to be differentiated, and the second code value before the difference is C ( x, y), it is calculated by the following equation (2).

  Furthermore, in the above embodiment, as shown in FIG. 1, the imaging unit 1, the measurement object 4, and the screen 2 b are arranged in a straight line in order, but the present invention is not limited to such an example. For example, as shown in FIG. 17, the projector 2a, the measuring object 4, and the screen 2b are arranged in a straight line in order, and are displayed on the screen 2b so as not to hit the light emitted from the projector 2a. The imaging unit 1 may be provided at a position where an image can be captured.

Furthermore, in the said embodiment, as shown in FIG. 1, the pattern display part 2 was comprised from the projector 2a and the screen 2b, However, This invention is not limited to this example. For example, as shown in FIG. 18, the pattern display unit 2 may be a device that can directly display an image on a screen such as a flat display 2c.
Furthermore, the present invention is a resin derived from perspective distortion caused by at least one of injection molding, press molding, component mounting, assembly, heat treatment, and finished product inspection using the method for measuring perspective distortion shown in the above embodiment. Alternatively, a perspective distortion inspection method for inspecting a quality defect of a glass molded product may be used.

Further, in the above embodiment, the first reference code value and the second reference code value are calculated by measuring the first code value and the second code value in the state where the measurement object 4 is not present. However, the present invention is not limited to such an example. For example, the first reference code value and the second reference code value may be geometrically calculated from the positional relationship between the pattern display unit 2 and the imaging unit 1. In addition, the first reference code value and the second reference code value are calculated by measuring the first code value and the second code value in a state where a measurement target that is a reference for perspective distortion is installed. May be.
Further, in the present invention, in addition to the configuration shown in FIG. 1, a display device capable of displaying a measurement result of perspective distortion, a map of perspective distortion, and the like may be connected to the control unit 3.

<5. Summary>
(1) A perspective distortion measuring apparatus according to the present invention includes a pattern display unit that sequentially displays a plurality of stripe pattern images composed of a plurality of alternating light and dark stripes extending in one direction as a light and dark pattern, and the pattern display unit. An imaging unit that images the plurality of brightness and dark patterns displayed by the imaging target, and a perspective distortion calculation unit that calculates a perspective distortion of the measurement target surface based on a plurality of imaging results of the imaging unit. The perspective distortion calculation unit extracts, from the imaging results of the plurality of stripe pattern images among the plurality of imaging results, a first appearance order that is the order of appearance of light and dark for each pixel of the imaging unit, A first code value indicating a coordinate value on the pattern display unit is calculated for each pixel from the first appearance order. According to the above configuration, the perspective distortion can be measured with high spatial resolution in units of pixels in one direction by measuring the perspective distortion continuously in one direction. For this reason, distortion like a fine wrinkle can be measured.

(2) The perspective distortion calculation unit calculates perspective distortion from a difference between the first code value and a first reference code value serving as a reference set in advance, and the first reference code value is The coordinates on the pattern display unit corresponding to the stripe pattern image observed in the absence of a measurement target are shown.
(3) The first reference code value is calculated in advance from a plurality of imaging results captured by the imaging unit without the measurement target.

(4) The first reference code value is geometrically calculated from a positional relationship between the pattern display unit and the imaging unit. According to the above configuration, since it is not necessary to measure the first reference code value in advance, the time required for measurement can be shortened.
(5) The first reference code value is calculated in advance from a plurality of imaging results captured by the imaging unit in a state where a measurement target that is a reference for perspective distortion is installed.
(6) The pattern display unit includes a plurality of types of stripe pattern images including stripes extending in the one direction, and a plurality of types of stripe pattern images including stripes extending in a direction perpendicular to the one direction. Are respectively displayed, and the perspective distortion calculation unit calculates perspective distortion in a direction orthogonal to the one direction and in the one direction.

  (7) The pattern display unit has a multi-slit pattern composed of a plurality of stripes extending in the one direction, the width being smaller than the minimum width of the stripes of the plurality of types of stripe pattern images, as the light / dark pattern. The stripe pattern image is further scanned and displayed in the direction perpendicular to the stripe extending direction by the minimum width or more, and the perspective distortion calculation unit is configured based on the imaging result of the multi-slit pattern among the plurality of imaging results. The timing at which the luminance is maximized for each pixel is obtained as an intermediate code value, and the first code value is synthesized by calculating the movement distance of the slit light on the screen from the timing and combining it with the first code value. A second code value indicating a coordinate value on the pattern display section that is finer than the code value of is calculated.

  According to the above configuration, by using the multi-slit pattern, the perspective distortion can be measured with a high spatial resolution in units of one pixel of the imaging unit 1 in the direction in which the stripe extends and the direction perpendicular to the direction. . Further, in such a measuring method, since the multi-slit pattern is composed of a plurality of stripes, the distance for scanning the stripes can be shortened. In other words, in order to measure perspective distortion with the same measurement accuracy as the measurement method described in (6), compared to a case where one thin stripe is scanned from one end of the pattern display unit 2 to the other opposite end. Measurement time can be shortened.

  (8) The perspective distortion calculation unit calculates perspective distortion from the second code value and a second reference code value that is a preset reference, and the second reference code value is the measurement target. The coordinates on the pattern display unit corresponding to the multi-slit pattern observed in the absence of the pattern are shown. According to the said structure, distortion like the process part with a big curvature can be measured.

(9) The second reference code value is calculated in advance from the imaging result of the multi-slit pattern imaged by the imaging unit in the absence of the measurement target.
(10) The second reference code value is geometrically calculated from a positional relationship between the pattern display unit and the imaging unit. According to the above configuration, since it is not necessary to measure the second reference code value in advance, the time required for measurement can be shortened.
(11) The second reference code value is calculated in advance from a plurality of imaging results captured by the imaging unit in a state in which a measurement target that is a reference for perspective distortion is installed.

  (12) The perspective distortion calculation unit calculates a moving average value of the second code value and subtracts the calculated moving average value from the second code value or calculates a second-order difference value. The perspective distortion is calculated by obtaining this value. According to the above configuration, it is possible to exclude distortion such as a processed part having a large curvature, and it is possible to measure only local distortion.

(13) The perspective distortion calculation unit displays an image of the generation position and generation amount of the perspective distortion.
(14) The pattern display unit includes a projector capable of projecting an arbitrary pattern and a screen.
(15) The pattern display unit includes a flat display capable of displaying an arbitrary pattern.

  (16) A perspective distortion measuring method according to the present invention includes a pattern display step of sequentially displaying a plurality of stripe pattern images composed of a plurality of alternating light and dark stripes extending in one direction as a light and dark pattern, and the pattern An imaging step of imaging the plurality of light and dark patterns displayed in the display step via a measurement target; and a perspective distortion calculation step of calculating a perspective distortion of the measurement target surface based on a plurality of imaging results of the imaging step; In the perspective distortion calculation step, a first detection order that is a light and dark detection order is extracted for each pixel from the imaging results of the plurality of stripe pattern images among the plurality of imaging results, A first code value indicating a coordinate value on the pattern display unit is calculated for each pixel from a first detection order.

  (17) In the pattern display step, as the light / dark pattern, a multi-slit pattern comprising a plurality of stripes having a width smaller than the minimum width of the stripes of the plurality of types of stripe pattern images and extending in the one direction. Are further displayed by scanning in the direction perpendicular to the direction in which the stripes of the stripe pattern image extend, over the minimum width of the stripes of the stripe pattern image, and in the perspective distortion calculation step, a plurality of the imaging results are displayed. Among these, the second image indicating the coordinate value on the pattern display section that is finer than the first code value based on the captured image of the multi-slit pattern, the imaging timing of the captured image, and the first code value. The code value of is calculated.

(18) In the perspective distortion calculation step, perspective distortion is calculated from the second code value and a second reference code value that is a preset reference, and the second reference code value is The coordinates on the pattern display unit corresponding to the multi-slit pattern observed in the absence of a measurement object are shown.
(19) In the perspective distortion calculation step, a moving average value of the second code value is calculated, and the calculated moving average value is subtracted from the second code value or a second-order difference is calculated. The perspective distortion is calculated by obtaining the value.
(20) Further, the surface quality inspection method according to the present invention includes a resin or glass molded product derived from perspective distortion caused by at least one of injection molding, press molding, component mounting, assembly, heat treatment, and finished product inspection. Quality defects are inspected using the method for measuring perspective distortion described in any one of (16) to (19).

<Example 1>
Next, Example 1 performed by the present inventors will be described.
In Example 1, the perspective distortion of the measurement object 4 was measured using the perspective distortion measurement method according to the second and third embodiments. In Example 1, as shown in FIG. 20, a transparent acrylic plate whose inner side was bent in an arc shape on the left and right in the x-axis direction was used as the measurement object 4. In Example 1, perspective distortion was measured using a stripe pattern image and a multi-slit pattern in which stripes extend parallel to the x-axis direction, which is the horizontal direction of the imaging target region.

  The measurement result of the perspective distortion by the perspective distortion measurement method according to the second embodiment in Example 1 will be described.

  First, steps S200 to S214 were performed, and a stripe pattern image and a multi-slit pattern were projected and imaged. In Example 1, five captured images shown in FIGS. 21A to 21E were obtained by using the stripe pattern images shown in FIGS. 11A to 11E. Furthermore, a plurality of captured images as shown in FIGS. 21F to 21I were obtained by projecting the multi-slit pattern shown in FIG. 11F and continuously capturing images while scanning. The vertical and horizontal pixels of the captured image are 640 pixels × 480 pixels.

  Next, by performing the processing of steps S216 to S222, the difference between the second code value and the second reference code value was calculated as the distortion amount, and a distortion amount map was created. FIG. 22A is a diagram showing an image in which the first code value calculated in step S216 is mapped for each pixel with black and white shading. In FIG. 22B, when the second code value is calculated in step S218, the intermediate code value obtained by converting the imaging timing for each pixel having the maximum luminance is mapped for each pixel with black and white shading. FIG. FIG. 22C is a diagram illustrating an image in which the second code value calculated in step S218 is mapped for each pixel with black and white shading. FIG. 22D is a diagram illustrating an image in which the second reference code value calculated in advance in the second embodiment is mapped for each pixel with black and white shading. FIG. 22E shows an image in which the distortion amount calculated from the difference between the second code value and the second reference code value created in step S222 is mapped for each pixel in black and white shades. It is.

  In the first embodiment, as shown in FIGS. 22A and 22B, the second code value is calculated from the state in which the boundary portion divided by the first code value is known, thereby FIG. It was confirmed that a smooth code value distribution was obtained. Further, as shown in FIG. 22E, it was confirmed that the perspective distortion can be measured by calculating the difference between the second code value and the second reference code value.

  FIG. 23 is an enlarged view of FIG. The region indicated by the II line in FIG. 23 represents one pixel near the center in the x-axis direction arranged in the y-axis direction. 24 is a graph showing the distribution of code values in the region indicated by the line II in FIG. 23 in FIGS. 22C and 22D, and FIG. 25 is indicated by the line I-I in FIG. 24 shows the difference between the second code value and the second reference code value in each pixel on the horizontal axis in FIG. In FIG. 24, unlike the graph of the second reference code value, it was confirmed that the second code value fluctuated in the vicinity of 200 pixels on the horizontal axis, and perspective distortion occurred at this position. Furthermore, in FIG. 25, it was confirmed that a small distortion occurred in the vicinity of 130 pixels in addition to the distortion in the vicinity of 200 pixels on the horizontal axis. In the measurement result of Example 1, it was confirmed that the perspective distortion in the x-axis direction and the y-axis direction can be detected in units of pixels, so that the perspective distortion can be measured with high spatial resolution.

  Next, the measurement result of the perspective distortion by the perspective distortion measurement method according to the third embodiment in Example 1 will be described. In the measurement according to the third embodiment of Example 1, the measurement is performed in the same manner as in the second embodiment described above, and in step S320, the second reference code value calculated from the moving average is used, thereby local distortion. Extracted. FIG. 26 is a graph showing the second code value and the second reference code value calculated using the third embodiment in the same area as the area of the line I-I in FIG. FIG. 27 is a graph showing the difference between the second code value and the second reference code value in each pixel on the horizontal axis of FIG. FIG. 28 is a map of the distortion amount created in step S322, and is an image in which the distortion amount calculated from the difference between the second code value and the second reference code value is mapped for each pixel with black and white shading. FIG. As shown in FIGS. 26 to 28, it was confirmed that the perspective distortion can be measured with high spatial resolution in the measurement method according to the third embodiment as in the second embodiment.

<Example 2>
Next, Example 2 performed by the present inventors will be described.
In the second embodiment, as shown in FIG. 19, the use of a stripe pattern image and a multi-slit pattern extending in parallel to the y-axis direction is different from the first embodiment, but other conditions are the same as in the first embodiment. Perspective distortion was measured.

  The measurement result of the perspective distortion by the perspective distortion measurement method according to the second embodiment in Example 1 will be described.

  First, the processing of steps S300 to S314 was performed, and a stripe pattern image and a multi-slit pattern were projected and imaged. In Example 1, five captured images shown in FIGS. 29A to 29E were obtained by using the striped pattern images shown in FIGS. 19A to 19E. Furthermore, a plurality of captured images as shown in FIGS. 29F to 29I were obtained by projecting the multi-slit pattern shown in FIG. 19F and continuously capturing images while scanning.

  Next, by performing the processing of steps S316 to S322, the difference between the second code value and the second reference code value was calculated as the distortion amount, and a distortion amount map was created. FIG. 30A is a diagram showing an image in which the first code value calculated in step S316 is mapped for each pixel with black and white shading. In FIG. 30B, when the second code value is calculated in step S318, the intermediate code value obtained by converting the imaging timing for each pixel having the maximum luminance is mapped for each pixel with black and white shading. FIG. FIG. 30C is a diagram illustrating an image in which the second code value calculated in step S318 is mapped for each pixel with black and white shading. FIG. 30D is a diagram showing an image in which the second reference code value calculated in advance in the second embodiment is mapped for each pixel with black and white shading. FIG. 30E is a diagram showing an image in which the distortion amount calculated from the difference between the second code value and the second reference code value created in step S322 is mapped for each pixel with black and white shading. It is.

  In the second embodiment, as shown in FIGS. 30A and 30B, the second code value is calculated from the state in which the boundary portion divided by the first code value is known, and as shown in FIG. It was confirmed that a smooth code value distribution was obtained. Further, as shown in FIG. 30 (e), it was confirmed that the perspective distortion can be measured by calculating the difference between the second code value and the second reference code value. Further, in FIG. 30E, it was confirmed that the entire region where the measurement object 4 was imaged has a code value distribution in the x-axis direction. This is a measurement of the perspective distortion that occurs because the measurement object 4 has a shape that is bent into a circular arc with a large curvature.

  FIG. 31 is an enlarged view of FIG. The region indicated by the line II-II in FIG. 31 shows one pixel near the center in the y-axis direction aligned in the x-axis direction. 32 is a graph showing the distribution of code values in the region indicated by the line II-II in FIG. 31 in FIGS. 30C and 30D, and FIG. 33 is indicated by the line II-II in FIG. 33 shows the difference between the second code value and the second reference code value in each pixel on the horizontal axis in FIG. In FIG. 32, unlike the graph of the second reference code value, it was confirmed that the second code value fluctuated in the vicinity of 300 pixels on the horizontal axis, and perspective distortion occurred at this position. Further, in FIG. 33, it was confirmed that distortion occurred overall in the range of 100 pixels to 500 pixels in addition to the distortion around 300 pixels on the horizontal axis. In the measurement result of Example 2, it was confirmed that the perspective distortion in the x-axis direction and the y-axis direction can be detected in units of pixels, so that the perspective distortion can be measured with high spatial resolution. Furthermore, it has been confirmed that the perspective distortion caused by the measurement object 4 itself being curved with a large curvature can also be measured.

  Next, the measurement result of the perspective distortion by the perspective distortion measurement method according to the third embodiment in Example 2 will be described. In the measurement according to the third embodiment of Example 2, the measurement is performed in the same manner as in the second embodiment described above, and in step S320, the second reference code value calculated from the moving average is used, whereby local distortion is obtained. Extracted. FIG. 34 is a graph showing the second code value and the second reference code value calculated using the third embodiment in the same region as the region of the line II-II in FIG. FIG. 35 is a graph showing the difference between the second code value and the second reference code value in each pixel on the horizontal axis of FIG. FIG. 36 is a map of the distortion amount created in step S322, and is an image in which the distortion amount calculated from the difference between the second code value and the second reference code value is mapped for each pixel with black and white shading. FIG. As shown in FIG. 34, in the measurement method according to the third embodiment, it was confirmed that the perspective distortion can be measured with high spatial resolution, as in the second embodiment. Furthermore, as shown in FIG. 35 and FIG. 36, in the third embodiment, unlike the second embodiment, it is confirmed that the perspective distortion due to a large curvature that distorts the entire region where the measurement object 4 is imaged is not observed. did. This is because the use of the moving average value excludes the influence of a constant gentle distortion change due to a large curvature. For this reason, according to the measurement method according to the third embodiment, even when the measurement object 4 has a processed portion having a large curvature, the influence on the perspective distortion by the processed portion can be excluded, and the local It was confirmed that only the perspective distortion can be measured.

DESCRIPTION OF SYMBOLS 1: Image pick-up part 2: Pattern display part 2a: Projector 2b: Screen 2c: Flat display 3: Control part 3a: Sequence controller 3b: Perspective distortion calculation part 3c: Pattern projection part 3d: Image buffer 4: Measurement object 6a-6e : Stripe pattern image 6f: Multi-slit pattern 7a to 7e: Stripe pattern image 7f: Multi-slit pattern 11A to 11H: Region 11a to 11g: Boundary portion 12A to 12H: Region 12a to 12f: Boundary portion 14: Region 40: Region

Claims (9)

A pattern display unit that sequentially displays a plurality of stripe pattern images composed of a plurality of stripes of alternating light and darkness extending in one direction as light and dark patterns,
An imaging unit that images the plurality of bright and dark patterns displayed by the pattern display unit via a measurement target;
Based on a plurality of imaging results by the imaging unit, a perspective distortion calculation unit that calculates perspective distortion of the measurement target surface;
With
The perspective distortion calculation unit
From the imaging results of the plurality of stripe pattern images among the plurality of imaging results, the order of appearance of light and dark is extracted for each pixel of the imaging unit,
A first code value indicating a coordinate value on the pattern display unit is calculated for each pixel from the appearance order ,
Calculating perspective distortion from the difference between the first code value and a first reference code value that is a preset reference;
The perspective distortion measuring apparatus according to claim 1, wherein the first reference code value indicates coordinates on the pattern display unit corresponding to the stripe pattern image observed without the measurement target . The pattern display unit displays a plurality of types of stripe pattern images composed of stripes extending in the one direction and a plurality of types of stripe pattern images composed of stripes extending in a direction perpendicular to the one direction, respectively. And
The perspective distortion measuring device according to claim 1, wherein the perspective distortion calculation unit calculates perspective distortion in each of a direction orthogonal to the one direction and the one direction. The pattern display unit has a multi-slit pattern made up of a plurality of stripes extending in the one direction and having a width smaller than the minimum width of the stripes of the plurality of types of stripe pattern images. More than the minimum width of the stripe of the stripe pattern image in the direction perpendicular to the direction in which the stripe of the image extends is further displayed,
The perspective distortion calculation unit is finer than the first code value based on the captured image of the multi-slit pattern, the imaging timing of the captured image, and the first code value among the plurality of imaging results. perspective distortion of the measuring device according to claim 1 or 2, characterized in that calculating a second code value that indicates the coordinate values on Do the pattern display unit. The perspective distortion calculation unit calculates perspective distortion from the second code value and a second reference code value serving as a preset reference,
4. The measurement of perspective distortion according to claim 3 , wherein the second reference code value indicates coordinates on the pattern display unit corresponding to the multi-slit pattern observed in the absence of the measurement target. apparatus. The perspective distortion calculation unit calculates a moving average value of the second code value, and subtracts the calculated moving average value from the second code value or obtains a second-order difference value. The perspective distortion measuring apparatus according to claim 3 , wherein the perspective distortion is calculated. A pattern display step of sequentially displaying a plurality of stripe pattern images consisting of a plurality of stripes of light and dark alternating extending in one direction on the pattern display section as the light and dark patterns,
An imaging step of imaging the plurality of bright and dark patterns displayed in the pattern display step via a measurement object;
A perspective distortion calculation step of calculating a perspective distortion of the measurement target surface based on a plurality of imaging results of the imaging step;
With
During the perspective distortion calculation step,
From the imaging results of the plurality of stripe pattern images among the plurality of imaging results, the first appearance order that is the appearance order of light and dark is extracted for each pixel,
A first code value indicating a coordinate value on the pattern display unit is calculated for each pixel from the first appearance order ,
Calculating perspective distortion from the difference between the first code value and a first reference code value that is a preset reference;
The method for measuring a perspective distortion, wherein the first reference code value indicates coordinates on the pattern display unit corresponding to the stripe pattern image observed in a state where there is no measurement target . In the pattern display step, a multi-slit pattern consisting of a plurality of stripes extending in the one direction and having a width smaller than the minimum width of the stripes of the plurality of types of stripe pattern images as the light and dark patterns, Further displaying the stripe pattern image by scanning more than the minimum width of the stripe of the stripe pattern image in a direction perpendicular to the direction in which the stripe of the stripe pattern image extends,
Based on the first code value based on the captured image of the multi-slit pattern, the imaging timing of the captured image, and the first code value among the plurality of imaging results in the perspective distortion calculation step. 7. The method for measuring perspective distortion according to claim 6 , further comprising: calculating a second code value indicating a coordinate value on the fine pattern display unit. In the perspective distortion calculation step, a perspective distortion is calculated from the second code value and a second reference code value serving as a preset reference;
The measurement of perspective distortion according to claim 7 , wherein the second reference code value indicates coordinates on the pattern display unit corresponding to the multi-slit pattern observed without the measurement target. Method. During the perspective distortion calculated out step, wherein calculating a moving average value of the second code value, the calculated the moving average value, or subtract from the second code value, or a second-order difference The method for measuring a perspective distortion according to claim 7 , wherein the perspective distortion is calculated by obtaining.